This action might not be possible to undo. Are you sure you want to continue?

Table Of Contents

Quality Planning Tools ........................................................................................................................................................... 1 Overview........................................................................................................................................................................... 1 Run Chart ......................................................................................................................................................................... 2 Pareto Chart ..................................................................................................................................................................... 5 Cause-and-Effect Diagram ............................................................................................................................................... 9 Multi-Vari Chart............................................................................................................................................................... 13 Symmetry Plot ................................................................................................................................................................ 14 Control Charts...................................................................................................................................................................... 19 Overview......................................................................................................................................................................... 19 Box-Cox Transformation................................................................................................................................................. 25 Variables Charts for Subgroups ..................................................................................................................................... 27 Variables Charts for Individuals...................................................................................................................................... 90 Attributes Charts........................................................................................................................................................... 123 Time-weighted Charts .................................................................................................................................................. 152 Multivariate Charts........................................................................................................................................................ 177 Measurement Systems Analysis........................................................................................................................................ 209 Overview....................................................................................................................................................................... 209 Gage Run Chart ........................................................................................................................................................... 211 Gage Linearity and Bias Study ..................................................................................................................................... 214 Gage R&R Study (Crossed) ......................................................................................................................................... 216 Gage R&R Study (Nested) ........................................................................................................................................... 225 Attribute Agreement Analysis ....................................................................................................................................... 229 Attribute Gage Study (Analytic Method) ....................................................................................................................... 235 Process Capability ............................................................................................................................................................. 239 Overview....................................................................................................................................................................... 239 Individual Distribution Identification .............................................................................................................................. 245 Johnson Transformation............................................................................................................................................... 250 Capability Analysis - Normal......................................................................................................................................... 253 Capability Analysis - Between/Within ........................................................................................................................... 260 Capability Analysis - Nonnormal .................................................................................................................................. 264 Capability Analysis for Multiple Variables - Normal ...................................................................................................... 269 Capability Analysis for Multiple Variables - Nonnormal................................................................................................ 278 Capability Analysis - Binomial ...................................................................................................................................... 285 Capability Analysis - Poisson ....................................................................................................................................... 289 Capability Sixpack - Normal ......................................................................................................................................... 293 Capability Sixpack - Between/Within............................................................................................................................ 298 Capability Sixpack - Nonnormal ................................................................................................................................... 302 Index .................................................................................................................................................................................. 305

2003 Minitab Inc.

i

**Quality Planning Tools
**

Overview

Quality Planning Tool

Minitab offers several graphical tools to help you explore and detect quality problems and improve your process: Run charts detect patterns in your process data, and perform two tests for non-random behavior. Pareto charts help you identify which of your problems are most significant, so you can focus improvement efforts on areas where the largest gains can be made. Cause-and-effect (fishbone) diagrams can help you organize brainstorming information about the potential causes of a problem. Multi-vari charts present analysis of variance data in graphical form to give you a "look" at your data. Symmetry plots can help you assess whether your data come from a symmetric distribution.

Quality Tools

Stat > Quality Tools Choose one of the following: Run Chart Pareto Chart Cause-and-Effect Individual Distribution Identification Johnson Transformation Capability Analysis Normal Between/Within Nonnormal Multiple Variables (Normal) Multiple Variables (Nonnormal) Binomial Poisson Capability Sixpack Normal Nonnormal Between/Within Gage Study Gage Run Chart Gage Linearity and Bias Study Gage R&R Study (Crossed) Gage R&R Study (Nested) Attribute Gage Study (Analytic Method) Attribute Agreement Analysis Multi-Vari Chart Symmetry Plot

2003 Minitab Inc.

1

Quality Control

**Examples of Quality Tools
**

The following examples illustrate how to use the various quality tools. Choose an example below: Run Chart Pareto Chart Cause-and-Effect Individual Distribution Identification Johnson Transformation Capability Analysis Capability Sixpack Gage Run Chart Gage Linearity and Bias Study Gage R&R Study (Crossed) (ANOVA method) Gage R&R Study (Crossed) (X-bar and R method) Gage R&R Study (Nested) Attribute Gage Study (Analytic Method) Attribute Agreement Analysis Multi-Vari Chart Symmetry Plot

**References − Quality Tools
**

[1] [2] [3] J.D. Gibbons (1986). "Randomness, Tests of," Encyclopedia of Statistical Sciences, 7, 555−562. T.P. Ryan (1989). Statistical Methods for Quality Improvement. John Wiley & Sons. W.A. Taylor (1991). Optimization & Variation Reduction in Quality. McGraw-Hill, Inc.

Run Chart

Run Chart Overview

Variation occurs in all processes. Common cause variation is a natural part of the process. Another type of variation, called special causes, comes from outside the system and causes recognizable patterns, shifts, or trends in the data. The run chart shows if special causes are influencing your process. A process is in control when only common causes affect the process output. Run Chart performs two tests for randomness that provide information on the non-random variation due to trends, oscillation, mixtures, and clustering. For details, see Using the Tests for Randomness.

Run Chart

Stat > Quality Tools > Run Chart Use Run Chart to look for evidence of patterns in your process data, and perform two tests for non-random behavior. Run Chart plots all of the individual observations versus the subgroup number, and draws a horizontal reference line at the median. When the subgroup size is greater than one, Run Chart also plots the subgroup means or medians and connects them with a line. The two tests for nonrandom behavior detect trends, oscillation, mixtures, and clustering in your data. Such patterns suggest that the variation observed is due to special causes − causes arising from outside the system that can be corrected. Common cause variation is variation that is inherent or a natural part of the process. A process is in control when only common causes affect the process output.

**Dialog box items
**

Data are arranged as Single column: Choose if data is in one column. Enter a column. Subgroup size (use a constant or an ID column): Enter the subgroup size (for equally sized subgroups) or a column of subscripts (for unequally sized subgroups). Subgroups across rows of: Choose if subgroups are arranged in rows across several columns. Enter the columns.

2

2003 Minitab Inc.

enter a series of columns in Subgroups across rows of. Plot subgroup means: Choose to plot the subgroup means as points on the graph. Minitab uses the points to count the number of runs in tests for randomness. 3 . See [1] for details. In Subgroup size. Assume the test for randomness in the examples is significant at an α-value of 0. then uses the normal distribution to obtain p-values. Subgroup data can be structured in a single column or in rows across several columns.Quality Planning Tools For data in subgroups You can plot either the subgroup means or medians as points on the graph. then set up a second column of subgroup indicators. the data exhibit random behavior. 2003 Minitab Inc. the null hypothesis is that the data have a random sequence. use any dialog box items. When either is smaller than your α-value (significance level). When you have subgroups of unequal size. See Data for examples. <Options> Data − Run Chart You can use individual observations or subgroup data. Do one of the following: • When subgroups or individual observations are in one column. the tests are based on either the subgroup means (the default) or the subgroup medians. If you like. When the subgroup size is greater than one. Test for randomness Number of runs about the median Condition More runs observed than expected Fewer runs observed than expected Number of runs up or down More runs observed than expected Fewer runs observed than expected Indicates Mixed data from two population Clustering of data Oscillation − data varies up and down rapidly Trending of data Both tests are based on individual observations when the subgroup size is equal to one. Interpreting the tests for randomness A normal pattern for a process in control is one of randomness. Dialog box items Title: Type the desired text to replace the default title with your own custom title. enter the data column in Single column. then click OK. Plot subgroup medians: Choose to plot the subgroup medians as points on the graph. The following table illustrates what the two tests for randomness can tell you. With both tests. 3 Run Chart − Options Stat > Quality Tools > Run Chart > Options You can create your own custom title. For individual observations. reject the hypothesis of randomness. enter the subgroups in a single column. enter a subgroup size of 1.05. Run Chart converts the observed number of runs into a test statistic that is approximately standard normal. The two p-values correspond to the one-sided probabilities associated with the test statistic. • When subgroups are in rows. To make a Run Chart 1 2 Choose Stat > Quality Tools > Run Chart. enter a subgroup size or column of subgroup indicators. If only common causes of variation exist in your process. See Interpreting the test for number of runs about the median and Interpreting the test for number of runs up or down for more details.

Quality Control

**Interpreting the test for number of runs about the median
**

This test is based on the total number of runs that occur both above and below the median. A run, with this test, is one or more consecutive points on the same side of the median. When the points are connected with a line, a run ends when the line crosses the median. A new run begins with the next plotted point. The test for the number of runs about the median is sensitive to two types of nonrandom behavior − mixtures and clustering: • • An observed number of runs that is statistically greater than expected supports the alternative of mixing (which corresponds to a right-tail rejection region). An observed number of runs that is statistically less than expected supports the alternative of clustering (which corresponds to a left-tail rejection region).

**Interpreting the test for number of runs up or down
**

This test is based on the number of runs up or down − increasing or decreasing. A run, with this test, is one or more consecutive points in the same direction. A new run begins each time there is a change in the direction (either ascending or descending) in the sequence of data. For example, with increasing values, a run up continues until a value is larger than the next point, then a run down begins. The test for the number of runs up or down is sensitive to two types of nonrandom behavior − oscillation and trends. • • An observed number of runs that is statistically greater than expected supports the alternative of oscillation (which corresponds to a right-tail rejection region). An observed number of runs that is statistically less than expected supports the alternative of trends (which corresponds to a left-tail rejection region).

**Comparing Run Chart and Runs Tests
**

Minitab provides two tests for randomness: Runs Test and Run Chart. Use Runs Test with individual observations and tests for randomness without looking for specific nonrandom patterns. Runs Test bases its test for randomness on the number of runs above and below the mean by default, but you can specify the median. When the subgroup size is one, and you specify the median (instead of the default mean), Runs Test and Run Chart perform the same test. Run Chart displays the one-sided probabilities associated with the test statistic. In contrast, Runs Test uses a two-sided test. Thus, the p-value reported by Runs Test is approximately twice as large as the smaller p-value reported by Run Chart.

**Example of Run Chart
**

Suppose you work for a company that produces several devices that measure radiation. As the quality engineer, you are concerned with a membrane type device's ability to consistently measure the amount of radiation. You want to analyze the data from tests of 20 devices (in groups of 2) collected in an experimental chamber. After every test, you record the amount of radiation that each device measured. As an exploratory measure, you decide to construct a run chart to evaluate the variation in your measurements. 1 2 3 4 Open the worksheet RADON.MTW. Choose Stat > Quality Tools > Run Chart. In Single column, enter Membrane. In Subgroup size, enter 2. Click OK.

4

2003 Minitab Inc.

Quality Planning Tools Graph window output

**Interpreting the results
**

The test for clustering is significant at the 0.05 level. Because the probability for the cluster test (p = 0.02209) is less than the α value of 0.05, you can conclude that special causes are affecting your process, and you should investigate possible sources. Clusters may indicate sampling or measurement problems. Note The 0.05 level of significance was chosen for illustrative purposes, because it is commonly used in many fields. You can evaluate the significance of tests for nonrandom patterns at any level. When the p-value displayed is less than the chosen level of significance, you reject the null hypothesis − a random sequence of data − in favor of one of the alternatives. See Interpreting the tests for randomness for a complete discussion.

Pareto Chart

Pareto Chart

Stat > Quality Tools > Pareto Chart Pareto charts are a type of bar chart in which the horizontal axis represents categories of interest, rather than a continuous scale. The categories are often "defects." By ordering the bars from largest to smallest, a Pareto chart can help you determine which of the defects comprise the "vital few" and which are the "trivial many." A cumulative percentage line helps you judge the added contribution of each category. Pareto charts can help to focus improvement efforts on areas where the largest gains can be made. Pareto chart can draw one chart for all your data (the default), or separate charts for groups within your data.

**Dialog box items
**

Chart defects data in: Choose if the data is entered in raw form and enter the column containing the data. The defects can be either text or numeric data, so you can use the defect names or numeric codes representing the defects. If you use text, each defect name can have up to 72 characters, though Minitab uses only the first 15, so be sure defect names are distinct within the first 15 characters. Minitab also creates a bar for missing data. By variable in__(optional): Enter the column containing the indicator variable you want to use to designate groups. The indicator column designates which group each observation belongs in. Thus, the By variable in: column must be the same length as the data column. Note, the column can contain text or numeric data. Numeric data does not need to be integers. The maximum number of levels of the indicator variable is 25. Default (all on same graph, same ordering of bars): Choose to display a matrix of Pareto diagrams, one for each group, using the same ordering of bars on each chart. The order of the bars in all of the Pareto diagrams is determined by the first group. That is, the first group's bars will be represented in Pareto order, from largest to smallest. All other groups will have their bars represented in the same order as the first group, which in most cases will mean that bars in subsequent groups will not be in Pareto order. However, this can be useful for comparing importance of categories relative to a baseline, which is the first group. One chart per graph, same ordering of bars: Choose to draw a separate full-page Pareto diagram for each group using the same ordering of bars on each page. The order of the bars in all of the Pareto diagrams is determined by

2003 Minitab Inc.

5

Quality Control

the first group. All other groups will have their bars represented in the same order as the first group, which in most cases will mean that bars in subsequent groups will not be in Pareto order. However, this can be useful for comparing importance of categories relative to a baseline, which is the first group. Each chart displays in a separate Graph window. One chart per graph, independent ordering of bars: Choose to order the bars in each group's Pareto diagram independently of all other groups. In other words, the bars in each Pareto diagram will be in Pareto order, which means that in most cases the order will be different between groups. Each chart displays in a separate Graph window. Chart defects table: Choose if the data is entered as defect names or codes and frequency counts rather than raw data. You can use any numeric data in the Frequencies in column. If you want to obtain a Pareto chart for something like variance components, where the numbers are not integers, convert them to integers by multiplying by the appropriate value. Labels in: Enter the column containing the defect names or codes. Frequencies in: Enter the column containing the frequency of occurrence corresponding the names or codes specified in the Labels in box. Combine defects after the first__% into one: Enter the cumulative percentage for which you want to generate bars. Minitab generates bars for defects until the cumulative percentage surpasses the % specified, then groups the remaining defects into a bar labeled "Others." The default cumulative percentage is set at 95%. <Options>

**Data − Pareto Chart
**

You can structure your data in one of two ways: • • As one column of raw data, where each observation is an occurrence of a type of defect As two columns: one column of defect names and a corresponding column of counts

**To make a Pareto chart
**

1 2 Choose Stat > Quality Tools > Pareto Chart. Do one of the following: • If you have a column of raw data, enter the column in Chart defects data in. • If you have a column of defect names and a column of counts: − − − 3 Choose Chart defects table. In Labels in, enter a column of defect names. In Frequencies in, enter a column of counts.

If you like, use any dialog box items, then click OK.

**Data limitations for Pareto chart
**

If you use text for the defects, each defect name can have as many as 72 characters. When using Chart defects table option, the values in the counts column must be integers. If you want to obtain a Pareto chart for something like variance components, where the numbers are not integers, convert them to integers by multiplying by the appropriate value.

**Missing Data in Pareto Charts
**

When you use Pareto Chart with raw data, counts the number of occurrences of each unique value in the column and displays a bar for each unique value. With text, a blank line is a missing value. For numeric data, a missing value is represented as an ∗. In either case, Minitab interprets the missing value as a unique value of the defect column. Pareto Chart includes a missing data bar as part of the plot. If you do not want a "missing" bar in the chart, use Data > Delete Rows or Data > Copy > Columns to Columns to remove rows with missing data. When you use Chart Defects Table option, the defect label is matched up with the count in the same row of the frequencies column. If a value in the frequencies column is missing (∗), Minitab removes that row of data. If a row in the names column is blank, Pareto Chart includes the data in the chart when count in that row is not missing. When the chart is drawn, no name appears for the defect.

6

2003 Minitab Inc.

Quality Planning Tools

**Pareto Chart − Options
**

Stat > Quality Tools > Pareto Chart > Options Allows you to specify the x and y axis labels, not display the cumulative percent, and type your own custom title.

**Dialog box items
**

X axis label: Type the x axis label. If you do not specify a label, the default label is "Defect." Y axis label: Type the y axis label. If you do not specify a label, the default label is "Count." Do not chart cumulative percent: Check to not show cumulative percent symbols, connecting lines, and percent scale. Title: Type the desired text to replace the default title with your own custom title.

**Example of using Pareto chart with raw data
**

The company you work for manufactures metal bookcases. During final inspection, a certain number of bookcases are rejected due to scratches, chips, bends, or dents. You want to make a Pareto chart to see which defect is causing most of your problems. First you count the number of times each defect occurred, then you enter the name of the defect each time it occurs into a worksheet column called Damage. 1 2 3 Open the worksheet EXH_QC.MTW. Choose Stat > Quality Tools > Pareto Chart. Choose Chart defects data in and enter Damage. Click OK.

Graph window output

**Interpreting the results
**

Focus on improvements to scratches and chips because 75% of the damage is due to these defects.

**Example of using Pareto chart with count data
**

Suppose you work for a company that manufactures motorcycles. You hope to reduce quality costs arising from defective speedometers. During inspection, a certain number of speedometers are rejected, and the types of defects recorded. You enter the name of the defect into a worksheet column called Defects, and the corresponding counts into a column called Counts. You know that you can save the most money by focusing on the defects responsible for most of the rejections. A Pareto chart will help you identify which defects are causing most of your problems. 1 2 3 4 Open the worksheet EXH_QC.MTW. Choose Stat > Quality Tools > Pareto Chart. Choose Chart defects table. In Labels in, enter Defects. In Frequencies in, enter Counts. Click OK.

2003 Minitab Inc.

7

Quality Control Graph window output

**Interpreting the results
**

Focus on improving the number of missing screws because over half of your speedometers are rejected due to this defect.

**Example of using Pareto chart with a by variable
**

Imagine you work for a company that manufactures dolls. Lately, you have noticed that an increasing number of dolls are being rejected at final inspection due to scratches, peels, and smudges in their paint. You want to see if a relationship exists between the type and number of flaws, and the work shift producing the dolls. 1 2 3 Open the worksheet EXH_QC.MTW. Choose Stat > Quality Tools > Pareto Chart. Choose Chart defects data in and enter Flaws in the text box. In BY variable in, enter Period. Click OK.

8

2003 Minitab Inc.

Quality Planning Tools Graph window output

**Interpreting the results
**

The night shift is producing more flaws overall. Most of the problems are due to scratches and peels. You may learn a lot about the problem if you examine that part of the process during the night shift.

**Cause-and-Effect Diagram
**

Cause-and-Effect Diagram Overview

A cause-and-effect, or fishbone, diagram depicts potential causes of a problem. The problem (effect) displays on the right side and the list of causes on the left side in a treelike structure. The branches of the tree are often associated with major categories of causes. Each branch has a listing of more specific causes in that category. You can also add sub-branches to any branch. Fishbone diagrams are a convenient tool for organizing information about causes of a problem. Although there is no "correct" way to construct a fishbone diagram, some specific types lend themselves well to many different situations. One of these is the "5M" diagram, so called because five of the categories on the branches begin with the letter M ("Personnel" is also referred to as "Man"). Minitab draws a 5M diagram by default.

**Cause-and-Effect Diagram
**

Stat > Quality Tools > Cause-and-Effect Use a cause-and-effect (fishbone or Ishikawa) diagram to organize brainstorming information about potential causes of a problem. Diagramming helps you to see relationships among potential causes. You can draw a blank diagram, or a diagram filled in as much as you like, including sub-branches. Although there is no "correct" way to construct a fishbone diagram, some types lend themselves well to many different situations. Note You can change or add branches or sub-branches after you create a cause-and-effect diagram. See Add/ Change Branches.

**Dialog box items
**

Causes In column: Choose if the causes are listed in a column, then enter the list of causes for the corresponding branch of the diagram. Entries in the columns can be up to 72 characters wide. Odd branches are on top (5, 3, then 1, left to right), and even branches are on the bottom (6, 4, then 2). To display diagram with the main branches but no causes, do not select any columns. Constants: Choose to enter the causes as constants, then type the list of causes for the corresponding branch of the diagram. Use a blank space between causes. Odd branches are on top (5, 3, then 1, left to right), and even branches are on the bottom (6, 4, then 2). To display diagram with the main branches but no causes, do not select any columns. Label: Type the label you want to display to change the default branch labels. By default, the labels for branches 1 through 6 are (in order) Personnel, Machines, Materials, Methods, Measurements, and Environment. Odd branches are

2003 Minitab Inc.

9

Quality Control

on top (5, 3, then 1, left to right), and even branches are on the bottom (6, 4, then 2). To display a blank diagram, check Do not label branches. Sub: Click to add sub-branches for the corresponding branch. Effect: Type the desired text to display a label for the effect or problem that you are trying to solve. The text is displayed to the right of the diagram. You may use up to 72 characters. Title: Type the desired text to replace the default title with your own custom title. Do not label the branches: Check to suppress the labeling of branches. Do not display empty branches: Check to suppress empty branches for which you have not specified data. <Sub-branches>

**Data − Cause-and-Effect Diagram
**

If you want to enter causes on the branches or sub-branches of the diagram, create a column of causes for each branch and sub-branch. There is no limit on the number of causes you can list for a branch. You can reduce the font to incorporate the desired number of causes.

**To make a cause-and-effect diagram
**

1 2 Choose Stat > Quality Tools > Cause-and-Effect. If you like, use any dialog box items, then click OK.

**Cause-and-Effect Diagram − Sub-Branches
**

Stat > Quality Tools > Cause-and-Effect > Sub-Branches Add sub-branches to any branch in the cause-and-effect diagram.

**Dialog box items
**

Causes In column: Choose if the causes are listed in a column, then enter the list of causes for the corresponding sub-branch of the diagram. Entries in the columns can be up to 72 characters wide. Odd branches are on top (5, 3, then 1, left to right), and even branches are on the bottom (6, 4, then 2). To display diagram with the main branches but no causes, do not select any columns. Constants: Choose to enter the causes as constants, then type the list of causes for the corresponding branch of the diagram. Use a blank space between causes. Odd branches are on top (5, 3, then 1, left to right), and even branches are on the bottom (6, 4, then 2). To display diagram with the main branches but no causes, do not select any columns.

**Example of using cause-and-effect to draw a blank diagram
**

Using a Pareto chart, you discovered that your parts were rejected most often due to surface flaws. This afternoon, you are meeting with members of various departments to brainstorm potential causes for these flaws. Beforehand, you decide to print a cause-and-effect (fishbone) diagram to help organize your notes during the meeting. The example below illustrates how to generate a blank cause-and-effect diagram. 1 2 Choose Stat > Quality Tools > Cause-and-Effect. Check Do not label the branches, then click OK.

10

2003 Minitab Inc.

Quality Planning Tools Graph window output

**Example of using cause-and-effect to draw a diagram with empty branches
**

Using a Pareto chart, you discovered that your parts were rejected most often due to surface flaws. This afternoon, you are meeting with members of various departments to brainstorm potential causes for these flaws. Beforehand, you decide to print a cause-and-effect (fishbone) diagram to help organize your notes during the meeting. The example below illustrates how to generate a cause-and-effect diagram with empty branches and custom title. 1 2 Choose Stat > Quality Tools > Cause-and-Effect. In Title, type Sample FISHBONE Diagram, then click OK.

Graph window output

**Example of using cause-and-effect to draw a completed diagram
**

Using a Pareto chart, you discovered that your parts were rejected most often due to surface flaws. This afternoon, you are meeting with members of various departments to brainstorm potential causes for these flaws. Beforehand, you decide to print a cause-and-effect (fishbone) diagram to help organize your notes during the meeting. The example below illustrates how to generate a complete cause-and-effect diagram.

2003 Minitab Inc.

11

<Sub-branches> To add or change branches 1 2 3 After creating a cause-and-effect diagram. In Effect. Material. To display a blank diagram. the labels for branches 1 through 6 are (in order) Personnel. Do not display empty branches: Check to suppress empty branches for which you have not specified data. 12 2003 Minitab Inc. Add/Change Branches − Sub-Branches Editor > Graph Options > Sub-branches Add or change the sub-branches on the cause-and-effect diagram. Measurements. . Graph window output Add/Change Branches Editor > Graph Options Add or change the branches on the cause-and-effect diagram. Enter Man. choose In column for rows 1-6. Odd branches are on top (5.Quality Control 1 2 3 4 5 Open the file SURFACEFLAWS. Odd branches are on top (5. then click OK. Choose Stat > Quality Tools > Cause-and-Effect. Methods. Click OK. 4. Method. left to right). Label: Type the label you want to display to change the default branch labels. Materials. Under Causes. Under Causes. Machine. then 1. Dialog box items Causes: Type a list of causes for the corresponding branch of the diagram. check Do not label branches. 3. then 2). double-click within any row to add or change a branch.MTW. choose Editor > Graph Options. then 1. Machines. and even branches are on the bottom (6. Measure. Sub-branches: Click to add or change sub-branches for the corresponding branch. use any dialog box items. then 2). Do not label the branches: Check to suppress the labeling of branches. By default. left to right). 3. If you like. 4. type Surface Flaws. respectively. Entries in the columns can be up to 72 characters wide. and even branches are on the bottom (6. and Enviro in rows 1 through 6. and Environment.

enter a factor level column. Connect means for Factor 3: Check to connect the factor level 3 means with a line. Each row contains the data for a single observation. Click OK in each dialog box. <Options> Data − Multi Vari Chart You need one numeric column for the response variable and up to four numeric. or Factor 4 as needed. double-click within any row to add or change a branch.Quality Planning Tools Dialog box items Causes: Type a list of causes for the corresponding sub-branch of the diagram. text. Under Causes. If you wish. then 2). Text categories (factor levels) are processed in alphabetical order by default. or date/time factor columns. Choose Sub-branches for the branch you want to change. Odd branches are on top (5. To add or change sub-branches 1 2 3 4 After creating a cause-and-effect diagram. If you have more than one factor. Multi-vari charts are a way of presenting analysis of variance data in a graphical form providing a "visual" alternative to analysis of variance. 2003 Minitab Inc. use any dialog box items. 4. enter columns in Factor 2. Multi-Vari Chart − Options Stat > Quality Tools > Multi-Vari Chart > Options Allows you to choose output display options. Dialog box items Response: Enter the column containing the response (measurement) data. then click OK. Factor 3: Enter an additional factor level column. Factor 4: Enter an additional factor level column. Multi-Vari Chart Multi-Vari Chart Stat > Quality Tools > Multi-Vari Chart Minitab draws multi-vari charts for up to four factors. In Response. Entries in the columns can be up to 72 characters wide. To draw a Multi-Vari Chart 1 2 3 4 5 Choose Stat > Quality Tools > Multi-Vari Chart. enter the column containing the response (measurement) data. and even branches are on the bottom (6. Factor 1: Enter the factor level column. Factor 3. 13 . Minitab automatically omits missing data from the calculations. The chart displays the means at each factor level for every factor. In Factor 1. then 1. If you like. 3. you can define your own order − see Ordering Text Categories. Connect means for Factor 2: Check to connect the factor level 2 means with a line. These charts may also be used in the preliminary stages of data analysis to get a look at the data. left to right). Factor 2: Enter an additional factor level column. Dialog box items Display options Display individual data points: Check to draw individual data points on the chart Connect means for Factor 1: Check to connect the factor level 1 means with a line. choose Editor > Graph Options.

such as nonparametric methods. .MTW. so having data from a symmetric distribution is often sufficient. Example of a Multi-Vari Chart You are responsible for evaluating the effects of sintering time on the compressive strength of three different metals. 150 minutes. The greatest compressive strength for Metal Type 1 is obtained by sintering for 100 minutes. <Options> 14 2003 Minitab Inc. However. Symmetry Plot Symmetry Plot Stat > Quality Tools > Symmetry Plot Symmetry plots can be used to assess whether sample data come from a symmetric distribution. many procedures are robust to violations of normality. In Factor 2. In Response. Dialog box items Variables: Enter the columns containing the numeric data you want to plot. Therefore. Choose Stat > Quality Tools > Multi-Vari Chart. Click OK. 1 2 3 4 Open the worksheet SINTER. a symmetry plot is a useful tool in many circumstances. Other procedures. Before you engage in a full data analysis. To quantify this interaction. Graph window output Interpreting the results The multi-vari chart indicates that an interaction exists between the type of metal and the length of time it is sintered. enter SinterTime. Compressive strength was measured for five specimens for each metal type at each of the sintering times: 100 minutes. you could further analyze this data using techniques such as analysis of variance or general linear model. enter Strength. and 200 minutes. Many statistical procedures assume that data come from a normal distribution. for Metal Type 2 sintering for 150 minutes. assume symmetric distributions rather than normal distributions. In Factor 1.Quality Control Title: Type the desired text to replace the default title with your own custom title. enter MetalType. and for Metal Type 3 sintering for 200 minutes. you want to view the data to see if there are any visible trends or interactions by creating a multi-vari chart.

you can expect to see runs of points above or below the line. you can assess the degree of symmetry present in the data. The important thing to look for is whether the points remain close to or parallel to the line. If you enter more than one data column. Minitab draws a line on the plot to represent exact X-Y equality (a perfectly symmetric sample). Interpreting the Symmetry Plot When the sample data follow a symmetric distribution. 15 . enter the columns containing the numeric data you want to plot. Even with normally distributed data. 2003 Minitab Inc. If you like. then click OK. use any dialog box items. By comparing the data points to the line. Minitab draws a separate symmetry plot for each column. the closer the points will be to the line. the X and Y coordinates will be approximately equal for all points and the data will fall in a straight line. The more symmetric the data. versus the points diverging from the line. Minitab automatically omits missing data from the calculations.Quality Planning Tools Data − Symmetry Plot The data columns must be numeric. To draw a symmetry plot 1 2 3 Choose Stat > Quality Tools > Symmetry Plot. • Data points diverging below the line indicate skewness to the right. In Variables. You can detect the following asymmetric conditions: • Data points diverging above the line indicate skewness to the left.

. In Variables. Choose Stat > Quality Tools > Symmetry Plot. Example of a Symmetry Plot Before doing further analyses. Interpreting a plot with too few data points may lead to incorrect conclusions. Click OK. 16 2003 Minitab Inc. Caution As rule of thumb. Dialog box items Title: Type the desired text to replace the default title with your own custom title. enter Faults. you would like to determine whether or not the sample data come from a symmetric distribution. 1 2 3 Open the worksheet EXH_QC. Symmetry Plot − Options Stat > Quality Tools > Symmetry Plot > Options Let you replace the default title with your own title.Quality Control • Points far away from the line in the upper right corner (where distances are large) indicate some degree of skewness in the tails of the distribution.MTW. you should have at least 25 to 30 data points.

You can also see this skewness in the histogram. These points indicate skewness in the left tail of the distribution. 17 . 2003 Minitab Inc.Quality Planning Tools Graph window output Interpreting the results Notice the few points above the line in the upper right corner.

.

3σ below the center line by default Special causes result in variation that can be detected and controlled. such as subsetting your data and changing the control chart's title. A process is in control when points fall within the bounds of the control limits. shift. Minitab draws the: • • • Center line at the average of the statistic by default Upper control limit. Minitab offers a variety of options for customizing your control charts: • • • See Control Chart Options for options specific to control charts. See Graph Editing Overview for options that are available after graph creation. Common cause variation. you can use control charts to estimate process parameters needed to determine capability. such as using stages and tests for special causes. 3σ above the center line by default Lower control limit. is inherent in the process. such as changing font size and figure location. individual observation. See Control Chart Display Options for options that are available before graph creation. When a process is in control. 19 . 2003 Minitab Inc. versus sample number or time. A process is in control when only common causes − not special causes − affect the process output. on the other hand. Minitab plots a process statistic. or day of the week.Control Charts Overview Control Charts Stat > Control Charts Choose from the following options: Box-Cox Transformation: Performs a Box-Cox procedure on process data used in control charts Minitab draws a wide variety of control charts: Variables chart for subgroups: Creates control charts for measurement data in subgroups Variables charts for individuals: Creates control charts for measurement data for individuals Attributes charts: Creates control charts for data with counts of defects or non-conformities in a sample or the presence or absence of characteristics or attributes Time-weighted charts: Creates control charts for time-weighted data Multivariate charts: Creates control charts for multivariate data Control Charts Overview data You can use control charts to track process statistics over time and to detect the presence of special causes. You can also perform a Box-Cox transformation on non-normal data. Examples include differences in supplier. weighted statistic. such as a subgroup mean. Use the tests for special causes offered with Minitab's control charts to detect nonrandom patterns in your data. and the points do not display any nonrandom patterns. or number of defects.

or a column containing different values for each stage. X-bar and S.. different value for each stages. different values for each stage. Multiple Graphs: Use to set the same y-scale for different models. gridlines and reference lines. This option is available only for the Zone chart. EWMA. you can select options to customize your chart: • • • • • • • • • • • Parameters: Use to enter historical data for estimation. for instance. X-bar and R. Confidence Limits: Use to specify confidence limits for multivariate control charts.Quality Control See Control Charts for the types of control charts available. U Historical σ: All control charts except Z-MR and I-MR-R/S For variables control charts. Historical charts with other control chart options In a sense you are creating two or more control charts in one: so how do the control chart options work with these "separate charts"? Most of the time what goes for one goes for all. Box-Cox: Use the Box-Cox transformation to normalize your data. Moving Average. Alternatively. Historical standard deviation. Weights/Reset: Use to control the weight assigned to each zone and to reset the cumulative score after each signal. Minitab estimates them from the data.. Enter one value to be used for all stages. Display: Use to display the control chart by stages or subgroup. and display the test results in the Session window. Use historical values Charts Historical µ: X-bar. Multiple Graphs is available for all control charts except the Multivariate charts and the T2 and generalized variance chart. I-MR-R/S. Tests: Use to choose tests for special causes. There are exceptions. Stages: Use to display stages or groups in your process. Historical mean. axes and ticks. three lines are drawn for each stage. Individuals.. however. Control Charts Options In each control chart's data source dialog box. C. Control Charts Display Options You can choose from the following options to customize the appearance of the control charts before creating your graphs: • • • • Scale: Use to customize the time scale. Each time a new stage begins. Storage: Use to store your choice of statistics in the worksheet. Zone. Labels: Use to customize the title and/or footnotes. or your goals. This option is available only for the CUSUM chart. Data Options: Use to select which rows in your Minitab worksheet to use in the control chart. you can enter known process parameters. estimates obtained from past data. respectively.. the tests restart. If µ and σ are not specified. The mean and standard deviation of a population distribution are denoted by mu (µ) and sigma (σ). The tests for special causes. or a column containing different values for each stage. 20 2003 Minitab Inc. Estimate: Use to omit certain subgroups for the analysis or to choose an estimation method. and Historical p Box-Cox power transformation You can. split the chart. Enter one value to be used for all stages. Some options offer choices: With this option. Plan/Type: Use to specify the CUSUM plan and type. I-MR. If you specify that you want three control limit lines drawn. S Limits: Use to determine standard deviation limit positions and to place a lower or upper limits. the process is assumed to produce data from a stable population that often follows a normal distribution. for instance. are performed independently for each stage. .

I-MR-R/S. Unbiasing estimates Check Use unbiasing constants to use unbiased values for the estimation of σ. Capability Sixpack (Normal). is to estimate σ with a pooled standard deviation. CUSUM When subgroup sizes are not equal. Capability Analysis Multiple Variables (Normal). Choose Rbar to base the estimate on the average of the subgroup ranges. EWMA. EWMA. Capability Analysis Multiple Variables (Normal). > control chart Options > S Limits Charts X-bar and R. I-MR. For the CUSUM chart. and Moving Range 2003 Minitab Inc. Capability Sixpack (Normal). Capability Analysis Multiple Variables (Normal). Zone. pooled standard deviation. Capability Sixpack (Between/Within). The center line of charts for ranges and standard deviations also varies with the subgroup size. since consecutive values have the greatest chance of being alike. except for the R. NP. If the sizes do not vary much. Capability Analysis (Normal). 21 . By default. use Median moving range or Square root of MSSD. Capability Analysis (Normal). you may want to force these lines to be constant. Zone. EWMA. Individuals. CUSUM. Choose Sbar to base your estimate on the average of the subgroup standard deviations. Click Use Moving range of length to change the length of the moving range. I-MR. EWMA. Sbar: X-bar. CUSUM. I-MR. Capability Sixpack (Normal). See [2] for a discussion of the relative merits of each estimator.. X-bar and S. Force control limits and center line to be constant . X-bar and S. Pooled standard deviation: X-bar. When you use this option. Xbar and R. I-MR-R/S. Capability Analysis (Normal). Zone. Capability Analysis Multiple Variables (Normal). you can force the control limits to be constant using CUSUM Options > Plan/Type. Capability Analysis (Between/Within). Capability Sixpack (Between/Within). Capability Analysis (Between/Within). CUSUM. X-bar. you could enter the average sample size as the subgroup size. R. Capability Sixpack (Normal). Capability Sixpack (Between/Within). S. Moving Range. the moving range is of length 2. Capability Analysis (Normal). Moving Average. Z-MR. Capability Analysis (Between/Within). Average moving range: Individuals. and square root of MSSD). EWMA. Control how the standard deviation is estimated Minitab has several methods of estimating σ. Moving Range. Zone. Capability Sixpack (Between/Within). R. Eight tests for special causes Charts X-bar and R. Alternatively. I-MR-R/S. CUSUM.Control Charts This option overrides any options relating to estimating µ or σ from the data − specifically. Capability Sixpack (Between/Within). but will vary with the subgroup size. Median moving range: Individuals. X-bar and S. Capability Sixpack (Normal). the average of the moving range divided by an unbiasing constant. Individual observations The estimate of σ is based on MRbar / d2. Omit the following samples when estimating parameters and any of the Methods for estimating standard deviation. Capability Sixpack (Normal). Moving Average. The pooled standard deviation is the most efficient method of estimating the standard deviation when you can assume constant variation across subgroups. EWMA. X-bar and S. Capability Sixpack (Normal). Zone. S. Zone. depending on whether your data is in subgroups or individual observations. S. the plot points themselves are not changed. Moving Average. I-MR-R/S. Square root of MSSD: Individuals. S. For instance. Capability Analysis (Between/Within). I-MR-R/S. Capability Analysis (Normal). Capability Analysis Multiple Variables (Normal). CUSUM. and Capability Sixpack (Between/Within) Tests 1-4 only: R. Commands Rbar: Xbar. Capability Analysis (Between/Within). Moving Average. Unbiased estimation is only available with estimation methods that use the standard deviation (Sbar. S. I-MR-R/S. only the control limits and center line.. R. each control limit will not be a single straight line. Capability Analysis Multiple Variables (Normal). and X-bar and S charts. X-bar and R. Capability Analysis (Between/Within). CUSUM. Capability Analysis (Normal). Z-MR. Moving Average. U. X-bar. Data in subgroups The default for all commands. EWMA. Moving Average. I-MR-R/S. P. Capability Sixpack (Between/Within). X-bar and R. Moving Average.

there is a gap in the Moving Range chart corresponding to each of the moving ranges that includes the missing observation. alternating up and down Test 5 2 out of 3 points > 2σ from center line (same side) Test 6 4 out of 5 points > 1σ from center line (same side) Test 7 15 points in a row within 1σ of center line (either side) Test 8 8 points in a row > 1σ from center line (either side) Missing Data with Individual Observations If an observation is missing. there is a gap in the Individuals chart where that observation would have been plotted. When calculating moving ranges. it is not calculated. If any of the observations for a particular moving range are missing. See [1] and [29] for guidance on using these tests. one that should be investigated. The occurrence of a pattern suggests a special cause for the variation. all increasing or all decreasing Test 4 Fourteen points in a row. If a point fails more than one test. the number of the first test in your list is the number printed on the chart. where K is the length of the moving ranges. . Hence. detects a specific pattern in the data plotted on the chart. Eight Tests for Special Causes Test 1 1 point more than 3σ from center line Test 2 9 points in a row on same side of center line Test 3 Six points in a row. it is marked with the test number on the chart. shown below. 22 2003 Minitab Inc. When a point fails a test. each value is the range of K consecutive observations.Quality Control Each of the tests for special causes.

However. Moving average charts plot a moving average of K subgroup averages. Unequal-size subgroups All control charts will handle unequal subgroup sizes. they are affected by unequal-size subgroups. it is omitted from the calculations of the summary statistics for the subgroup it is in. This may cause the control chart limits and the center line to have different values for that subgroup. once it finds a missing subgroup. This may also cause the Moving Average chart to produce control chart limits that are not straight lines. Hence. Like EWMA Chart. Because the control limits are functions of the subgroup size. the chart will be blank beginning with the missing subgroup. 23 . when the CUSUM chart encounters a missing subgroup. it cannot calculate any more values.Control Charts Missing Data with Data in Subgroups If a single observation is missing. you may want to force the control limits to be constant. it is omitted from the calculations of the summary statistics for the subgroup it was in. If an entire subgroup is missing. EWMA and Zone charts work with equal or unequal-size subgroups. Choose an example below: Variables Charts for Subgroups X-bar and R X-bar and S I-MR-R/S (Between/Within) X-bar R R chart with time stamp S Zone Variables Charts for Individuals I-MR Z-MR Individuals Individuals. Missing Data for Control Charts that use Weighted Statistics If a single observation is missing and you have data in subgroups. All formulas are adjusted accordingly. there is a gap in the chart where the summary statistic for that subgroup would have been plotted. When the span of the moving average is three (default) and three or more consecutive samples are missing. This may cause the EWMA chart to produce control chart limits that are not straight lines. If the sample sizes do not vary by very much. See Force control limits and center line to be constant for details. All formulas are adjusted accordingly. Moving Average charts treat missing data differently. Suppose an entire subgroup is missing: • • • EWMA plots an exponentially weighted moving average of all past subgroup means. but CUSUM charts require all subgroups to be the same size. CUSUM plots a cumulative sum of deviations. The chart will be blank starting with the missing subgroup. historical Moving Range Attributes Charts P NP C 2003 Minitab Inc. there will be a gap in the chart corresponding to all of the moving averages that would have used that subgroup mean. Zone leaves a gap in the chart corresponding to the missing subgroup. You must have at least one sample within a span to calculate the moving average. Control Chart Examples The following examples illustrate how to generate control charts with Minitab.

W. Analysis of Variance. "A Cusum for a Scale Parameter." Technometrics. N. 199−205. [17] J. [16] J. Applied Linear Statistical Models: Regression. Kincaid. April.E." Journal of Quality Technology. "Cumulative Sum Charts.R. [24] E. (1995). Burr (1976). L. [19] R. and S. 29. "Multivariate Control Charts for Individual Observations. Nachtschiem. Ishikawa (1967). Lowry.L. 20. 1−9.W. Inc. W. Statistical Quality Control Methods. McGraw-Hill.E. and J.Quality Control U Time-Weighted Charts Moving average EWMA CUSUM Multivariate Charts T2 and Generalized Variance T2 Generalized Variance Multivariate EWMA References − Control Charts [1] [2] [3] [4] [5] [6] [7] [8] [9] Automotive Industry Action Group (1991). C. Saccucci (1990). 24. [15] J. 4th edition. McGraw-Hill/Irwin. MI.M. 19.M. Inc. and J. Tracy.J. Woodall (1990). C. Lucas (1976). Crosier (1982). I. [23] C. G.W. [13] M. D. C.E. (1992). R." Journal of Quality Technology.A. Wadsworth Publishing." Communications in Statistics − Theory and Methods. April. Guide to Quality Control. Statistical Quality Control. E. Hawkins (1981). 27. B. Southfield. (2000). ASQC Quality Press. (1988). Mason. 3. . Young. Homer. 49-57. A." Journal of Quality Technology. Prins and D. Marcel Dekker. 1−12. W.S. 24. 8. pp. Farnum (1994)." Quality Engineering. Griffith (1989). AIAG. Marcel Dekker.H. Numerical Mathematics and Computing.K. "Exponentially Weighted Moving Average Control Schemes: Properties and Enhancements. 7th Edition. 3rd edition.228−231." Technometrics. pp. C. Woodall. Montgomery (2001)." Technometrics..B. 88-95. [22] L. Mader (1997-98).B. "Performance of the Zone Control Chart. Cheney and D." Journal of Quality Technology. "Performance of the Control Chart Trend Rule Under Linear Shift. "Decomposition of T2 for Multivariate Control Chart Interpretation. [14] C. Woodall (1988). 21. "The Shewhart Control Chart − Tests for Special Causes. 4th edition. (1999).H. [10] D. Neter (1996). C.M. Brooks/Cole Publishing Company. Case (1989). "Multivariate Control Charts for Grouped and Individual Observations. Kane (1989). 32. [20] D. "Exact Results for Shewhart Control Charts with Supplementary Runs Rule. Tracy. Champ. "Fast Initial Response for CUSUM Quality-Control Schemes: Give Your CUSUM a Head Start." Technometrics. Grant and R. Kutner.H. Rigdon (1992). "Development and Evaluation of Control Charts Using Exponentially Weighted Moving Averages. 24 2003 Minitab Inc. John Wiley & Sons. Ng and K. [18] R.H. L. 393-399. Lucas and M. H. Leavenworth (1996). H. and Experimental Designs. N. Page (1961). Classical and Modern Regression with Applications. W." Journal of Quality Technology. 899-108. Davis. [25] J. 4th Edition. Introduction to Statistical Quality Control." Journal of Quality Technology.S. R.W. Davis and W. Milwaukee. Mason. "The Design and Use of V-Mask Control Schemes. 10. [11] K. pp. Statistical Process Control Methods for Long and Short Runs. 46-53. 242−250. Fundamental Statistical Process Control. Young. 16. [12] V. 13. Wasserman.M. " Journal of Quality Technology. 260-262. N. 237−239. "A Multivariate Exponentially Weighted Moving Average Control Chart. D. Brook/Cole Publishing Company. Myers.S.S. Defect Prevention: Use of Simple Statistical Tools. 1−12. Nelson (1984). Lucas and R. Asian Productivity Organization." Technometrics. Modern Statistical Quality Control and Improvement. Inc. and J. Champ and W. [21] R.C. 1581-1587. Woodall.

and A. J. John Wiley & Sons. Woodall (2000). Wheeler and D. when you enter the control chart command. To do a Box-Cox transformation 1 2 Choose Stat > Control Charts > Box-Cox Transformation. Structure individual observations down a single column. Advanced Topics in Statistical Process Control: The Power of Shewhart Charts." Journal of the American Statistical Association. When you include or exclude rows using control chart options > Estimate. Subgroup sizes: Enter a number or a column of subscripts. Statistical Methods for Quality Improvement. Wheeler [31] and Wheeler and Chambers [30] suggest that it is not necessary to transform data that are used in control charts. because control charts work well in situations where data are not normally distributed. SPC Press.H. Second Edition.H.R. then enter a series of columns. [31] D. Modern Methods for Quality Control and Improvement. Ryan (2000). The Box-Cox transformation can be useful for correcting both nonnormality in process data and subgroup process variation that is related to the subgroup mean. enter a number or a column of subscripts. T. J. 2nd edition.P. Observations for a subgroup are in one row of columns: Choose if subgroups are arranged in rows across several columns. then enter one or more columns. Note You can only use this procedure with positive data. then enter the columns. Indianapolis. You can use these procedures in tandem. Under most conditions. choose Observations for a subgroup are in one row of columns. [27] Z. [29] Western Electric (1956). described in this section. Minitab only uses the non-omitted data to find lambda. 341-378. Western Electric Corporation. SPC Press. Stoumbos. K. Inc. Inc. "Controversies and Contradictions in Statistical Process Control. Minitab provides two Box-Cox transformations: a standalone command. and W. When you ask Minitab to estimate lambda. 95. Stephens. the command can be used several ways: 3 2003 Minitab Inc. M. Do one of the following: • If subgroups are in one or more columns. Chambers.S. Godfrey (2001). 992-998. 25 . Dialog box items All observations for a chart are in one column: Choose if data are in one or more columns. use the standalone command as an exploratory tool to help you determine the best lambda value for the transformation. 2nd edition.Control Charts [26] T. choose All observations for a chart are in one column. Then. Wheeler (1995). • If subgroups are in rows. use the transformation option to transform the data at the same time you draw the chart. John Wiley & Sons. 32. Statistical Quality Control Handbook. [28] H. Understanding Statistical Process Control. then enter the columns. except the Attributes charts. In Subgroup sizes. Ryan. Structure subgroup data in a single column or in rows across several columns − see Data for examples. it is not necessary to correct for nonnormality unless the data are highly skewed. First. Indiana. and a transformation option provided with all control charts.M. you get graphical output. [30] D. See Box-Cox Transformation − Graphical Output for an explanation of the output. At this point. Box-Cox Transformation Box-Cox Transformation Stat > Control Charts > Box-Cox Transformation Performs a Box-Cox procedure for process data used in control charts. Woodall (2000). "The State of Statistical Process Control as We Proceed into the 21st Century. Reynolds. the data must be positive.B. To use Box-Cox. They give an excellent demonstration of the performance of control charts when data are collected from a variety of nonsymmetric distributions. S. [32] W. (1992).P." Journal of Quality Technology. <Options> Data − Box-Cox Transformation Use this command with subgroup data or individual observations. Wadsworth.

1 2 3 4 5 Open the worksheet BOXCOX. you can look at the spread of the data both before and after the transformation using Graph > Histogram. 26 2003 Minitab Inc.Quality Control To. enterC2. Click Options. Use overall standard deviation: Check to use the overall standard deviation in the estimation of lambda. Choose Stat > Control Charts > Box-Cox Transformation. Estimate the best lambda value for the Click Options and in Store transformed data in. choose Constant(s) and enter the values or choose Column and enter a column containing the values. Box-Cox Transformation − Options Stat > Control Charts > Box-Cox Transformation > Options Lets you bypass the lambda estimation step by entering your own value of lambda to use in the transformation. then enter Skewed. then enter lambda values or a column of lambda values. As another example. . For example. Click OK.. since this transformation is simple and understandable. Click OK in each dialog box. transformed data in the column(s) you specify Transform the data with a lambda value you enter. Click OK in each dialog box.. Note In some cases. Choose All observations for a chart are in one column. and consist of 50 subgroups each of size 5. Example of Box-Cox transformation The data used in the example are highly right skewed. you get a graph that displays: • • • The best estimate of lambda for the transformation Two closely competing values of lambda A 95% confidence interval for the true value of lambda See Example of a Box-Cox data transformation for an illustration. enter a column(s) in transformation.. enter a column(s) in which to store the transformed data.5. You must enter a storage column in the main dialog box to see the transformed data. In Other (enter value(s) between -5 and 5). and store the which to store the transformed data. In Subgroup sizes. transform the data. Store transformed data in: Enter storage columns for the transformed data. The graph can be used to assess the appropriateness of the transformation. one of the closely competing values of lambda may end up having a slightly smaller standard deviation than the best estimate. since a lambda of 1 indicates that a transformation should not be done. If you like. you can use the 95% confidence interval for lambda to determine whether the optimal lambda value is "close" to 1. you could simply take the square root of the data. then click OK in each dialog box.MTW. and store the transformed data in a column(s) you specify Click Options and in Store transformed data in. Estimate the best lambda value for the transformation Do this.. you would gain very little by performing the transformation. if the optimal lambda is "close" to 0. Box-Cox Transformation − Graphical Output When you ask Minitab to estimate the lambda value. Under Store transformed data in. Dialog box items Optimal lambda(s): Choose to have Minitab estimate an optimal value of lambda (default). Other (enter value(s) between -5 and 5) : Choose to enter your own value of lambda to bypass the lambda estimation step. enter 5. In the case that the optimal lambda is close to 1.

such as the square root (a lambda of 0.039501) and the best value (0. and multivariate charts also plot measurement data. this corresponds to an interval of − 0. In this example. 2003 Minitab Inc. such as length or pressure. which is the value used in the transformation. Choosing a variables control chart for subgroups Minitab offers five variables control charts: • • • X-bar and R − an X-bar chart and R chart displayed in one window X-bar and S − an X-bar chart and S chart displayed in one window I-MR-R/S (Between/Within) − a three-way control chart that uses both between-subgroup and within-subgroup variation. such as the number of defects or defective units.207308). time-weighted charts. Variables control charts for individuals.5) or the natural log (a lambda of 0). for subgroup data.000000).292850) and lower CI (−0. 0 is a reasonable choice because it falls within the 95% confidence interval. Attributes control charts plot count data. see Control Charts Overview. which are marked on the graph by vertical lines. Therefore. the natural log transformation may be preferred to the transformation defined by the best estimate of lambda. any lambda value which has a standard deviation close to the dashed line is also a reasonable value to use for the transformation. For more information about control charts. and a R or S chart.Control Charts Graph window output Interpreting the results The Lambda table contains an estimate of lambda (0. In this example. The 95% confidence interval includes all lambda values which have a standard deviation less than or equal to the horizontal line. Therefore. Variables Charts for Subgroups Xbar-R Chart Variables Control Charts for Subgroups Overview data Variables control charts for subgroups plot statistics from measurement data. The Lambda table also includes the upper CI (0. 27 .292850. in any practical situation you want a lambda value that corresponds to an understandable transformation. a MR chart. Although the best estimate of lambda is a very small negative number. An I-MR-R/S chart consists of an I chart.207308 to 0.

Variables charts include: • • • • Variables charts for subgroups Variables charts for individuals Time-weighted charts Multivariate charts Structure your data for these charts using the guidelines below. By default.Quality Control • • • • X-bar − a chart of subgroup means R − a chart of subgroup ranges S − a chart of subgroup standard deviations Zone − a chart of the cumulative scores based on each point's distance from the center line The I-MR-R/S (Between/Within) chart requires that you have two or more observations in at least one subgroup. Multivariate data must be entered down columns. the R chart in the lower half. . Observations for a subgroup are in one row of columns: Choose if subgroups are arranged in rows across several columns. using the following table as a guide. X-bar and R charts are typically used to track the process level and process variation for samples of size 8 or less. as well as detect the presence of special causes. with one column for each variable. while X-bar and S charts are used for larger samples. Worksheet Structure Structure your data down a column or across rows. Minitab's X-bar and R chart bases the estimate of the process variation. see Control Charts Overview and Variables Control Charts for Subgroups Overview. Subgroup sizes: Enter a number or a column of subscripts. Seeing both charts together allows you to track both the process level and process variation at the same time. If the subgroups are not equal. Subgroups are equal size Univariate (one variable) Multivariate (more than one variable) Down columns or across rows Down columns Subgroups are unequal size Down columns with subgroup indicator column Down columns with subgroup indicator column 28 2003 Minitab Inc. each control limit is not a single straight line but varies with the subgroup size. The X-bar chart is drawn in the upper half of the screen. Subgroups do not need to be the same size. you may want to force the control limits to be constant by specifying a fixed subgroup size using Xbar-R Options > Estimate. Minitab calculates summary statistics for each subgroup. If the subgroup sizes do not vary much. or enter a historical value for σ. which are plotted on the chart and used to estimate process parameters. See [29] for a discussion of how to interpret joint patterns in the two charts. X-bar and R Chart Stat > Control Charts > Variables Charts for Subgroups > Xbar-R Displays a control chart for subgroup means (an X-bar chart) and a control chart for subgroup ranges (an R chart) in the same graph window. <Scale> <Labels> <Multiple Graphs> <Data Options> <Xbar-R Options> Data − Variables Control Chart Organize the data for all variables control charts in the same way. For more information. then enter the columns. You can also use a pooled standard deviation. on the average of the subgroup ranges. σ. then enter the columns. Dialog box items All observations for a chart are in one column: Choose if data are in one or more columns.

with subgroups of size 5. and so on. For information on the Box-Cox transformation option. then click OK. For information on the stand alone Box-Cox transformation command. choose Observations for a subgroup are in one row of columns. Use the stand alone command as an exploratory tool to help you determine the best lambda value for the transformation.Control Charts Structure subgroup data down a column or across rows. In this example. or by using the stand alone Box-Cox command. you can use the transformation option to transform the data at the same time you draw the control chart. then create a second column of subscripts which serve as subgroup indicators. 3 2003 Minitab Inc. see Box-Cox Transformation. then enter a series of columns. To display an X-bar and R chart 1 2 Choose Stat > Control Charts > Variables Charts for Subgroups > Xbar-R. Nonnormal data To properly interpret Minitab's control charts. • If subgroups are in rows. a new subgroup begins in C1. Do one of the following: • If subgroups are in one or more columns. If the data are highly skewed. and so on. Here is the same data set. When subgroups are of unequal size. C1 contains the process data and C2 contains subgroup indicators: Each time a subscript changes in C2. the second 5 observations are the second row. subgroup 2 has six observations. enter a number or a column of subscripts. see Options − Box-Cox. Note that the first five observations in the left data set (subgroup 1) are the first row of the right-side data set. you may want to use the Box-Cox transformation to induce normality. In Subgroup sizes. Missing data See Missing data in control charts for information on how to handle missing data for different types of control charts. If you like. choose All observations for a chart are in one column. use any dialog box options. Then. you must enter data that approximate a normal distribution. you must enter your data in one column. 29 . You can access the Box-Cox transformation two ways: by using the Box-Cox transformation option provided with the control chart commands. subgroup 1 has three observations. then enter one or more columns. structured both ways. In the following example.

You can also select one of two methods to estimate σ and use biased estimates of σ. Use the following samples when estimating parameters: Choose to specify subgroups of data to compute µ and σ. if you have a known process parameter or an estimate obtained from past data. EWMA Use to enter historical data for estimating µ and σ. Use unbiasing constant: Uncheck to use biased estimates of σ. Standard deviation: Enter the historical standard deviations as constants. X-bar.Quality Control Options − Parameters . Minitab only uses the non-omitted data to find lambda. you can prevent these subgroups from influencing the estimation. enter one or more values. Each stage for each variable uses a different value. enter one or more values. Enter entire subgroups. 3 • • • You can enter historical data in several ways: Options − Estimate . Minitab estimates it from the data. A single value.. 30 2003 Minitab Inc. one for each variable. one for each stage within each variable. then decide to omit subgroup 2 from the parameter estimation (using control chart Options > Estimate). Dialog box items Omit the following samples when estimating parameters: Choose to specify subgroups of data to omit from the computations for µ and σ. When you include or exclude rows using control chart Options > Estimate and choose to perform a Box-Cox transformation. Because Minitab considers an omitted sample (subgroup 1) as part of the entire data set. Minitab estimates the process parameters from all the data.. you exclude subgroup 1 from your analysis (using control chart > Data Options > Subset). Do any of the following: • In Mean. X-bar and S. If you do not specify a value for µ or σ. the omitted or used samples apply to the entire data set. Dialog box items Mean: Enter the historical means as constants. then enter individual subgroup numbers or a range of subgroups. But you may want to use or omit certain data if they show abnormal behavior. you must enter subgroup 2 as the subgroup to omit. you can enter these values. Individuals. Method for estimating standard deviation: See [2] for a discussion of the relative merits of the methods listed below. Subgroup size > 1 Rbar: Choose to estimate σ based on the average of the subgroup ranges. To use historical mean and standard deviation values 1 2 In the control chart dialog box. not just the subsetted data. I-MR. For example. A list of values. All stages for each variable use the same value. R Omits or includes certain subgroups to estimate µ and σ. You can set preferences for the estimation of σ using Tools > Options > Control Charts and Quality Tools > Estimating Standard Deviation. A list of values. click control chart Options. • In Standard deviation. . For example. To omit one observation from a subgroup.. Moving Average. if some subgroups have erratic data due to assignable causes that you have corrected.. For example. > control chart Options > Parameters Charts X-bar and R. Caution When you include or exclude rows using control chart > Data Options > Subset and also omit or use samples to estimate parameters. Pooled standard deviation: Choose to estimate σ using a pooled standard deviation. Click OK. To omit or use subgroups for the estimates of parameters By default. Zone. not individual observations. which applies to all variables and stages. change that value to missing (∗). then enter individual subgroup numbers or a range of subgroups. You can omit or include certain subgroups when estimating parameters using Estimate. > control chart Options > Estimate Charts X-bar and R.

Only the control limits are affected. If you specify more than one value. Tip You can also modify the control limits using Minitab's graph editing features. the plotted data are not changed. 31 . Assuming all subgroups have size: Choose to estimate the control limits using a specified subgroup size. Moving Range. If the calculated upper control limit is greater than the upper bound. and I-MR charts. If the sizes do not vary much. Click the Estimate tab. To customize the control limits for X-bar and R charts 1 2 In the control chart data source dialog box. Then click OK. If the calculated upper control limit is greater than the upper bound. Minitab excludes any moving ranges with omitted data from the calculations. Place bounds on control limits of R Chart Lower standard deviation limit bound: Check to set a lower bound for the control limits for the R chart. The values entered are the number of standard deviations above and below the center line. You can also force the control limits to be constant when sample sizes are unequal. click Xbar-R Options. If the calculated lower control limit is less than the lower bound. Minitab assumes the values you enter are subgroup numbers. If the calculated lower control limit is less than the lower bound. Click the Estimate tab. a horizontal line labeled UB will be drawn at the upper bound instead. Do any of the following: 2003 Minitab Inc. When subgroup sizes are unequal. For an example. When you use Omit the following samples when estimating parameters with Length of moving range. a horizontal line labeled LB will be drawn at the lower bound instead. a pair of control limits is drawn for each number in the list. Click the S Limits tab. a horizontal line labeled LB is drawn at the lower bound instead. the values are interpreted as observation numbers.Control Charts 1 2 In the control chart dialog box. click control chart Options. Under Estimating standard deviation. Note 3 Click OK. Place bounds on control limits of Xbar Chart Lower standard deviation limit bound: Check to set a lower bound for the control limits for the X-bar chart. calculate control limits Using actual sizes of the subgroups: Choose to use the subgroups defined in the main dialog box to estimate σ. Upper standard deviation limit bound: Check to set an upper bound for the control limits for the R chart. Similarly. a horizontal line labeled LB will be drawn at the lower bound instead. but varies with the subgroup size. if the calculated lower control limit is less than the lower bound. Do one of the following: • Choose Omit the following samples when estimating parameters and enter the subgroups or observation numbers you want to omit from the calculations. Dialog box items Display control limits at These multiples of the standard deviation: Enter one or more values. a horizontal line labeled UB is drawn at the upper bound instead. Note Xbar and R − Options − S Limits Stat > Control Charts > Variables Charts for Subgroups > Xbar-R > Xbar-R Options > S Limits Displays control limits above and below the mean at the multiples of any standard deviation. then type the subgroup size. With these charts. You can set bounds on the upper and/or lower control limits. When subgroup sizes are not equal. each control limit is not a single straight line. Upper standard deviation limit bound: Check to set an upper bound for the control limits for the X-bar chart. When the calculated upper control limit is greater than the upper bound. a horizontal line labeled UB will be drawn at the upper bound instead. • Choose Use the following samples when estimating parameters and enter the subgroups or observation numbers you want to use for the calculations. To choose how the standard deviation is estimated 1 2 In the control chart dialog box. you may want to force the control limits to be constant by entering a sample size in this text box. see Example of an X-bar chart with tests and customized control limits. click control chart Options. except with the Individuals. choose a method.

If the data change and the control chart is live-linked. For example. See Updating Graphs for more information. all increasing or all decreasing 14 points in a row. Click OK. Minitab marks it with the test number on the plot.. The occurrence of a pattern suggests a special cause for the variation. The test results are displayed in the Session window. the test results are not updated in the Session window. If a point fails more than one test. click control chart Options. I-MR Selects a subset of the tests shown in Eight Tests for Special Causes. Display test results in Session window: Check to display the test results in the Session window. check Lower standard deviation limit bound (and/or Upper standard deviation limit bound) and enter a value. X-bar and S. You can set preferences for the sensitivity of these tests using Tools > Options > Control Charts and Quality Tools > Define Tests. enter a value of 6 to calculate the control limits and center line as if all subgroup sizes were 6. > control chart Options > Tests Charts X-bar and R. To set bounds on the control limits for the moving range chart: under Place bounds on control limits of R chart. . Under When subgroup sizes are unequal. I-MR-R/S (Between/Within). then check any combination of the tests listed below. above. Each test detects a specific pattern in the data plotted on the chart. To set bounds on the control limits for the individuals chart: under Place bounds on control limits of Xbar chart. For example. check Lower standard deviation limit bound (and/or Upper standard deviation limit bound) and enter a value. 32 2003 Minitab Inc. X-bar. You should force the control limits and center line to be constant only when the difference in size between the largest and smallest subgroups is no more than 25%.. calculate control limits. You can choose to change the color of a control chart when the last row of data causes a new test failure for any point. one that should be investigated. then enter a value. alternating up and down 2 out of 3 points > 2 standard deviations from center line (same side) 4 out of 5 points > 1 standard deviation from center line (same side) 15 points in a row within 1 standard deviation of center line (either side) 8 points in a row > 1 standard deviation from center line (either side) Perform no tests: Choose to perform no tests for special causes. When a point fails a test. See [1] and [29] for guidance on using these tests. suppose the largest subgroup is size 125 and the smallest is 100. Minitab will perform tests when subgroup sizes are unequal. enter one or more values or a column of values in These multiples of the standard deviation. choose Assuming all subgroups have size.Quality Control • • • To specify where control limits are drawn: under Display control limits at. See Tools > Options > Control Charts and Quality Tools > Other. Individuals. Minitab marks it by the lowest numbered test. To force control limits and center line to be constant 1 2 In the control chart dialog box. Each value represents the number of standard deviations below and above the mean. 3 Click OK in each dialog box. 1 point more than 3 standard deviations from center line 9 points in a row on same side of center line 6 points in a row. You can use this method because the size difference is 25% (125 / 100 = 25%). Click the S Limits tab. Each value is the number of standard deviations the lines should be drawn at. 3 Note Options − Tests . Each value represents the number of standard deviations below and above the mean. Perform the following tests for special causes: Choose to specify a subset of the eight tests. Dialog box items Tests For Special Causes (default definitions): Perform all tests for special causes: Choose to perform all eight tests shown below. and below the mean.

You can enter the same value more than once. You can also choose any value between − 5 and 5 for λ. Enter the values. Click OK. S. To display a historical chart To define stages in your process.. and X-bar and S. The column must be the same length as the data column (or columns. Enclose date/time or text entries in double quotes. When executing the command. you must have at least one subgroup with two or more observations: R. you should not choose a λ outside the range of −2 and 2.) To use this option. (λ is lambda. The indicators can be numbers. in which case the natural log is taken. 3 Options − Stages . Minitab only uses the non-omitted data to find lambda.. dates. Minitab treats each repeated value as a separate occurrence.5). Enclose date/time or text entries in double quotes. enter the column which contains the stage indicators. then click OK. do one of the following: • To start a new stage each time the value in the column changes. In most cases. when subgroups are across rows). the data must be positive. • To start a new stage at the first occurrence of a certain value. or text. 5 Options − Box-Cox . unless λ = 0. If you like. choose Perform the following tests for special causes and check the tests to perform. You can enter the same value more than once. You may want to first run the command Stat > Control Charts > Box-Cox Transformation to help you find the optimal transformation value. In Define stages (historical groups) with this variable. Note With the following charts. Minitab treats each repeated value as a separate occurrence. choose Perform all tests for special causes. > control chart Options > Stages Charts All charts except Z-MR You can display stages in your process by drawing a "historical chart" − a control chart in which the control limits and center line are estimated independently for different groups in your data. Do one of the following: • To do all tests.. then enter the values. choose Perform no tests. choose With each new value. • To select certain tests. 33 . With the first occurrence of these values: Choose to start a new stage at the first occurrence of a certain value. Historical charts are particularly useful for comparing data before and after a process improvement. Dialog box items Define stages (historical groups) with this variable: Enter the column that contains the stage indicators. you must set up a column of grouping indicators." The transformation takes the original data to the power λ. use any dialog box options. The Options subdialog box lists the common transformations natural log (λ = 0) and square root (λ = 0. choose With the first occurrence of these values. X-bar and R. > control chart Options > Box-Cox Charts All except the Attributes charts You can use the Box-Cox power transformation when your data are very skewed or when the within-subgroup variation is unstable to make the data "more normal. Under When to start a new value. you can tell Minitab to start a new stage in one of two ways: • • 1 2 3 4 Each time the value in the column changes At the first occurrence of one or more values In the control chart dialog box. When to start a new stage With each new value: Choose to start a new stage each time the value in the column changes.Control Charts To perform the tests for special causes 1 2 Choose Stat > Control Charts > control chart type > control chart > control chart Options > Tests. • To perform no tests. Click the Stages tab. click control chart Options. 2003 Minitab Inc.. When you include or exclude rows using control chart Options > Estimate.

Minitab does not accept any values you enter in the Parameters tab for historical means or historical standard deviations.Quality Control When you use this transformation.5 (square root): Choose to use the square root of the data. Split chart into series of segments for display purposes Do not split: Choose not to split the chart. then enter the lambda values. Lambda = 0 (natural log): Choose to use the natural log of the data. . Test results Display test results in Session window: Check to display the test results in the Session window.. Click the Box-Cox tab. To display subgroups 1 2 In the control chart dialog box. • Choose Lambda=0. > control chart Options > Display Charts All charts except Z-MR Use to display the control chart by stage or by number of plotted points and to display the test results in the Session window. Optimal lambda for each stage (in a chart with stages): Choose to have Minitab search for optimal values for each stage. under Subgroups to display. Subgroups to display Display all subgroups: Choose to display all subgroups.5 (square root) to use the square root of the data. Lambda=0. Click the Display tab. For help choosing a lambda value. Do one of the following: • To display all subgroups in the control chart.. and enter the number of subgroups. Use overall standard deviation: Check to use the overall standard deviation in the estimation of the optimal lambda. • Choose Optimal lambda to have Minitab search for an optimal value for all stages. Display last __ subgroups: Choose to display a specific number of subgroups. Dialog box items Box-Cox power transformation (W = Y**Lambda): Choose to use when your data are very skewed or when the withinsubgroup variation is unstable. • Choose Optimal lambda for each stage (in a chart with stages) to have Minitab search for optimal values for each stage. then do one of the following: • Choose Lambda = 0 (natural log) to use the natural log of the data. then enter one or more values. Options − Display . Each segment contains __ subgroups: Choose to split the control chart into subgroups of a specified number of points. 3 Click OK. Check Box-Cox power transformation (W = Y**Lambda). Each segment corresponds to a stage (if chart has stages): Choose to split the control chart by stage or stage when you have specified stages in control chart Options > Stage. click control chart Options. 34 2003 Minitab Inc. To do the Box-Cox power transformation 1 2 In the control chart dialog box. Caution If you use Stat > Control Charts > Box-Cox Transformation to find the optimal λ value and choose to store the transformed data with that command. doing so will double-transform the data. click control chart Options. Other (enter value(s) between -5 and 5): Choose to transform the data using another lambda value or values. see the independent Box-Cox transformation command Stat > Control Charts > Box-Cox Transformation. • Choose Other (enter value(s) between -5 and 5) to transform the data using another lambda value or values. and enter the number of subgroups for each segment. choose Display all subgroups. Optimal lambda: Choose to have Minitab search for an optimal value. do not select the Box-Cox option when you make a control char.

choose Each segment corresponds to a stage (if chart has stages). • To split the chart by stage. Your supervisor wants to run X-bar and R charts to monitor this characteristic. Control limit values: Check to store the control limit values. Center line value: Check to store the center line value. one row for each plotted point. one row for each plotted point. EWMA. Minitab stores one column for the lower control limit and one column for the upper control limit. 2003 Minitab Inc. then click OK. 35 . one row for each plotted point.MTW. Example of X-bar and R chart You work at an automobile engine assembly plant. which causes poor-fitting assemblies. a camshaft. one row for each stage. Click the Display tab. click control chart Options. One of the parts. Click OK. Stage: Check to store the stage. Test results: Check to store the results of any test performed. Choose All observations for a chart are in one column.Control Charts • 3 To display a specific number of the last subgroups. Standard deviations: Check to store the estimates of the standard deviations. Choose Stat > Control Charts > Variables Charts for Subgroups > Xbar-R. Subgroup size: Check to store the subgroup sizes. First you will look at camshafts produced by Supplier 2. 3 Options − Storage Stat > Control Charts > control chart type > control chart > control chart Options > Storage Charts X-bar and R. and 100 observations from each of your suppliers. one row for each historical stage. one row for each plotted point. Stores these values for each point Point plotted: Check to store the plotted points. one row for each plotted point. CUSUM Use to store your choice of statistics in the worksheet. resulting in high scrap and rework rates. Click OK. 1 2 3 4 Open the worksheet CAMSHAFT. under Subgroups to display. Check the statistics to store in the worksheet. Click OK. X-bar and S. must be 600 mm +2 mm long to meet engineering specifications. enter 5. There has been a chronic problem with camshaft length being out of specification. so for a month. one row for each plotted point. you collect a total of 100 observations (20 samples of 5 camshafts each) from all the camshafts used at the plant. choose Display last __ subgroups and enter the number of subgroups to display. In Subgroup sizes. choose Each segment contains __ subgroups. X-bar. To split a control chart 1 2 In the control chart dialog box. and enter the number of subgroups for each segment. Dialog box items Store these estimates for each chart Means: Check to store the estimates of the means. To store statistics in the worksheet 1 2 Choose Stat > Control Charts > control chart type > control chart > control chart Options > Storage. then enter Supp2. Moving Average. Do one of the following: • To split the chart into a specific number of subgroups.

Choosing a variables control chart for subgroups Minitab offers five variables control charts: • • • • • X-bar and R − an X-bar chart and R chart displayed in one window X-bar and S − an X-bar chart and S chart displayed in one window I-MR-R/S (Between/Within) − a three-way control chart that uses both between-subgroup and within-subgroup variation. For more information about control charts. 3. such as the number of defects or defective units. but two of the points fall outside the control limits. Graph window output Interpreting the results The center line on the X-bar chart is at 600. Xbar-S Chart Variables Control Charts for Subgroups Overview data Variables control charts for subgroups plot statistics from measurement data. implying that your process is falling within the specification limits.00 standard deviations from center line. X-bar − a chart of subgroup means R − a chart of subgroup ranges 36 2003 Minitab Inc. Variables control charts for individuals. such as length or pressure. Attributes control charts plot count data. time-weighted charts. There may be excess variability in your process. the results above may no * longer be correct. . Test Failed at points: 2. and multivariate charts also plot measurement data. see Control Charts Overview. One point more than 3. for subgroup data. The center line on the R chart. is also quite large considering the maximum allowable variation is +2 mm.890. An I-MR-R/S chart consists of an I chart.23. implying an unstable process. 14 * WARNING * If graph is updated with new data. a MR chart. and a R or S chart.Quality Control Session window output Xbar-R Chart of Supp2 Test Results for Xbar Chart of Supp2 TEST 1.

See [29] for a discussion of how to interpret joint patterns in the two charts. Subgroups do not need to be the same size. which are plotted on the chart and used to estimate process parameters. structured both ways. or enter a historical value for σ. By default. on the average of the subgroup standard deviations.Control Charts • • S − a chart of subgroup standard deviations Zone − a chart of the cumulative scores based on each point's distance from the center line The I-MR-R/S (Between/Within) chart requires that you have two or more observations in at least one subgroup. For more information. then enter the columns. 37 . with subgroups of size 5. Worksheet Structure Structure your data down a column or across rows. <Scale> <Labels> <Multiple Graphs> <Data Options> <Xbar-S Options> Data − Variables Control Chart Organize the data for all variables control charts in the same way. then enter the columns. Note that the first five observations in the left data set (subgroup 1) are the first row of the right-side data set. Subgroups are equal size Univariate (one variable) Multivariate (more than one variable) Down columns or across rows Down columns Subgroups are unequal size Down columns with subgroup indicator column Down columns with subgroup indicator column Structure subgroup data down a column or across rows. using the following table as a guide. Seeing both charts together allows you to track both the process level and process variation at the same time. Dialog box items All observations for a chart are in one column: Choose if data are in one or more columns. 2003 Minitab Inc. Subgroup sizes: Enter a number or a column of subscripts. and so on. If the subgroup sizes do not vary much. Minitab's X-bar and S Chart command bases the estimate of the process variation. while X-bar and R charts are used for smaller samples. you may want to force the control limits to be constant by specifying a fixed subgroup size using Xbar-S Options > Estimate. the S chart in the lower half. Multivariate data must be entered down columns. You can also use a pooled standard deviation. as well as detect the presence of special causes. Observations for a subgroup are in one row of columns: Choose if subgroups are arranged in rows across several columns. each control limit is not a single straight line but varies with the subgroup size. with one column for each variable. The X-bar chart is drawn in the upper half of the screen. If the subgroups are not equal. σ. the second 5 observations are the second row. Variables charts include: • • • • Variables charts for subgroups Variables charts for individuals Time-weighted charts Multivariate charts Structure your data for these charts using the guidelines below. Minitab calculates summary statistics for each subgroup. X-bar and S Chart Stat > Control Charts > Variables Charts for Subgroups > Xbar-S Displays a control chart for subgroup means (an X-bar chart) and a control chart for subgroup standard deviations (an S chart) in the same graph window. X-bar and S charts are typically used to track process variation for samples of nine or more. see Control Charts Overview and Variables Control Charts for Subgroups Overview. Here is the same data set.

then enter a series of columns. If the data are highly skewed. or by using the stand alone Box-Cox command. If you like. you must enter your data in one column. Do one of the following: • If subgroups are in one or more columns. For information on the stand alone Box-Cox transformation command. Missing data See Missing data in control charts for information on how to handle missing data for different types of control charts. see Options − Box-Cox. you can use the transformation option to transform the data at the same time you draw the control chart. In this example. C1 contains the process data and C2 contains subgroup indicators: Each time a subscript changes in C2. a new subgroup begins in C1. In Subgroup sizes. Use the stand alone command as an exploratory tool to help you determine the best lambda value for the transformation. enter a number or a column of subscripts. you may want to use the Box-Cox transformation to induce normality. In the following example. then create a second column of subscripts which serve as subgroup indicators. see Box-Cox Transformation. Nonnormal data To properly interpret Minitab's control charts. choose Observations for a subgroup are in one row of columns. use any dialog box options. • If subgroups are in rows. subgroup 2 has six observations. You can access the Box-Cox transformation two ways: by using the Box-Cox transformation option provided with the control chart commands. subgroup 1 has three observations. you must enter data that approximate a normal distribution. . 3 38 2003 Minitab Inc. To display an X-bar and S chart 1 2 Choose Stat > Control Charts > Variables Charts for Subgroups > Xbar-S. then enter one or more columns. choose All observations for a chart are in one column. then click OK.Quality Control When subgroups are of unequal size. and so on. Then. For information on the Box-Cox transformation option.

If you do not specify a value for µ or σ. Click OK. 39 . Standard deviation: Enter the historical standard deviations as constants. X-bar.. Method for estimating standard deviation: See [2] for a discussion of the relative merits of the methods listed below. not just the subsetted data. Minitab estimates it from the data. A list of values. Dialog box items Omit the following samples when estimating parameters: Choose to specify subgroups of the data to omit from the computations for µ and σ. Zone. not individual observations. enter one or more values. To omit one observation from a subgroup. 2003 Minitab Inc. All stages for each variable use the same value. > control chart Options > Estimate Charts X-bar and S. A single value. Subgroup size > 1 Sbar: Choose to estimate σ based on the average of the subgroup standard deviations. X-bar and S. Minitab estimates the process parameters from all the data. • In Standard deviation. then decide to omit subgroup 2 from the parameter estimation (using control chart Options > Estimate). Enter entire subgroups. S Omits or includes certain subgroups to estimate µ and σ. For example.Control Charts Options − Parameters . But you may want to use or omit certain data if they show abnormal behavior. if you have a known process parameter or an estimate obtained from past data. Because Minitab considers an omitted sample (subgroup 1) as part of the entire data set. Moving Average. Use unbiasing constant: Uncheck to use biased estimates of σ. You can set preferences for the estimation of σ using Tools > Options > Control Charts and Quality Tools > Estimating Standard Deviation. which applies to all variables and stages. change that value to missing (∗) in the worksheet. EWMA Use to enter historical data for estimating µ and σ. 3 • • • You can enter historical data in several ways: Options − Estimate . the omitted or used samples apply to the entire data set. > control chart Options > Parameters Charts X-bar and R.. then enter individual subgroup numbers or a range of subgroups. I-MR. You can also select one of two methods to estimate σ and use biased estimates of σ. Dialog box items Mean: Enter the historical means as constants. enter one or more values. Caution When you include or exclude rows using control chart > Data Options > Subset and also omit or use samples to estimate parameters. To use historical mean and standard deviation values 1 2 In the control chart dialog box. you can enter these values. you can prevent these subgroups from influencing the estimation of process parameters. A list of values. Individuals. one for each variable. then enter individual subgroup numbers or ranges of subgroups. you exclude subgroup 1 from your analysis (using control chart > Data Options > Subset). Each stage for each variable uses a different value. if some subgroups have erratic data due to assignable causes that you have corrected. click control chart Options. Pooled standard deviation: Choose to estimate σ using a pooled standard deviation.. For example. Minitab only uses the non-omitted data to find lambda. Use the following samples when estimating parameters: Choose to specify subgroups of data to compute µ and σ. You can omit or include certain subgroups when estimating parameters using Estimate. To omit or use subgroups for the estimates of parameters By default. For example. one for each stage within each variable. Do any of the following: • In Mean. When you include or exclude rows using control chart Options > Estimate and choose to perform a Box-Cox transformation. you must enter subgroup 2 as the subgroup to omit..

Place bounds on control limits of Xbar Chart Lower standard deviation limit bound: Check to set a lower bound for the control limits for the X-bar chart. Similarly. but varies with the subgroup size. With these charts. click control chart Options. The values entered are the number of standard deviations above and below the center line. When subgroup sizes are not equal. choose a method. To choose how the standard deviation is estimated 1 2 In the control chart dialog box. Do one of the following: • Choose Omit the following samples when estimating parameters and enter the subgroups or observation numbers you want to omit from the calculations. Upper standard deviation limit bound: Check to set an upper bound for the control limits for the X-bar chart. You can also force the control limits to be constant when sample sizes are unequal. Assuming all subgroups have size: Choose to estimate the control limits using a specified subgroup size. Minitab excludes any moving ranges with omitted data from the calculations. a horizontal line labeled LB is drawn at the lower bound instead. If the calculated upper control limit is greater than the upper bound. click Xbar-S Options. a horizontal line labeled UB will be drawn at the upper bound instead. If the calculated upper control limit is greater than the upper bound. a horizontal line labeled UB is drawn at the upper bound instead. If the sizes do not vary much. calculate control limits Using actual sizes of the subgroups: Choose to use the subgroups defined in the main dialog box to estimate σ. To customize the control limits for X-bar and S charts 1 2 In the control chart data source dialog box. When subgroup sizes are unequal. the values are interpreted as observation numbers. Tip You can also modify the control limits using Minitab's graph editing features. Note Xbar and S − Options − S Limits Stat > Control Charts > Variables Charts for Subgroups > Xbar-S > Xbar-S Options > S Limits Allows you to draw control limits above and below the mean at the multiples of any standard deviation. Do any of the following: 40 2003 Minitab Inc. except with the Individuals. Moving Range. see Example of an Xbar chart with tests and customized control limits. If you specify more than one value. Only the control limits are affected. Minitab assumes the values you enter are subgroup numbers. Upper standard deviation limit bound: Check to set an upper bound for the control limits for the S chart. each control limit is not a single straight line. Click the Estimate tab. Under Estimating standard deviation. Click the S Limits tab. If the calculated lower control limit is less than the lower bound. When the calculated upper control limit is greater than the upper bound. . if the calculated lower control limit is less than the lower bound. then enter the subgroup size. Then click OK. When you use Omit the following samples when estimating parameters with Length of moving range. a horizontal line labeled UB will be drawn at the upper bound instead. a pair of control limits is drawn for each number in the list. click control chart Options. Dialog box items Display control limits at These multiples of the standard deviation: Enter one or more values. Click the Estimate tab. a horizontal line labeled LB will be drawn at the lower bound instead. the plotted data are not changed. • Choose Use the following samples when estimating parameters and enter the subgroups or observation numbers you want to use for the calculations. you may want to force the control limits to be constant by entering a sample size in this text box. a horizontal line labeled LB will be drawn at the lower bound instead. You can set bounds on the upper and/or lower control limits. and I-MR charts.Quality Control 1 2 In the control chart dialog box. If the calculated lower control limit is less than the lower bound. Note 3 Click OK. For an example. Place bounds on control limits of S Chart Lower standard deviation limit bound: Check to set a lower bound for the control limits for the S chart.

. choose Assuming all subgroups have size. See Tools > Options > Control Charts and Quality Tools > Other. Each value represents the number of standard deviations below and above the mean. The test results are displayed in the Session window. Minitab marks it with the test number on the plot. 2003 Minitab Inc. If a point fails more than one test. The occurrence of a pattern suggests a special cause for the variation. 1 point more than 3 standard deviations from center line 9 points in a row on same side of center line 6 points in a row. See Updating Graphs for more information. alternating up and down 2 out of 3 points > 2 standard deviations from center line (same side) 4 out of 5 points > 1 standard deviation from center line (same side) 15 points in a row within 1 standard deviation of center line (either side) 8 points in a row > 1 standard deviation from center line (either side) Perform no tests: Choose to perform no tests for special causes. all increasing or all decreasing 14 points in a row. suppose the largest subgroup is size 125 and the smallest is 100. To set bounds on the control limits for the individuals chart: under Place bounds on control limits of Xbar chart. Individuals. check Lower standard deviation limit bound (and/or Upper standard deviation limit bound). then enter a value. Dialog box items Tests For Special Causes (default definitions): Perform all tests for special causes: Choose to perform all eight tests shown below. enter one or more values or a column of values in These multiples of the standard deviation. the test results are not updated in the Session window. Minitab marks it by the lowest numbered test. For example. > control chart Options > Tests Charts X-bar and R. click control chart Options.Control Charts • • • To specify where control limits are drawn: under Display control limits at. Minitab will perform tests when subgroup sizes are unequal.. one that should be investigated. Each value represents the number of standard deviations below and above the mean. Click the S Limits tab. enter a value of 6 to calculate the control limits and center line as if all subgroup sizes were 6. I-MR Selects a subset of the tests shown in Eight Tests for Special Causes. 3 Note Options − Tests . X-bar and S. I-MR-R/S (Between/Within). Perform the following tests for special causes: Choose to specify a subset of the eight tests. You can choose to change the color of a control chart when the last row of data causes a new test failure for any point. To set bounds on the control limits for the moving range chart: under Place bounds on control limits of S chart. 3 Click OK in each dialog box. and enter a value. You can set preferences for the sensitivity of these tests using Tools > Options > Control Charts and Quality Tools > Define Tests. calculate control limits. Under When subgroup sizes are unequal. Display test results in Session window: Check to display the test results in the Session window. 41 . Click OK. Each test detects a specific pattern in the data plotted on the chart. then check any combination of the tests listed below. If the data change and the control chart is live-linked. When a point fails a test. X-bar. check Lower standard deviation limit bound (and/or Upper standard deviation limit bound). above. and below the mean. You should force the control limits and center line to be constant only when the difference in size between the largest and smallest subgroups is no more than 25%. For example. See [1] and [29] for guidance on using these tests. To force control limits and center line to be constant 1 2 In the control chart dialog box. You can use this method because the size difference is 25% (125 / 100 = 25%). Each value is the number of standard deviations the lines should be drawn at. and enter a value.

You can enter the same value more than once. Do one of the following: • To do all tests. unless λ = 0. Historical charts are particularly useful for comparing data before and after a process improvement. > control chart Options > Box-Cox Charts All except the Attributes charts You can use the Box-Cox power transformation when your data are very skewed or when the within-subgroup variation is unstable to make the data "more normal. The column must be the same length as the data column (or columns. Dialog box items Define stages (historical groups) with this variable: Enter the column that contains the stage indicators. Enclose date/time or text entries in double quotes. Minitab treats each repeated value as a separate occurrence. You can enter the same value more than once. In most cases.) To use this option. Click the Stages tab. dates. you can tell Minitab to start a new stage in one of two ways: • • 1 2 3 4 Each time the value in the column changes At the first occurrence of one or more values In the control chart dialog box. When you include or exclude rows using control chart Options > Estimate.. • To perform no tests. use any dialog box options. when subgroups are across rows). choose Perform the following tests for special causes and check the tests to perform. When to start a new stage With each new value: Choose to start a new stage each time the value in the column changes.. When executing the command. you must set up a column of grouping indicators. do one of the following: • To start a new stage each time the value in the column changes. With the first occurrence of these values: Choose to start a new stage at the first occurrence of a certain value. 3 Options − Stages . you should not choose a λ outside the range of −2 and 2. in which case the natural log is taken. enter the column which contains the stage indicators.5). You can also choose any value between − 5 and 5 for λ. Note With the following charts. Click OK. You may want to first run the command Stat > Control Charts > Box-Cox Transformation to help you find the optimal transformation value. choose With the first occurrence of these values. or text. then enter the values. and X-bar and S. then click OK. The Options subdialog box lists the common transformations natural log (λ = 0) and square root (λ = 0. . • To select certain tests. Under When to start a new value. you must have at least one subgroup with two or more observations: R. To display a historical chart To define stages in your process.. click control chart Options. 5 Options − Box-Cox . the data must be positive. • To start a new stage at the first occurrence of a certain value.Quality Control To perform the tests for special causes 1 2 Choose Stat > Control Charts > control chart type > control chart > control chart Options > Tests. If you like. choose Perform no tests. Minitab only uses the non-omitted data to find lambda. Minitab treats each repeated value as a separate occurrence. 42 2003 Minitab Inc. X-bar and R. > control chart Options > Stages Charts All charts except Z-MR You can display stages in your process by drawing a "historical chart" − a control chart in which the control limits and center line are estimated independently for different groups in your data. (λ is lambda. Enclose date/time or text entries in double quotes. In Define stages (historical groups) with this variable. The indicators can be numbers.. choose Perform all tests for special causes. Enter the values. choose With each new value." The transformation takes the original data to the power λ. S.

2003 Minitab Inc. see the independent Box-Cox transformation command Stat > Control Charts > Box-Cox Transformation. 43 . 3 Click OK. Click the Display tab. click control chart Options. then enter one or more values. Each segment contains __ subgroups: Choose to split the control chart into subgroups of a specified number of points. Dialog box items Box-Cox power transformation (W = Y**Lambda): Choose to use when your data are very skewed or when the withinsubgroup variation is unstable. Split chart into series of segments for display purposes Do not split: Choose not to split the chart.Control Charts When you use this transformation. > control chart Options > Display Charts All charts except Z-MR Use to display the control chart by stage or by number of plotted points and to display the test results in the Session window. choose Display all subgroups. and enter the number of subgroups for each segment. For help choosing a lambda value. To do the Box-Cox power transformation 1 2 In the control chart dialog box.. Caution If you use Stat > Control Charts > Box-Cox Transformation to find the optimal λ value and choose to store the transformed data with that command. Click the Box-Cox tab. Lambda=0. Subgroups to display Display all subgroups: Choose to display all subgroups. • Choose Optimal lambda to have Minitab search for an optimal value for all stages. under Subgroups to display. Minitab does not accept any values you enter in the Parameters tab for historical means or historical standard deviations. Optimal lambda: Choose to have Minitab search for an optimal value. Display last __ subgroups: Choose to display a specific number of subgroups. To display subgroups 1 2 In the control chart dialog box. Do one of the following: • To display all subgroups in the control chart. Use overall standard deviation: Check to use the overall standard deviation in the estimation of the optimal lambda. Optimal lambda for each stage (in a chart with stages): Choose to have Minitab search for optimal values for each stage. and enter the number of subgroups..5 (square root) to use the square root of the data. then do one of the following: • Choose Lambda = 0 (natural log) to use the natural log of the data. Check Box-Cox power transformation (W = Y**Lambda).5 (square root): Choose to use the square root of the data. • Choose Other (enter value(s) between -5 and 5) to transform the data using another lambda value or values. doing so will double-transform the data. click control chart Options. Test results Display test results in Session window: Check to display the test results in the Session window. do not select the Box-Cox option when you make a control char. Each segment corresponds to a stage (if chart has stages): Choose to split the control chart by stage or stage when you have specified stages in control chart Options > Stage. Other (enter value(s) between -5 and 5): Choose to transform the data using another lambda value or values. then enter the lambda values. • Choose Lambda=0. • Choose Optimal lambda for each stage (in a chart with stages) to have Minitab search for optimal values for each stage. Options − Display . Lambda = 0 (natural log): Choose to use the natural log of the data.

Center line value: Check to store the center line value.Quality Control • 3 To display a specific number of the last subgroups. X-bar. 3 Options − Storage Stat > Control Charts > control chart type > control chart > control chart Options > Storage Charts X-bar and R. CUSUM Use to store your choice of statistics in the worksheet. Stage: Check to store the stage. one row for each plotted point. choose Display last __ subgroups and enter the number of subgroups to display. enter 9. • To split the chart by stage. Subgroup size: Check to store the subgroup sizes. Choose Stat > Control Charts > Variables Charts for Subgroups > Xbar-S. To store statistics in the worksheet 1 2 Choose Stat > Control Charts > control chart type > control chart > control chart Options > Storage. then click OK. one row for each historical stage. one row for each plotted point. 1 2 3 4 Open the worksheet BLOODSUGAR. 44 2003 Minitab Inc. create an X-bar and S chart. Dialog box items Store these estimates for each chart Means: Check to store the estimates of the means. EWMA. choose Each segment contains __ subgroups. . Check the statistics to store in the worksheet. and enter the number of subgroups for each segment. Test results: Check to store the results of any test performed. choose Each segment corresponds to a stage (if chart has stages). Stores these values for each point Point plotted: Check to store the plotted points. Click the Display tab. Moving Average. one row for each stage. Control limit values: Check to store the control limit values. one row for each plotted point. In Subgroup sizes. You take a blood glucose reading every day for each patient for 20 days. Click OK. click control chart Options. To monitor the mean and standard deviation of the blood glucose levels of your patients. Click OK.MTW. Minitab stores one column for the lower control limit and one column for the upper control limit. Click OK. one row for each plotted point. X-bar and S. one row for each plotted point. Example of X-bar and S chart You are conducting a study on the blood glucose levels of 9 patients who are on strict diets and exercise routines. Choose All observations for a chart are in one column. under Subgroups to display. one row for each plotted point. Standard deviations: Check to store the estimates of the standard deviations. then enter Glucoselevel. Do one of the following: • To split the chart into a specific number of subgroups. To split a control chart 1 2 In the control chart dialog box.

Subgroups do not need to be the same size. For more information about control charts. a MR chart. X-bar − a chart of subgroup means R − a chart of subgroup ranges S − a chart of subgroup standard deviations Zone − a chart of the cumulative scores based on each point's distance from the center line The I-MR-R/S (Between/Within) chart requires that you have two or more observations in at least one subgroup. The glucose level and its variability are in control for the nine patients on the diet and exercise program. and a R or S chart. Variables control charts for individuals.Control Charts Graph window output Interpreting the results The glucose level means and standard deviations over the 10-day period fall within the control limits. see Control Charts Overview. time-weighted charts. 45 . such as length or pressure. such as the number of defects or defective units. I-MR-R/S (Between/Within) Chart Variables Control Charts for Subgroups Overview data Variables control charts for subgroups plot statistics from measurement data. Choosing a variables control chart for subgroups Minitab offers five variables control charts: • • • • • • • X-bar and R − an X-bar chart and R chart displayed in one window X-bar and S − an X-bar chart and S chart displayed in one window I-MR-R/S (Between/Within) − a three-way control chart that uses both between-subgroup and within-subgroup variation. which are plotted on the chart and used to estimate process parameters. Minitab calculates summary statistics for each subgroup. and multivariate charts also plot measurement data. Attributes control charts plot count data. An I-MR-R/S chart consists of an I chart. I-MR-R/S (Between/Within) Chart Stat > Control Charts > Variables Charts for Subgroups > I-MR-R/S (Between/Within) Produces a three-way control chart using both between-subgroup and within-subgroup variations. An I-MR-R/S chart consists of: • An individuals chart 2003 Minitab Inc. for subgroup data.

an S chart is displayed. Dialog box items All observations for a chart are in one column: Choose if data are in one or more columns. the only differences should be due to random error. Use this chart. using the between-sample component of variation. For example. and the within-sample standard deviation no longer estimates random error. This eliminates the within-sample component of variation in the control limits. then each column would contain the measurements from one of the locations. This variation due to location is not accounted for. For more information. see Control Charts Overview and Variables Control Charts for Subgroups Overview. and each row would contain the measurements from a single part. If you base estimates on the pooled standard deviation and your subgroup size is nine or greater. the measurements taken at the five locations can also be consistently different in all parts. Minitab displays an S chart if you estimate using pooled standard deviation or an R chart if you estimate using Rbar. Otherwise. Whether Minitab displays an R chart or an S chart depends on the chosen estimation method and the size of the subgroup. to track both process location and process variation. If you specify stages. with each row representing a subgroup. an R chart is displayed. Over time. If you base estimates on the average of subgroup standard deviations. then enter the columns. Suppose you sample one part every hour. . If you base estimates on the pooled standard deviation and your subgroup size is eight or less. and the within-sample component of variation. This process appears to be too good. the overall process variation is due to both between-sample variation and random error. then an R chart is displayed. then enter the columns. causing control limits that are too wide. If you base estimates on the average of subgroup ranges. but actually estimates both random error and the location effect. Subgroup sizes: Enter a number or a column of subscripts. Under these conditions. This results in a standard deviation that is too large. using a moving range to estimate the standard deviation of the distribution of sample means is similar to estimating just the random error component. Perhaps one location almost always produces the largest measurement or is consistently smaller. For example. If the subgroups are not equal. Observations for a subgroup are in one row of columns: Choose if subgroups are arranged in rows across several columns. if you took 5 measurements of each part. R chart or S chart: Charts process variation using the within-sample component of variation. While the parts can vary hour to hour. You can solve this problem by using I-MR-R/S (Between/Within) to create three separate evaluations of process variation: Individuals chart: Charts the means from each sample on an individuals control chart.Quality Control • • A moving range chart An R chart or S chart When collecting data in subgroups. Variation within each sample also contributes to overall process variation. the combination of the three charts provides a method of assessing the stability of process location. along with the Individuals chart. rather than on an X-bar chart. This chart uses a moving range between consecutive means to determine the control limits. Since the distribution of the sample means is related to the random error. and it probably is. with most points on the control chart placed very close to the centerline. Thus. the process could shift or drift. you may want to force the control limits to be constant by specifying a fixed subgroup size using I-MR-R/S Options > Estimate. If you base estimates on the pooled standard deviation and the minimum of the mode sample sizes is eight or less. Do one of the following: 46 2003 Minitab Inc. then an S chart is displayed. Moving range chart: Charts the subgroup means using a moving range to remove the within-sample variation. so the next sample of five parts may be different from the previous sample. random error may not be the only source of variation. the betweensample component of variation. but in different locations. an R chart is displayed. If the subgroup sizes do not vary much. To display a I-MR-R/S (Between/Within) chart 1 2 Choose Stat > Control Charts > Variables Charts for Subgroups > I-MR-R/S (Between/Within). and measure five locations across the part. if you sample five parts in close succession every hour. the display of a R or S chart depends on the minimum of the mode sample size for all stages. each control limit is not a single straight line but varies with the subgroup size. <Scale> <Labels> <Multiple Graphs> <Data Options> <I-M R/S Options> Data − I-MR-R/S (Between/Within) Chart Set up the data in several columns.

you must enter subgroup 2 as the subgroup to omit. enter the historical between subgroups standard deviations as constants. For example. Do any of the following: • In Mean. choose Observations for a subgroup are in one row of columns. then enter one or more columns. Click OK. Dialog box items Mean: Enter the historical means as constants. not just the subsetted data. then click OK. For example. then decide to omit subgroup 2 from the parameter estimation (using control chart Options > Estimate). I-MR-R/S Chart – Options – Parameters Stat > Control Charts > Variables Charts for Subgroups > I-MR-R/S (Between/Within) > I-MR-R/S Options > Parameters Enter historical data for estimating µ and σ. If you enter one constant. the omitted or used samples apply to the entire data set. 47 . otherwise. enter a number or a column of subscripts. • Under Standard deviations. you can prevent these subgroups from influencing the estimation of process parameters. If you like. you exclude subgroup 1 from your analysis (using control chart > Data Options > Subset). A list of values. Within subgroups: Enter the historical standard deviations for within subgroups as constants. all stages use that value. Because Minitab considers an omitted sample (subgroup 1) as part of the entire data set. enter a value for each stage. in Within subgroups. Minitab estimates it from the data. enter a value for each stage. 3 • • • You can enter historical data in several ways: I-MR-R/S Chart − Options − Estimate Stat > Control Charts > Variables Charts for Subgroups > I-MR-R/S (Between/Within) > I-MR-R/S Options > Estimate Omits or includes certain subgroups to estimate µ and σ. if some subgroups have erratic data due to assignable causes that you have corrected. Caution When you include or exclude rows using control chart > Data Options > Subset and also omit or use samples to estimate parameters. one for each stage within each variable. Minitab only uses the nonomitted data to find lambda. enter the historical means as constants. one for each variable. Each stage for each variable uses a different value. force the control limits to be constant when you have unequal sample sizes. otherwise. To use historical values for the mean and standard deviation 1 2 Choose Stat > Control Charts > Variables Charts for Subgroups > I-MR-R/S (Between/Within) > I-MR-R/S Options > Parameters. choose All observations for a chart are in one column. For example. A list of values. If you enter one constant. and use biased estimates of σ. then enter a series of columns. which applies to all variables and stages. When you include or exclude rows using control chart Options > Estimate and choose to perform a Box-Cox transformation. 2003 Minitab Inc. You can set preferences for the estimation of σ and the length of the moving range using Tools > Options > Control Charts and Quality Tools > Estimating Standard Deviation. A single value. In Subgroup sizes. all stages use that value. You can omit or include certain subgroups when estimating parameters using Estimate. Standard deviations Between subgroups: Enter the historical standard deviations for between subgroups as constants. If you do not specify a value for the mean or standard deviation. All stages for each variable use the same value. enter the historical within subgroups standard deviations as constants. use any dialog box options. you may enter this value. You can also select one of several methods to estimate σ. if you have a known process parameter or an estimate obtained from past data.Control Charts • • 3 If subgroups are in one or more columns. in Between subgroups. • Under Standard deviations. If subgroups are in rows. change the length of the moving range.

choose a method. The individuals chart of the subgroup means gives an estimate. except with the Individuals. By default. Enter entire subgroups. For example. To choose how the standard deviation is estimated 1 2 In the control chart dialog box. The within components. Minitab excludes any moving ranges with omitted data from the calculations. Xij = µ + Bi + Wj(i) i = 1. • Choose Use the following samples when estimating parameters and enter the subgroups or observation numbers you want to use for the calculations. Click the Estimate tab. Xij. To omit or use subgroups for the estimates of parameters By default.. . and a within component. Sbar: Choose to estimate σ based on the average of the subgroup standard deviations. Subgroup Size > 1 Rbar: Choose to estimate σ based on the average of the subgroup ranges. Do one of the following: • Choose Omit the following samples when estimating parameters and enter the subgroups or observation numbers you want to omit from the calculations. Then click OK. . With these charts.Quality Control Dialog box items Omit the following samples when estimating parameters (eg. Moving Range. not individual observations. Bi. Use unbiasing constant: Uncheck to use biased estimates of σ. of Var (xbari). Use the following samples when estimating parameters (eg. a span of two is used because consecutive values have the greatest chance of being alike. Minitab assumes the values you enter are subgroup numbers. The model assumes that an observation. click control chart Options. 3 12:15): Choose to specify subgroups of data to compute µ and σ.. 1 2 In the control chart dialog box.. you might want to use the estimate for the total variation as the historical standard deviation in a normal capability analysis. The total variance of the observation. as shown below. are assumed to arise from a normal distribution with mean 0 and variance σ2B. change that value to missing (∗). µ. Click the Estimate tab. Wj(i). Subgroup Size = 1 Average moving range: Choose to estimate σ using the average of the moving range. j = 1. Median moving range: Choose to estimate σ using the median of the moving range. and I-MR charts. you can choose to display three standard deviations in the Session window. is the sum of an overall mean. When you use Omit the following samples when estimating parameters with Length of moving range. then enter individual subgroup numbers or ranges of subgroups. are assumed to be independent of the Bi and arise from a normal distribution with mean 0 and variance σ2w. Note 3 Click OK. Length of moving range: Enter the number of observations used to calculate the moving range. Minitab estimates the process parameters from all the data. and the total variation.. σ2T. Here are the model assumptions used in estimating these three standard deviations. a between component. is the sum of σ2B and σ2w. then type individual subgroup numbers or a range of subgroups. the variance of the subgroup means: Var (xbari) = σ2 B + σ2w / J 48 2003 Minitab Inc. The span must be < 100. Note Estimates of three standard deviations If you do not estimate parameters by stages.. These standard deviations can then be used as parameters in other analyses.. I. Wj(i). the values are interpreted as observation numbers. the within component. Pooled standard deviation: Choose to estimate σ using a pooled standard deviation. To omit one observation from a sample. click control chart Options. J The between components. . say V1. These are the standard deviations for the between component. Bi. 3 12:15): Choose to specify subgroups of the data to omit from the computations for µ and σ. Method for estimating standard deviation: See [2] for a discussion of the relative merits of the methods listed below. Under Estimating standard deviation. But you may want to use or omit certain data if they show abnormal behavior. Xij.

• To set bounds on the control limits for the individuals chart: under Place bounds on control limits of Individuals chart. 49 . Place bounds on control limits of Moving range chart Lower standard deviation limit bound: Check to set a lower bound for the control limits for the moving range chart. We can combine V1 and V2 to get estimates of σ2B and σ2w. Dialog box items Display control limits at: These multiples of the standard deviation: Enter one or more values. If the calculated upper control limit is greater than the upper bound. It is possible that the estimate for σ2B is zero. When the calculated upper control limit is greater than the upper bound. V1 − V2 / J ) and estimate of σ2w = V1 + (J − 1) ∗ V2 / J. and below the mean. For an example. a horizontal line labeled UB will be drawn at the upper bound instead. If the calculated lower control limit is less than the lower bound. above. • To set bounds on the control limits for the moving range chart: under Place bounds on control limits of Moving range chart. If the calculated lower control limit is less than the lower bound. You can set bounds on the upper and lower control limits. a horizontal line labeled LB will be drawn at the lower bound instead. The values entered are the number of standard deviations above and below the center line. If you specify more than one value. If the calculated upper control limit is greater than the upper bound. Similarly. I-MR-R/S Chart − Options − S Limits Stat > Control Charts > Variables Charts for Subgroups > I-MR-R/S (Between/Within) > I-MR-R/S Options > S Limits Displays control limits above and below the mean at the multiples of any standard deviation. a horizontal line labeled LB will be drawn at the lower bound instead. 2003 Minitab Inc. check Lower standard deviation limit bound (and/or Upper standard deviation limit bound) and enter a value. Do any of the following: • To specify where control limits are drawn: under Display control limits at. a horizontal line labeled LB is drawn at the lower bound instead. see Example of an X-bar chart with tests and customized control limits. a pair of control limits is drawn for each number in the list. for σ2w. Upper standard deviation limit bound: Check to set an upper bound for the control limits for the moving range chart. the within component. Each value is the number of standard deviations the lines should be drawn at. a horizontal line labeled UB is drawn at the upper bound instead. as estimate of σ2B = max(0. check Lower standard deviation limit bound (and/or Upper standard deviation limit bound) and enter a value. Each value represents the number of standard deviations below and above the mean. if the calculated lower control limit is less than the lower bound. Each value represents the number of standard deviations below and above the mean. a horizontal line labeled LB will be drawn at the lower bound instead. If the calculated lower control limit is less than the lower bound. a horizontal line labeled UB will be drawn at the upper bound instead. click I-MR-R/S Options.Control Charts The Range Chart gives an estimate. say V2. Place bounds on control limits of R or S chart Lower standard deviation limit bound: Check to set a lower bound for the control limits for the R or S chart. Upper standard deviation limit bound: Check to set an upper bound for the control limits for the R or S chart. enter one or more values in These multiples of the standard deviation. If the calculated upper control limit is greater than the upper bound. Click the S Limits tab. Place bounds on control limits of Individuals chart Lower standard deviation limit bound: Check to set a lower bound for the control limits for the individuals chart. This is true because the R chart for the Xbarij is the same as the R chart for the (unobservable) Wj(i). a horizontal line labeled UB will be drawn at the upper bound instead. To customize the control limits for I-MR-R/S charts 1 2 In the control chart data source dialog box. Upper standard deviation limit bound: Check to set an upper bound for the control limits for the individuals chart. Tip You can also modify the control limits using the Minitab graph editing features.

choose Perform all tests for special causes. I-MR-R/S (Between/Within). choose Perform the following tests for special causes and check the tests to perform. See Tools > Options > Control Charts and Quality Tools > Other. If a point fails more than one test. You can set preferences for the sensitivity of these tests using Tools > Options > Control Charts and Quality Tools > Define Tests. See [1] and [29] for guidance on using these tests. The test results are displayed in the Session window. When a point fails a test. . 3 Click OK in each dialog box. Individuals. 50 2003 Minitab Inc. See Updating Graphs for more information. • To perform no tests. choose Perform no tests. • To select certain tests. Options − Tests . Perform the following tests for special causes: Choose to specify a subset of the eight tests.. check Lower standard deviation limit bound (and/or Upper standard deviation limit bound) and enter a value. Dialog box items Tests For Special Causes (default definitions): Perform all tests for special causes: Choose to perform all eight tests shown below. The occurrence of a pattern suggests a special cause for the variation. Minitab marks it with the test number on the plot.. Each value represents the number of standard deviations below and above the mean.. alternating up and down 2 out of 3 points > 2 standard deviations from center line (same side) 4 out of 5 points > 1 standard deviation from center line (same side) 15 points in a row within 1 standard deviation of center line (either side) 8 points in a row > 1 standard deviation from center line (either side) Perform no tests: Choose to perform no tests for special causes. 1 point more than 3 standard deviations from center line 9 points in a row on same side of center line 6 points in a row. Minitab will perform tests when subgroup sizes are unequal. the test results are not updated in the Session window. > control chart Options > Tests Charts X-bar and R. Click OK. Minitab marks it by the lowest numbered test.Quality Control • To set bounds on the control limits for the R or S chart: under Place bounds on control limits of R or S Chart. all increasing or all decreasing 14 points in a row. Each test detects a specific pattern in the data plotted on the chart. then check any combination of the tests listed below. > control chart Options > Stages Charts All charts except Z-MR You can display stages in your process by drawing a "historical chart" − a control chart in which the control limits and center line are estimated independently for different groups in your data. 3 Options − Stages . Display test results in Session window: Check to display the test results in the Session window. You can choose to change the color of a control chart when the last row of data causes a new test failure for any point.. To perform the tests for special causes 1 2 Choose Stat > Control Charts > control chart type > control chart > control chart Options > Tests. X-bar. Do one of the following: • To do all tests. one that should be investigated. If the data change and the control chart is live-linked. Historical charts are particularly useful for comparing data before and after a process improvement. X-bar and S. I-MR Selects a subset of the tests shown in Eight Tests for Special Causes.

." The transformation takes the original data to the power λ. you must have at least one subgroup with two or more observations: R. dates.Control Charts Note With the following charts. the data must be positive.5 (square root): Choose to use the square root of the data. Under When to start a new value. With the first occurrence of these values: Choose to start a new stage at the first occurrence of a certain value. and X-bar and S. Enclose date/time or text entries in double quotes. in which case the natural log is taken. Minitab does not accept any values you enter in the Parameters tab for historical means or historical standard deviations. do not select the Box-Cox option when you make a control char.) To use this option. When to start a new stage With each new value: Choose to start a new stage each time the value in the column changes. The indicators can be numbers. do one of the following: • To start a new stage each time the value in the column changes. choose With each new value. 51 . You can also choose any value between − 5 and 5 for λ. When you use this transformation. In Define stages (historical groups) with this variable. The column must be the same length as the data column (or columns. or text. Minitab only uses the non-omitted data to find lambda. 2003 Minitab Inc. 5 Options − Box-Cox . When you include or exclude rows using control chart Options > Estimate. you can tell Minitab to start a new stage in one of two ways: • • 1 2 3 4 Each time the value in the column changes At the first occurrence of one or more values In the control chart dialog box. (λ is lambda. use any dialog box options. Caution If you use Stat > Control Charts > Box-Cox Transformation to find the optimal λ value and choose to store the transformed data with that command. You can enter the same value more than once. then click OK. Optimal lambda: Choose to have Minitab search for an optimal value.5). Dialog box items Define stages (historical groups) with this variable: Enter the column that contains the stage indicators. The Options subdialog box lists the common transformations natural log (λ = 0) and square root (λ = 0. Lambda = 0 (natural log): Choose to use the natural log of the data. X-bar and R. Minitab treats each repeated value as a separate occurrence. click control chart Options. If you like. Dialog box items Box-Cox power transformation (W = Y**Lambda): Choose to use when your data are very skewed or when the withinsubgroup variation is unstable. • To start a new stage at the first occurrence of a certain value. You can enter the same value more than once. When executing the command. Use overall standard deviation: Check to use the overall standard deviation in the estimation of the optimal lambda. you should not choose a λ outside the range of −2 and 2. unless λ = 0. Minitab treats each repeated value as a separate occurrence. Enter the values.. Click the Stages tab. enter the column which contains the stage indicators. To display a historical chart To define stages in your process. You may want to first run the command Stat > Control Charts > Box-Cox Transformation to help you find the optimal transformation value. then enter the values. In most cases. Lambda=0. you must set up a column of grouping indicators. Enclose date/time or text entries in double quotes. when subgroups are across rows). Optimal lambda for each stage (in a chart with stages): Choose to have Minitab search for optimal values for each stage. S. doing so will double-transform the data. > control chart Options > Box-Cox Charts All except the Attributes charts You can use the Box-Cox power transformation when your data are very skewed or when the within-subgroup variation is unstable to make the data "more normal. choose With the first occurrence of these values.

• To split the chart by stage. and enter the number of subgroups for each segment. • Choose Other (enter value(s) between -5 and 5) to transform the data using another lambda value or values. click control chart Options. then do one of the following: • Choose Lambda = 0 (natural log) to use the natural log of the data. click control chart Options. To do the Box-Cox power transformation 1 2 In the control chart dialog box. • To display a specific number of the last subgroups. Display last __ subgroups: Choose to display a specific number of subgroups. Each segment corresponds to a stage (if chart has stages): Choose to split the control chart by stage or stage when you have specified stages in control chart Options > Stage.. • Choose Optimal lambda to have Minitab search for an optimal value for all stages. Options − Display . • Choose Lambda=0. Check Box-Cox power transformation (W = Y**Lambda). click control chart Options.Quality Control Other (enter value(s) between -5 and 5): Choose to transform the data using another lambda value or values. choose Display all subgroups. Split chart into series of segments for display purposes Do not split: Choose not to split the chart. under Subgroups to display. then enter the lambda values. Click the Box-Cox tab. Click OK. • Choose Optimal lambda for each stage (in a chart with stages) to have Minitab search for optimal values for each stage. choose Each segment corresponds to a stage (if chart has stages). then enter one or more values. To display subgroups 1 2 In the control chart dialog box. . Do one of the following: • To split the chart into a specific number of subgroups.5 (square root) to use the square root of the data. choose Each segment contains __ subgroups. For help choosing a lambda value. Click OK. Click the Display tab. 3 Click OK. Subgroups to display Display all subgroups: Choose to display all subgroups. Do one of the following: • To display all subgroups in the control chart. see the independent Box-Cox transformation command Stat > Control Charts > Box-Cox Transformation. under Subgroups to display. and enter the number of subgroups. choose Display last __ subgroups and enter the number of subgroups to display. Each segment contains __ subgroups: Choose to split the control chart into subgroups of a specified number of points. 3 To split a control chart 1 2 In the control chart dialog box.. 3 52 2003 Minitab Inc. > control chart Options > Display Charts All charts except Z-MR Use to display the control chart by stage or by number of plotted points and to display the test results in the Session window. Click the Display tab. and enter the number of subgroups for each segment. Test results Display test results in Session window: Check to display the test results in the Session window.

Store these values for each point Point plotted: Check to store the plotted points.Control Charts I-MR-R/S Chart − Options − Storage Stat > Control Charts > Variables Charts for Subgroups > I-MR-R/S (Between/Within) > I-MR-R/S Options > Storage Stores your choice of statistics in the worksheet.0000 13. Session window output I-MR-R/S (Between/Within) Chart of Coating Test Results for I Chart of Subgroup Means of Coating TEST 1. 8. one row for each plotted point. You take 3 samples from 15 consecutive rolls and measure coating weight. Choose All observations for a chart are in one column.1225 2003 Minitab Inc.00 standard deviations from center line. Control limit values: Check to store the control limit values. one row for each plotted point. Between standard deviation: Check to store the between standard deviation. one row for each plotted point. Within standard deviation: Check to store the within standard deviation. Standard deviations: Check to store the estimates of the standard deviations. Because you are interested in whether or not the coating is even throughout a roll and whether each roll is correctly coated. then enter Coating. 1 2 3 4 Open the worksheet COATING. you use Minitab to create an I-MR-R/S chart. Center line value: Check to store the center line value. Between/within standard deviation: Check to store the between/within standard deviation. You are concerned that the paper is being coated with the correct thickness of film and that the coating is evenly distributed across the length of the roll. Check the statistics to store in the worksheet. one row for each plotted point. Test results: Check to store the results of any test performed. Example of I-MR-R/S Chart Suppose you are interested in determining whether or not a process that coats rolls of paper with a thin film is in control [31]. one row for each plotted point.1225 13. 53 . 15 I-MR-R/S Standard Deviations of Coating Standard Deviations Between Within Between/Within 0. One point more than 3. Test Failed at points: 7. Subgroup size: Check to store the subgroup sizes. Minitab stores one column for the lower control limits and one column for the upper control limits.MTW. Dialog box items Store these estimates for each chart Means: Check to store estimates of the means. Stage: Check to store the stage. one row for each stage. In Subgroup sizes. To store statistics in the worksheet 1 2 Choose Stat > Control Charts > control chart type > control chart > control chart Options > Storage. enter Roll. one row for each plotted point. then click OK. 14. 9. one row for each stage. Click OK. Choose Stat > Control Charts > Variables Charts for Subgroups > I-MR-R/S (Between/Within).

which are plotted on the chart and used to estimate process parameters. 54 2003 Minitab Inc. For more information about control charts. a MR chart.Quality Control * WARNING * If graph is updated with new data. An I-MR-R/S chart consists of an I chart. Subgroups do not need to be the same size. suggesting that this process is out of control. and multivariate charts also plot measurement data. such as the number of defects or defective units. X-bar Chart Stat > Control Charts > Variables Charts for Subgroups > Xbar A control chart of subgroup means. Attributes control charts plot count data. . see Control Charts Overview. and a R or S chart. Graph window output Interpreting the results The individuals chart shows five points outside the control limits. such as length or pressure. time-weighted charts. Variables control charts for individuals. Xbar Variables Control Charts for Subgroups Overview data Variables control charts for subgroups plot statistics from measurement data. X-bar − a chart of subgroup means R − a chart of subgroup ranges S − a chart of subgroup standard deviations Zone − a chart of the cumulative scores based on each point's distance from the center line The I-MR-R/S (Between/Within) chart requires that you have two or more observations in at least one subgroup. Minitab calculates summary statistics for each subgroup. You can use X-bar charts to track the process level and detect the presence of special causes. the results above may no * longer be correct. Choosing a variables control chart for subgroups Minitab offers five variables control charts: • • • • • • • X-bar and R − an X-bar chart and R chart displayed in one window X-bar and S − an X-bar chart and S chart displayed in one window I-MR-R/S (Between/Within) − a three-way control chart that uses both between-subgroup and within-subgroup variation. for subgroup data.

and what tests to perform. You can set options for control charts in all Minitab sessions using Tools > Options > Control Charts and Quality Tools. all affected dialog box settings automatically reflect your preferences. σ. Available options include how to estimate σ. then enter the columns. For more information. Worksheet Structure Structure your data down a column or across rows. If the subgroups are not equal. Dialog box items All observations for a chart are in one column: Choose if data are in one or more columns. see Control Charts Overview and Variables Control Charts for Subgroups Overview. Observations for a subgroup are in one row of columns: Choose if subgroups are arranged in rows across several columns. Subgroups are equal size Univariate (one variable) Multivariate (more than one variable) Down columns or across rows Down columns Subgroups are unequal size Down columns with subgroup indicator column Down columns with subgroup indicator column Structure subgroup data down a column or across rows. 2003 Minitab Inc. 55 .Control Charts By default. using the following table as a guide. the second 5 observations are the second row. Variables charts include: • • • • Variables charts for subgroups Variables charts for individuals Time-weighted charts Multivariate charts Structure your data for these charts using the guidelines below. <Scale> <Labels> <Multiple Graphs> <Data Options> <Xbar Options> Data − Variables Control Chart Organize the data for all variables control charts in the same way. You can also base the estimate on the average of the subgroup ranges or standard deviations. with one column for each variable. or enter a historical value for σ. using a pooled standard deviation. test definitions. Minitab's X-bar chart estimates the process variation. Multivariate data must be entered down columns. Subgroup sizes: Enter a number or a column of subscripts. you may want to force the control limits to be constant by specifying a fixed subgroup size using Xbar Options > Estimate. Here is the same data set. structured both ways. each control limit is not a single straight line but varies with the subgroup size. and so on. If the subgroup sizes do not vary much. When you set options. then enter the columns. with subgroups of size 5. Note that the first five observations in the left data set (subgroup 1) are the first row of the right-side data set.

3 56 2003 Minitab Inc. In this example. • If subgroups are in rows. In Subgroup sizes. or by using the stand alone Box-Cox command. C1 contains the process data and C2 contains subgroup indicators: Each time a subscript changes in C2. Missing data See Missing data in control charts for information on how to handle missing data for different types of control charts. you must enter your data in one column. subgroup 2 has six observations. a new subgroup begins in C1. use any dialog box options. you may want to use the Box-Cox transformation to induce normality. see Options − Box-Cox. If you like. choose All observations for a chart are in one column. . and so on. then click OK. Use the stand alone command as an exploratory tool to help you determine the best lambda value for the transformation. You can access the Box-Cox transformation two ways: by using the Box-Cox transformation option provided with the control chart commands. subgroup 1 has three observations. you can use the transformation option to transform the data at the same time you draw the control chart. you must enter data that approximate a normal distribution. then create a second column of subscripts which serve as subgroup indicators.Quality Control When subgroups are of unequal size. Nonnormal data To properly interpret Minitab's control charts. then enter one or more columns. In the following example. see Box-Cox Transformation. For information on the stand alone Box-Cox transformation command. choose Observations for a subgroup are in one row of columns. If the data are highly skewed. Do one of the following: • If subgroups are in one or more columns. For information on the Box-Cox transformation option. To display an X-bar chart 1 2 Choose Stat > Control Charts > Variables Charts for Subgroups > Xbar. then enter a series of columns. Then. enter a number or a column of subscripts.

. you can enter these values. • In Standard deviation. you exclude subgroup 1 from your analysis (using control chart > Data Options > Subset). 57 . To use historical mean and standard deviation values 1 2 In the control chart dialog box. A single value. 3 12:15): Choose to specify subgroups of data to compute µ and σ. Enter entire subgroups.. Dialog box items Mean: Enter the historical means as constants. When you include or exclude rows using control chart Options > Estimate and choose to perform a Box-Cox transformation. EWMA Use to enter historical data for estimating µ and σ. the omitted or used samples apply to the entire data set. Caution When you include or exclude rows using control chart > Data Options > Subset and also omit or use samples to estimate parameters. For example. For example. not individual observations. change that value to missing (∗). Minitab only uses the non-omitted data to find lambda. then decide to omit subgroup 2 from the parameter estimation (using control chart Options > Estimate). If you do not specify a value for µ or σ. not just the subsetted data. 2003 Minitab Inc. Each stage for each variable uses a different value. if some subgroups have erratic data due to assignable causes that you have corrected. Dialog box items Omit the following samples when estimating parameters (eg. A list of values. one for each variable. Sbar: Choose to estimate σ based on the average of the subgroup standard deviations. 3 • • • You can enter historical data in several ways: X-bar Chart − Options − Estimate Stat > Control Charts > Variables Charts for Subgroups > Xbar Chart > Xbar Options > Estimate Allows you to omit or use certain subgroups to estimate µ and σ. then enter individual subgroup numbers or a range of subgroups. you must enter subgroup 2 as the subgroup to omit. force the control limits to be constant when you have unequal sample sizes. You can also select one of two methods to estimate σ. For example. and use biased estimates of σ. Use the following samples when estimating parameters (eg. which applies to all variables and stages. Minitab estimates it from the data. All stages for each variable use the same value. if you have a known process parameter or an estimate obtained from past data. X-bar and S. Method for estimating standard deviation: See [2] for a discussion of the relative merits of the methods listed below. click control chart Options. Use unbiasing constants: Uncheck to use biased estimates of σ. Moving Average. Click OK. one for each stage within each variable. You can omit or include certain subgroups when estimating parameters using Estimate. A list of values. 3 12:15): Choose to specify subgroups of data to omit from the computations for µ and σ. Because Minitab considers an omitted sample (subgroup 1) as part of the entire data set. Standard deviation: Enter the historical standard deviations as constants. I-MR.Control Charts Options − Parameters . Do any of the following: • In Mean. X-bar. enter one or more values. Individuals. you can prevent these subgroups from influencing the estimation. Zone. You can set preferences for the estimation of σ using Tools > Options > Control Charts and Quality Tools > Estimating Standard Deviation. Pooled standard deviation: Choose to estimate σ using a pooled standard deviation. enter one or more values. To omit one observation from a subgroup. > control chart Options > Parameters Charts X-bar and R. Subgroup size > 1 Rbar: Choose to estimate σ based on the average of the subgroup ranges. then enter individual subgroup numbers or a range of subgroups.

Assuming all subgroups have size: Choose to estimate the control limits using a specified subgroup size. You can set bounds on the upper and/or lower control limits. Then click OK. Dialog box items Display control limits at These multiples of the standard deviation: Enter one or more values.. Only the control limits are affected. Similarly. Note 3 Click OK. each control limit is not a single straight line. The values entered are the number of standard deviations above and below the center line. a horizontal line labeled LB is drawn at the lower bound instead. choose a method. except with the Individuals. . Note S Limits . U. P. Upper standard deviation limit bound: Check to set an upper bound for the control limits. When subgroup sizes are not equal. To choose how the standard deviation is estimated 1 2 In the control chart dialog box. click control chart Options. • Choose Use the following samples when estimating parameters and enter the subgroups or observation numbers you want to use for the calculations. If the calculated upper control limit is greater than the upper bound. Click the Estimate tab. a pair of control limits is drawn for each value in the list. Each value is the number of standard deviations the lines should be drawn at. then type the subgroup size. S. and below the mean. 58 2003 Minitab Inc. R.Quality Control To omit or use subgroups for the estimates of parameters By default. But you may want to use or omit certain data if they show abnormal behavior. When the calculated upper control limit is greater than the upper bound. Minitab estimates the process parameters from all the data. Click the S Limits tab. Place bounds on control limits Lower standard deviation limit bound: Check to set a lower bound for the control limits. Under Estimating standard deviation. Minitab assumes the values you enter are subgroup numbers. Click the Estimate tab. If the sizes do not vary much. and I-MR charts. calculate control limits Using actual sizes of the subgroups: Choose to use the subgroups defined in the main dialog box to estimate σ. Tip You can also modify the control limits using Minitab's graph editing features. you may want to force the control limits to be constant by entering a sample size in this text box. if the calculated lower control limit is less than the lower bound. When you use Omit the following samples when estimating parameters with Length of moving range.. When subgroup sizes are unequal. enter one or more values in These multiples of the standard deviation. NP. see Example of an X-bar chart with tests and customized control limits. the values are interpreted as observation numbers. click control chart Options. > control chart Options > S Limits Charts X-bar. a horizontal line labeled UB will be drawn at the upper bound instead. Do one of the following: • Choose Omit the following samples when estimating parameters and enter the subgroups or observation numbers you want to omit from the calculations. Moving Average. but varies with the subgroup size. You can also force the control limits to be constant when sample sizes are unequal. To customize the control limits 1 2 In the control chart dialog box. If you specify more than one value. EWMA Allows you to draw control limits above and below the mean at the multiples of any standard deviation. Minitab excludes any moving ranges with omitted data from the calculations. 1 2 In the control chart dialog box. With these charts. Do any of the following: • To specify where control limits are drawn: under Display control limits at. If the calculated lower control limit is less than the lower bound. a horizontal line labeled LB will be drawn at the lower bound instead. a horizontal line labeled UB is drawn at the upper bound instead. For an example. Moving Range. the plotted data are not changed. click control chart Options. above.

See Updating Graphs for more information. choose Perform no tests.. 59 . alternating up and down 2 out of 3 points > 2 standard deviations from center line (same side) 4 out of 5 points > 1 standard deviation from center line (same side) 15 points in a row within 1 standard deviation of center line (either side) 8 points in a row > 1 standard deviation from center line (either side) Perform no tests: Choose to perform no tests for special causes. enter a value of 6 to calculate the control limits and center line as if all subgroup sizes were 6. You can choose to change the color of a control chart when the last row of data causes a new test failure for any point. You should force the control limits and center line to be constant only when the difference in size between the largest and smallest subgroups is no more than 25%. When a point fails a test. I-MR-R/S (Between/Within). Each test detects a specific pattern in the data plotted on the chart. To force control limits and center line to be constant 1 2 In the control chart dialog box. 3 Note Options − Tests . See [1] and [29] for guidance on using these tests. all increasing or all decreasing 14 points in a row. the test results are not updated in the Session window. 3 Click OK in each dialog box. click control chart Options.. You can set preferences for the sensitivity of these tests using Tools > Options > Control Charts and Quality Tools > Define Tests. If a point fails more than one test. X-bar.Control Charts • To set bounds on the control limits: check Lower standard deviation limit bound (and/or Upper standard deviation limit bound) and enter a value. Perform the following tests for special causes: Choose to specify a subset of the eight tests. The test results are displayed in the Session window. For example. • To select certain tests. Minitab will perform tests when subgroup sizes are unequal. I-MR Selects a subset of the tests shown in Eight Tests for Special Causes. Display test results in Session window: Check to display the test results in the Session window. then check any combination of the tests listed below. Dialog box items Tests For Special Causes (default definitions): Perform all tests for special causes: Choose to perform all eight tests shown below. Minitab marks it by the lowest numbered test. See Tools > Options > Control Charts and Quality Tools > Other. Do one of the following: • To do all tests. To perform the tests for special causes 1 2 Choose Stat > Control Charts > control chart type > control chart > control chart Options > Tests. If the data change and the control chart is live-linked. You can use this method because the size difference is 25% (125 / 100 = 25%). 2003 Minitab Inc. For example. > control chart Options > Tests Charts X-bar and R. X-bar and S. Each value represents the number of standard deviations below and above the mean. Click OK. calculate control limits. choose Perform all tests for special causes. Under When subgroup sizes are unequal. Minitab marks it with the test number on the plot. one that should be investigated. 1 point more than 3 standard deviations from center line 9 points in a row on same side of center line 6 points in a row. • To perform no tests. Click the S Limits tab. The occurrence of a pattern suggests a special cause for the variation. choose Assuming all subgroups have size. choose Perform the following tests for special causes and check the tests to perform. suppose the largest subgroup is size 125 and the smallest is 100. Individuals. then enter a value.

In Define stages (historical groups) with this variable. do one of the following: • To start a new stage each time the value in the column changes. choose With each new value. then click OK. Enter the values. 5 Options − Box-Cox . Options − Stages . you must set up a column of grouping indicators. When you use this transformation. Minitab treats each repeated value as a separate occurrence. Enclose date/time or text entries in double quotes. The Options subdialog box lists the common transformations natural log (λ = 0) and square root (λ = 0. the data must be positive. . With the first occurrence of these values: Choose to start a new stage at the first occurrence of a certain value. and X-bar and S. If you like. You may want to first run the command Stat > Control Charts > Box-Cox Transformation to help you find the optimal transformation value." The transformation takes the original data to the power λ. click control chart Options. Click the Stages tab. or text.. When executing the command.. To display a historical chart To define stages in your process. doing so will double-transform the data.) To use this option. use any dialog box options. When to start a new stage With each new value: Choose to start a new stage each time the value in the column changes. Enclose date/time or text entries in double quotes. you can tell Minitab to start a new stage in one of two ways: • • 1 2 3 4 Each time the value in the column changes At the first occurrence of one or more values In the control chart dialog box. > control chart Options > Stages Charts All charts except Z-MR You can display stages in your process by drawing a "historical chart" − a control chart in which the control limits and center line are estimated independently for different groups in your data.Quality Control 3 Click OK. Note With the following charts. enter the column which contains the stage indicators. The indicators can be numbers. 60 2003 Minitab Inc. S. The column must be the same length as the data column (or columns. X-bar and R. you should not choose a λ outside the range of −2 and 2. Historical charts are particularly useful for comparing data before and after a process improvement. Caution If you use Stat > Control Charts > Box-Cox Transformation to find the optimal λ value and choose to store the transformed data with that command. dates. In most cases. • To start a new stage at the first occurrence of a certain value. when subgroups are across rows). Dialog box items Define stages (historical groups) with this variable: Enter the column that contains the stage indicators. When you include or exclude rows using control chart Options > Estimate. You can enter the same value more than once. > control chart Options > Box-Cox Charts All except the Attributes charts You can use the Box-Cox power transformation when your data are very skewed or when the within-subgroup variation is unstable to make the data "more normal. unless λ = 0.. choose With the first occurrence of these values. you must have at least one subgroup with two or more observations: R.5). (λ is lambda. in which case the natural log is taken. You can also choose any value between − 5 and 5 for λ. Minitab treats each repeated value as a separate occurrence. Under When to start a new value. do not select the Box-Cox option when you make a control char. Minitab does not accept any values you enter in the Parameters tab for historical means or historical standard deviations. You can enter the same value more than once.. then enter the values. Minitab only uses the non-omitted data to find lambda.

To display subgroups 1 2 In the control chart dialog box. Split chart into series of segments for display purposes Do not split: Choose not to split the chart. choose Display last __ subgroups and enter the number of subgroups to display. then enter the lambda values. For help choosing a lambda value..5 (square root): Choose to use the square root of the data. 61 . see the independent Box-Cox transformation command Stat > Control Charts > Box-Cox Transformation. Each segment contains __ subgroups: Choose to split the control chart into subgroups of a specified number of points. choose Display all subgroups.. Do one of the following: • To display all subgroups in the control chart. Each segment corresponds to a stage (if chart has stages): Choose to split the control chart by stage or stage when you have specified stages in control chart Options > Stage. Subgroups to display Display all subgroups: Choose to display all subgroups. under Subgroups to display. Click the Display tab. Other (enter value(s) between -5 and 5): Choose to transform the data using another lambda value or values. Lambda = 0 (natural log): Choose to use the natural log of the data. then do one of the following: • Choose Lambda = 0 (natural log) to use the natural log of the data. • Choose Optimal lambda for each stage (in a chart with stages) to have Minitab search for optimal values for each stage. click control chart Options. and enter the number of subgroups.5 (square root) to use the square root of the data. click control chart Options.Control Charts Dialog box items Box-Cox power transformation (W = Y**Lambda): Choose to use when your data are very skewed or when the withinsubgroup variation is unstable. and enter the number of subgroups for each segment. • To display a specific number of the last subgroups. • Choose Optimal lambda to have Minitab search for an optimal value for all stages. Display last __ subgroups: Choose to display a specific number of subgroups. 3 2003 Minitab Inc. Click the Box-Cox tab. Options − Display . Lambda=0. Check Box-Cox power transformation (W = Y**Lambda). under Subgroups to display. Optimal lambda: Choose to have Minitab search for an optimal value. then enter one or more values. Use overall standard deviation: Check to use the overall standard deviation in the estimation of the optimal lambda. Optimal lambda for each stage (in a chart with stages): Choose to have Minitab search for optimal values for each stage. • Choose Other (enter value(s) between -5 and 5) to transform the data using another lambda value or values. To do the Box-Cox power transformation 1 2 In the control chart dialog box. • Choose Lambda=0. 3 Click OK. > control chart Options > Display Charts All charts except Z-MR Use to display the control chart by stage or by number of plotted points and to display the test results in the Session window. Click OK. Test results Display test results in Session window: Check to display the test results in the Session window.

Do one of the following: • To split the chart into a specific number of subgroups. Center line value: Check to store the center line value. In Subgroup sizes. Click the Display tab.Quality Control To split a control chart 1 2 In the control chart dialog box. • To split the chart by stage. parts of the crankshaft move up and down a certain distance from an ideal baseline position. enter 5. choose Each segment contains __ subgroups. one row for each plotted point. Stores these values for each point Point plotted: Check to store the plotted points. Dialog box items Store these estimates for each chart Means: Check to store the estimates of the means. one row for each historical stage. then click the Tests tab. Click Xbar Options. click control chart Options. Example of X-bar chart with tests and customized control limits Suppose you work at a car assembly plant in a department that assembles engines. and then ten per day from the 18th through the 25th. one row for each plotted point. choose Each segment corresponds to a stage (if chart has stages). 62 2003 Minitab Inc. Moving Average. Choose All observations for a chart are in one column. 3 Options − Storage Stat > Control Charts > control chart type > control chart > control chart Options > Storage Charts X-bar and R. AtoBDist is the distance (in mm) from the actual (A) position of a point on the crankshaft to the baseline (B) position. To ensure production quality. Under Display control limits at. You want to draw an X-bar chart to track the process level through that time period. Click OK in each dialog box. then click OK. from September 28 through October 15. one row for each stage. . EWMA. Choose Perform all tests for special causes. X-bar and S. X-bar. In an operating engine. 1 2 3 4 5 6 7 8 Open the worksheet CRANKSH. one row for each plotted point. Click OK. To store statistics in the worksheet 1 2 Choose Stat > Control Charts > control chart type > control chart > control chart Options > Storage. Test results: Check to store the results of any test performed. Stage: Check to store the stage. one row for each plotted point. Minitab stores one column for the lower control limit and one column for the upper control limit. enter 1 2 3 in These multiples of the standard deviation. Subgroup size: Check to store the subgroup sizes. Control limit values: Check to store the control limit values.MTW. you took five measurements each working day. CUSUM Use to store your choice of statistics in the worksheet. Standard deviations: Check to store the estimates of the standard deviations. and test for the presence of special causes. Check the statistics to store in the worksheet. then enter AtoBDist. one row for each plotted point. Choose Stat > Control Charts > Variables Charts for Subgroups > Xbar. Click the S Limits tab. one row for each plotted point. and enter the number of subgroups for each segment.

Attributes control charts plot count data. such as the number of defects or defective units. the results above may no * longer be correct. An I-MR-R/S chart consists of an I chart. and multivariate charts also plot measurement data. for subgroup data. X-bar − a chart of subgroup means R − a chart of subgroup ranges 2003 Minitab Inc. Test Failed at points: 5 * WARNING * If graph is updated with new data. Graph window output Interpreting the results Subgroup 5 failed Test 6. 4 out of 5 points more than 1 standard deviation from center line (on one side of CL). Choosing a variables control chart for subgroups Minitab offers five variables control charts: • • • • • X-bar and R − an X-bar chart and R chart displayed in one window X-bar and S − an X-bar chart and S chart displayed in one window I-MR-R/S (Between/Within) − a three-way control chart that uses both between-subgroup and within-subgroup variation. meaning it is the fourth point in a row in Zone B (1 to 2 standard deviations from the center line). such as length or pressure. time-weighted charts. This suggests the presence of special causes. For more information about control charts. and a R or S chart. 63 . Variables control charts for individuals. a MR chart.Control Charts Session window output Xbar Chart of AtoBDist Test Results for Xbar Chart of AtoBDist TEST 6. R Chart Variables Control Charts for Subgroups Overview data Variables control charts for subgroups plot statistics from measurement data. see Control Charts Overview.

Subgroups do not need to be the same size. R Chart Stat > Control Charts > Variables Charts for Subgroups > R A control chart of subgroup ranges.Quality Control • • S − a chart of subgroup standard deviations Zone − a chart of the cumulative scores based on each point's distance from the center line The I-MR-R/S (Between/Within) chart requires that you have two or more observations in at least one subgroup. then enter the columns. If the subgroup sizes do not vary much. see Control Charts Overview and Variables Control Charts for Subgroups Overview. each control limit is not a single straight line but varies with the subgroup size. Note that the first five observations in the left data set (subgroup 1) are the first row of the right-side data set. structured both ways. For more information. By default. while S charts are used for larger samples. Multivariate data must be entered down columns. 64 2003 Minitab Inc. and so on. then enter the columns. . Variables charts include: • • • • Variables charts for subgroups Variables charts for individuals Time-weighted charts Multivariate charts Structure your data for these charts using the guidelines below. the second 5 observations are the second row. <Scale> <Labels> <Multiple Graphs> <Data Options> <R Options> Data − Variables Control Chart Organize the data for all variables control charts in the same way. an R chart bases the estimate of the process variation. R charts are typically used to track process variation for samples of size 8 or less. you may want to force the control limits to be constant by specifying a fixed subgroup size using R Options > Estimate. Here is the same data set. with one column for each variable. on the average of the subgroup ranges. You can use R charts to track process variation and detect the presence of special causes. Subgroups are equal size Univariate (one variable) Multivariate (more than one variable) Down columns or across rows Down columns Subgroups are unequal size Down columns with subgroup indicator column Down columns with subgroup indicator column Structure subgroup data down a column or across rows. with subgroups of size 5. Minitab calculates summary statistics for each subgroup. or enter a historical value for s. σ. Observations for a subgroup are in one row of columns: Choose if subgroups are arranged in rows across several columns. Worksheet Structure Structure your data down a column or across rows. Dialog box items All observations for a chart are in one column: Choose if data are in one or more columns. If the subgroups are not equal. using the following table as a guide. which are plotted on the chart and used to estimate process parameters. Subgroup sizes: Enter a number or a column of subscripts. You can also use a pooled standard deviation.

then click OK.Control Charts When subgroups are of unequal size. Missing data See Missing data in control charts for information on how to handle missing data for different types of control charts. then enter a series of columns. In this example. see Box-Cox Transformation. use any dialog box options. Then. If the data are highly skewed. Do one of the following: • If subgroups are in one or more columns. you must enter data that approximate a normal distribution. then create a second column of subscripts which serve as subgroup indicators. In Subgroup sizes. If you like. Nonnormal data To properly interpret Minitab's control charts. 3 2003 Minitab Inc. For information on the stand alone Box-Cox transformation command. and so on. enter a number or a column of subscripts. C1 contains the process data and C2 contains subgroup indicators: Each time a subscript changes in C2. choose All observations for a chart are in one column. 65 . To display an R chart 1 2 Choose Stat > Control Charts > Variables Charts for Subgroups > R. you may want to use the Box-Cox transformation to induce normality. For information on the Box-Cox transformation option. then enter one or more columns. subgroup 1 has three observations. or by using the stand alone Box-Cox command. you can use the transformation option to transform the data at the same time you draw the control chart. you must enter your data in one column. In the following example. • If subgroups are in rows. Use the stand alone command as an exploratory tool to help you determine the best lambda value for the transformation. subgroup 2 has six observations. choose Observations for a subgroup are in one row of columns. see Options − Box-Cox. You can access the Box-Cox transformation two ways: by using the Box-Cox transformation option provided with the control chart commands. a new subgroup begins in C1.

1 2 In the control chart dialog box. You can set preferences for the estimation of σ using Tools > Options > Control Charts and Quality Tools > Estimating Standard Deviation. A list of values.. not individual observations. A list of values. For example. Subgroup size > 1 Rbar: Choose to estimate σ based on the average of the subgroup ranges. You can enter historical data in several ways: Options − Estimate . enter one or more values. But you may want to use or omit certain data if they show abnormal behavior. To use historical standard deviation values 1 2 3 • • • In the control chart dialog box. Moving Range. then enter individual subgroup numbers or a range of subgroups. To omit one observation from a subgroup. you can prevent these subgroups from influencing the estimation. one for each stage within each variable. you may enter this value or values. not just the subsetted data. which applies to all variables and stages. if some subgroups have erratic data due to assignable causes that you have corrected. Dialog box items Omit the following samples when estimating parameters: Choose to specify subgroups of data to omit from the computations for µ and σ. > control chart Options > Parameters Charts R. > control chart Options > Estimate Charts X-bar and R. CUSUM Enters historical data for estimating σ. if you have a known process parameter or an estimate obtained from past data.Quality Control Options − Parameters . R Omits or includes certain subgroups to estimate µ and σ. In Standard deviation. Enter entire subgroups. All stages for each variable use the same value. Click the Estimate tab. . You can also select one of two methods to estimate σ and use biased estimates of σ. then decide to omit subgroup 2 from the parameter estimation (using control chart Options > Estimate). When you include or exclude rows using control chart Options > Estimate and choose to perform a Box-Cox transformation. one for each variable. Do one of the following: 66 2003 Minitab Inc.. Each stage for each variable uses a different value. change that value to missing (∗). you must enter subgroup 2 as the subgroup to omit. Minitab estimates it from the data. click control chart Options. Because Minitab considers an omitted sample (subgroup 1) as part of the entire data set. For example. click control chart Options. To omit or use subgroups for the estimates of parameters By default. A single value. For example.. You can omit or include certain subgroups when estimating parameters using Estimate. Dialog box items Standard deviation: Enter the historical standard deviations as constants. the omitted or used samples apply to the entire data set. Caution When you include or exclude rows using control chart > Data Options > Subset and also omit or use samples to estimate parameters. Pooled standard deviation: Choose to estimate σ using a pooled standard deviation. Method for estimating standard deviation: See [2] for a discussion of the relative merits of the methods listed below. S. Minitab only uses the non-omitted data to find lambda. then enter individual subgroup numbers or a range of subgroups. Use the following samples when estimating parameters: Choose to specify subgroups of data to compute µ and σ. you exclude subgroup 1 from your analysis (using control chart > Data Options > Subset). Use unbiasing constant: Uncheck to use biased estimates of σ. Minitab estimates the process parameters from all the data.. If you do not specify a value for σ. Click OK.

choose a method. Each value represents the number of standard deviations below and above the mean. U. To customize the control limits 1 2 In the control chart dialog box. a horizontal line labeled LB is drawn at the lower bound instead. Click OK. Note S Limits . Moving Average. S. the values are interpreted as observation numbers. and below the mean. except with the Individuals. Dialog box items Display control limits at These multiples of the standard deviation: Enter one or more values. Under Estimating standard deviation. Upper standard deviation limit bound: Check to set an upper bound for the control limits. Tip You can also modify the control limits using Minitab's graph editing features. If you specify more than one value. the plotted data are not changed. 67 . then type the subgroup size. if the calculated lower control limit is less than the lower bound. Click OK in each dialog box. a horizontal line labeled UB is drawn at the upper bound instead. P. see Example of an X-bar chart with tests and customized control limits. If the calculated lower control limit is less than the lower bound. When subgroup sizes are unequal. you may want to force the control limits to be constant by entering a sample size in this text box. Click the Estimate tab. above.. click control chart Options. Moving Range. Assuming all subgroups have size: Choose to estimate the control limits using a specified subgroup size. Click the S Limits tab. You can also force the control limits to be constant when sample sizes are unequal. EWMA Allows you to draw control limits above and below the mean at the multiples of any standard deviation. Similarly. With these charts.Control Charts • • Note 3 Choose Omit the following samples when estimating parameters and enter the subgroups or observation numbers you want to omit from the calculations. When subgroup sizes are not equal. Each value is the number of standard deviations the lines should be drawn at. Minitab excludes any moving ranges with omitted data from the calculations. To choose how the standard deviation is estimated 1 2 In the control chart dialog box. You can set bounds on the upper and/or lower control limits. a horizontal line labeled UB will be drawn at the upper bound instead. When the calculated upper control limit is greater than the upper bound. Place bounds on control limits Lower standard deviation limit bound: Check to set a lower bound for the control limits. each control limit is not a single straight line. click control chart Options. Then click OK. and I-MR charts. • To set bounds on the control limits: check Lower standard deviation limit bound (and/or Upper standard deviation limit bound) and enter a value. calculate control limits Using actual sizes of the subgroups: Choose to use the subgroups defined in the main dialog box to estimate σ. For an example. a horizontal line labeled LB will be drawn at the lower bound instead. Choose Use the following samples when estimating parameters and enter the subgroups or observation numbers you want to use for the calculations. enter one or more values in These multiples of the standard deviation. If the calculated upper control limit is greater than the upper bound. > control chart Options > S Limits Charts X-bar.. 3 2003 Minitab Inc. If the sizes do not vary much. but varies with the subgroup size. NP. Minitab assumes the values you enter are subgroup numbers. Only the control limits are affected. When you use Omit the following samples when estimating parameters with Length of moving range. R. Do any of the following: • To specify where control limits are drawn: under Display control limits at. a pair of control limits is drawn for each value in the list. The values entered are the number of standard deviations above and below the center line.

then enter a value. all increasing or all decreasing Test 4 14 points in a row.Quality Control To force control limits and center line to be constant 1 2 In the control chart dialog box. alternating up and down Dialog box items Perform all tests for special causes: Choose to perform all tests. When a point fails a test. suppose the largest subgroup is size 125 and the smallest is 100. The test results are displayed in the Session window. See Tools > Options > Control Charts and Quality Tools > Other. all increasing or all decreasing 14 points in a row. Click OK. alternating up and down Perform no tests: Choose to perform no tests for special causes. Moving Range. detects a specific pattern in the plotted data. U. . Test 1 1 point more than 3σ from center line Test 2 9 points in a row on same side of center line Test 3 6 points in a row. the test results are not updated in the Session window.. For example. 1 point > 3 standard deviations from center line 9 points in a row on same side of center line 6 points in a row. 3 Note Options − Tests . You can use this method because the size difference is 25% (125 / 100 = 25%). You should force the control limits and center line to be constant only when the difference in size between the largest and smallest subgroups is no more than 25%. one that should be investigated. calculate control limits. Click the S Limits tab. Minitab marks it by the lowest numbered test. S.. If a point fails more than one test. Minitab will perform tests when sample sizes are unequal. The occurrence of a pattern suggests a special cause for the variation. C With any of the attributes control charts. P. you can perform the 4 tests for special causes. Minitab marks it with the test number on the plot. Under When subgroup sizes are unequal. Perform the following tests for special causes: Choose to specify a subset of the four tests. See Updating Graphs for more information. Set your preferences for the sensitivity of tests with Tools > Options > Control Charts and Quality Tools > Define Tests. choose Assuming all subgroups have size. enter a value of 6 to calculate the control limits and center line as if all subgroup sizes were 6. You can choose to change the color of a control chart when the last row of data causes a new test failure for any point. click control chart Options. 68 2003 Minitab Inc. then check any combination of the tests listed below. If the data change and the control chart is live-linked. Each test. as shown below. > control chart Options > Tests Charts R. See [1] and [10] for guidance on using these tests. NP. For example.

The Options subdialog box lists the common transformations natural log (λ = 0) and square root (λ = 0. When you include or exclude rows using control chart Options > Estimate. Note With the following charts. You 2003 Minitab Inc. The indicators can be numbers. 3 Options − Stages . > control chart Options > Stages Charts All charts except Z-MR You can display stages in your process by drawing a "historical chart" − a control chart in which the control limits and center line are estimated independently for different groups in your data. If you like. then click OK. or text. Historical charts are particularly useful for comparing data before and after a process improvement. Minitab treats each repeated value as a separate occurrence. 5 Options − Box-Cox . enter the column which contains the stage indicators. unless λ = 0. you must set up a column of grouping indicators. choose Perform no tests. You can enter the same value more than once. X-bar and R. 69 . when subgroups are across rows). In most cases. Enclose date/time or text entries in double quotes. (λ is lambda. you should not choose a λ outside the range of −2 and 2. • To select certain tests. do one of the following: • To start a new stage each time the value in the column changes.. Minitab treats each repeated value as a separate occurrence. Enter the values. you can tell Minitab to start a new stage in one of two ways: • • 1 2 3 4 Each time the value in the column changes At the first occurrence of one or more values In the control chart dialog box. dates.) To use this option. you must have at least one subgroup with two or more observations: R. Dialog box items Define stages (historical groups) with this variable: Enter the column that contains the stage indicators. in which case the natural log is taken. use any dialog box options. Minitab only uses the non-omitted data to find lambda. then enter the values.. choose Perform the following tests for special causes and check the tests to perform. You can enter the same value more than once. In Define stages (historical groups) with this variable. Enclose date/time or text entries in double quotes.5). To display a historical chart To define stages in your process. choose With each new value. Click the Stages tab. click control chart Options. Click OK. > control chart Options > Box-Cox Charts All except the Attributes charts You can use the Box-Cox power transformation when your data are very skewed or when the within-subgroup variation is unstable to make the data "more normal. choose Perform all tests for special causes. • To perform no tests. When executing the command. You can also choose any value between − 5 and 5 for λ. and X-bar and S.Control Charts To perform the tests for special causes 1 2 Choose Stat > Control Charts > control chart type > control chart > control chart Options > Tests. The column must be the same length as the data column (or columns. Under When to start a new value. With the first occurrence of these values: Choose to start a new stage at the first occurrence of a certain value. the data must be positive.." The transformation takes the original data to the power λ. S.. • To start a new stage at the first occurrence of a certain value. Do one of the following: • To do all tests. When to start a new stage With each new value: Choose to start a new stage each time the value in the column changes. choose With the first occurrence of these values.

then do one of the following: • Choose Lambda = 0 (natural log) to use the natural log of the data.5 (square root) to use the square root of the data. and enter the number of subgroups for each segment. Each segment corresponds to a stage (if chart has stages): Choose to split the control chart by stage or stage when you have specified stages in control chart Options > Stage. Click the Box-Cox tab. click control chart Options.5 (square root): Choose to use the square root of the data. 70 2003 Minitab Inc. • Choose Optimal lambda for each stage (in a chart with stages) to have Minitab search for optimal values for each stage. Options − Display . and enter the number of subgroups. Lambda=0. • Choose Lambda=0. then enter one or more values. • Choose Other (enter value(s) between -5 and 5) to transform the data using another lambda value or values. Dialog box items Box-Cox power transformation (W = Y**Lambda): Choose to use when your data are very skewed or when the withinsubgroup variation is unstable... Use overall standard deviation: Check to use the overall standard deviation in the estimation of the optimal lambda. Click the Display tab. Subgroups to display Display all subgroups: Choose to display all subgroups. . Optimal lambda: Choose to have Minitab search for an optimal value. When you use this transformation. Lambda = 0 (natural log): Choose to use the natural log of the data. doing so will double-transform the data. Split chart into series of segments for display purposes Do not split: Choose not to split the chart. Other (enter value(s) between -5 and 5): Choose to transform the data using another lambda value or values. Minitab does not accept any values you enter in the Parameters tab for historical means or historical standard deviations. For help choosing a lambda value. click control chart Options. Optimal lambda for each stage (in a chart with stages): Choose to have Minitab search for optimal values for each stage.Quality Control may want to first run the command Stat > Control Charts > Box-Cox Transformation to help you find the optimal transformation value. do not select the Box-Cox option when you make a control char. Caution If you use Stat > Control Charts > Box-Cox Transformation to find the optimal λ value and choose to store the transformed data with that command. Each segment contains __ subgroups: Choose to split the control chart into subgroups of a specified number of points. • Choose Optimal lambda to have Minitab search for an optimal value for all stages. > control chart Options > Display Charts All charts except Z-MR Use to display the control chart by stage or by number of plotted points and to display the test results in the Session window. To do the Box-Cox power transformation 1 2 In the control chart dialog box. Check Box-Cox power transformation (W = Y**Lambda). 3 Click OK. see the independent Box-Cox transformation command Stat > Control Charts > Box-Cox Transformation. Display last __ subgroups: Choose to display a specific number of subgroups. Test results Display test results in Session window: Check to display the test results in the Session window. To display subgroups 1 In the control chart dialog box. then enter the lambda values.

MTW. Stage: Check to store the stage. Click OK. then enter AtoBDist. choose Each segment corresponds to a stage (if chart has stages). Click OK in each dialog box. click control chart Options. Click OK. add two rows of tick labels (Month and Day) below the original tick labels (the subgroup number). • To display a specific number of the last subgroups. then click OK. one row for each stage. 71 . choose Each segment contains __ subgroups. 3 Options − Storage Stat > Control Charts > control chart type > control chart > control chart Options > Storage Charts R. one row for each plotted point. Choose All observations for a chart are in one column. one row for each plotted point. Do one of the following: • To split the chart into a specific number of subgroups. Control limit values: Check to store the control limit values. S Use to store your choice of statistics in the worksheet. Example of R chart with time stamp In this example. one row for each plotted point. under Subgroups to display. Enter Month Day. Choose Stat > Control Charts > Variables Charts for Subgroups > R. 3 To split a control chart 1 2 In the control chart dialog box. Subgroup size: Check to store the subgroup sizes. one row for each plotted point. and enter the number of subgroups for each segment. Click Scale. To store statistics in the worksheet 1 2 Choose Stat > Control Charts > control chart type > control chart > control chart Options > Storage. In Subgroup sizes. Click the Display tab. under Subgroups to display. • To split the chart by stage. 2003 Minitab Inc. one row for each plotted point. Minitab stores one column for the lower control limits and one column for the upper control limits. Under X Scale. one row for each plotted point. Stores these values for each point Point plotted: Check to store the plotted points. 1 2 3 4 5 6 Open the worksheet CRANKSHD. Check the statistics to store in the worksheet. choose Stamp. choose Display all subgroups. enter 5. Test results: Check to store the results of any test performed. Dialog box items Store these estimates for each chart Standard deviations: Check to store the estimates of the standard deviations. choose Display last __ subgroups and enter the number of subgroups to display. Center line value: Check to store the center line value.Control Charts 2 Do one of the following: • To display all subgroups in the control chart.

choose Rbar. Click R Options. AtoBDist is the distance (in mm) from the actual (A) position of a point on the crankshaft to the baseline (B) position. Choose Stat > Control Charts >Variables Charts for Subgroups > R. Under Method for estimating standard deviation.Quality Control Graph window output Interpreting the results The points are randomly distributed between the control limits. Choose Perform all tests for special causes. To ensure production quality. Choose All observations for a chart are in one column. and then ten per day from the 18th through the 25th. Click OK in each dialog box. . parts of the crankshaft move up and down a certain distance from an ideal baseline position. implying a stable process. Now you want to display an R chart to track the process variation using the same data. Subgroup size > 1. implying a stable process. 1 2 3 4 5 6 7 8 Open the worksheet CRANKSH. Example of R chart Suppose you work at a car assembly plant in a department that assembles engines. These do not − again. In Subgroup sizes. It is also important to compare points on the R chart with those on the X-bar chart for the same data − see Example of an X-bar chart with tests and customized control limits to see if the points follow each other. from September 28 through October 15. then click the Estimate tab. enter 5. Click the Tests tab. 72 2003 Minitab Inc. You have already displayed an X-bar chart with the data to track the process level and test for special causes. In an operating engine.MTW. you took five measurements each working day. then enter AtoBDist.

S charts are typically used to track process variation for samples of 9 or more. Minitab calculates summary statistics for each subgroup. 2003 Minitab Inc. such as the number of defects or defective units. For more information about control charts. while R charts are used for smaller samples. X-bar − a chart of subgroup means R − a chart of subgroup ranges S − a chart of subgroup standard deviations Zone − a chart of the cumulative scores based on each point's distance from the center line The I-MR-R/S (Between/Within) chart requires that you have two or more observations in at least one subgroup. Attributes control charts plot count data. a MR chart. such as length or pressure. Variables control charts for individuals. Subgroups do not need to be the same size. time-weighted charts. Choosing a variables control chart for subgroups Minitab offers five variables control charts: • • • • • • • X-bar and R − an X-bar chart and R chart displayed in one window X-bar and S − an X-bar chart and S chart displayed in one window I-MR-R/S (Between/Within) − a three-way control chart that uses both between-subgroup and within-subgroup variation. It is also important to compare points on the R chart with those on the X-bar chart for the same data − see Example of an X-bar chart with tests and customized control limits to see if the points follow each other. 73 . You can use S charts to track the process variation and detect the presence of special causes. and multivariate charts also plot measurement data. and a R or S chart.Control Charts Graph window output Interpreting the results The points are randomly distributed between the control limits. S Chart Stat > Control Charts > Variables Charts for Subgroups > S A control chart of subgroup standard deviations. implying a stable process. S Chart Variables Control Charts for Subgroups Overview data Variables control charts for subgroups plot statistics from measurement data. for subgroup data. which are plotted on the chart and used to estimate process parameters. implying a stable process. see Control Charts Overview. These do not − again. An I-MR-R/S chart consists of an I chart.

Observations for a subgroup are in one row of columns: Choose if subgroups are arranged in rows across several columns. then enter the columns. Dialog box items All observations for a chart are in one column: Choose if data are in one or more columns. .Quality Control By default. see Control Charts Overview and Variables Control Charts for Subgroups Overview. If the subgroups are not equal. Multivariate data must be entered down columns. with one column for each variable. For more information. σ. You can also use a pooled standard deviation. you may want to force the control limits to be constant by specifying a fixed subgroup size using S Options > Estimate. on the average of the subgroup standard deviations. Here is the same data set. Note that the first five observations in the left data set (subgroup 1) are the first row of the right-side data set. <Scale> <Labels> <Multiple Graphs> <Data Options> <S Options> Data − Variables Control Chart Organize the data for all variables control charts in the same way. Worksheet Structure Structure your data down a column or across rows. and so on. Variables charts include: • • • • Variables charts for subgroups Variables charts for individuals Time-weighted charts Multivariate charts Structure your data for these charts using the guidelines below. structured both ways. If the subgroup sizes do not vary much. or enter a historical value for σ. each control limit is not a single straight line but varies with the subgroup size. with subgroups of size 5. 74 2003 Minitab Inc. the second 5 observations are the second row. Subgroup sizes: Enter a number or a column of subscripts. Subgroups are equal size Univariate (one variable) Multivariate (more than one variable) Down columns or across rows Down columns Subgroups are unequal size Down columns with subgroup indicator column Down columns with subgroup indicator column Structure subgroup data down a column or across rows. Minitab's S Chart command bases the estimate of the process variation. then enter the columns. using the following table as a guide.

Use the stand alone command as an exploratory tool to help you determine the best lambda value for the transformation. Missing data See Missing data in control charts for information on how to handle missing data for different types of control charts. S. In the following example. which applies to all variables and stages. To display an S chart 1 2 Choose Stat > Control Charts > Variables Charts for Subgroups > S. > control chart Options > Parameters Charts R. In Subgroup sizes. if you have a known process parameter or an estimate obtained from past data. In Standard deviation. see Options − Box-Cox. Then. For example. Dialog box items Standard deviation: Enter the historical standard deviations as constants. use any dialog box options. then enter a series of columns. you can use the transformation option to transform the data at the same time you draw the control chart. 3 Options − Parameters . you must enter data that approximate a normal distribution. then create a second column of subscripts which serve as subgroup indicators. Do one of the following: • If subgroups are in one or more columns. 75 . then enter one or more columns. a new subgroup begins in C1. A single value. Moving Range. you must enter your data in one column.. To use historical standard deviation values 1 2 3 • In the control chart dialog box. C1 contains the process data and C2 contains subgroup indicators: Each time a subscript changes in C2. and so on. You can access the Box-Cox transformation two ways: by using the Box-Cox transformation option provided with the control chart commands. Click OK. For information on the stand alone Box-Cox transformation command.Control Charts When subgroups are of unequal size. You can omit or include certain subgroups when estimating parameters using Estimate. choose Observations for a subgroup are in one row of columns. click control chart Options.. enter one or more values. you may enter this value or values. In this example. If the data are highly skewed. or by using the stand alone Box-Cox command. If you do not specify a value for σ. choose All observations for a chart are in one column. For information on the Box-Cox transformation option. CUSUM Enters historical data for estimating σ. You can enter historical data in several ways: 2003 Minitab Inc. you may want to use the Box-Cox transformation to induce normality. Minitab estimates it from the data. see Box-Cox Transformation. • If subgroups are in rows. If you like. then click OK. subgroup 2 has six observations. enter a number or a column of subscripts. Nonnormal data To properly interpret Minitab's control charts. subgroup 1 has three observations.

then enter individual subgroup numbers or ranges of subgroups. To omit one observation from a subgroup. the values are interpreted as observation numbers. one for each stage within each variable. Use the following samples when estimating parameters: Choose to specify subgroups of data to compute µ and σ. All stages for each variable use the same value. Minitab only uses the non-omitted data to find lambda. Note 76 2003 Minitab Inc. Because Minitab considers an omitted sample (subgroup 1) as part of the entire data set. choose a method. Do one of the following: • Choose Omit the following samples when estimating parameters and enter the subgroups or observation numbers you want to omit from the calculations. not individual observations. Pooled standard deviation: Choose to estimate σ using a pooled standard deviation.. Note 3 Click OK. Click the Estimate tab. and I-MR charts. With these charts. click control chart Options. Minitab assumes the values you enter are subgroup numbers. Subgroup size > 1 Sbar: Choose to estimate σ based on the average of the subgroup standard deviations. You can set preferences for the estimation of σ using Tools > Options > Control Charts and Quality Tools > Estimating Standard Deviation. Dialog box items Omit the following samples when estimating parameters: Choose to specify subgroups of the data to omit from the computations for µ and σ. change that value to missing (∗) in the worksheet. > control chart Options > Estimate Charts X-bar and S. To omit or use subgroups for the estimates of parameters By default. Moving Range. Use unbiasing constant: Uncheck to use biased estimates of σ. except with the Individuals. . Caution When you include or exclude rows using control chart > Data Options > Subset and also omit or use samples to estimate parameters. A list of values. When you use Omit the following samples when estimating parameters with Length of moving range. When you include or exclude rows using control chart Options > Estimate and choose to perform a Box-Cox transformation. click control chart Options.. if some subgroups have erratic data due to assignable causes that you have corrected. S Omits or includes certain subgroups to estimate µ and σ. Enter entire subgroups. you must enter subgroup 2 as the subgroup to omit. But you may want to use or omit certain data if they show abnormal behavior.Quality Control • • A list of values. you exclude subgroup 1 from your analysis (using control chart > Data Options > Subset). Minitab estimates the process parameters from all the data. then enter individual subgroup numbers or a range of subgroups. you can prevent these subgroups from influencing the estimation of process parameters. Under Estimating standard deviation. For example. You can also select one of two methods to estimate σ and use biased estimates of σ. not just the subsetted data. Minitab excludes any moving ranges with omitted data from the calculations. Options − Estimate . Method for estimating standard deviation: See [2] for a discussion of the relative merits of the methods listed below. then decide to omit subgroup 2 from the parameter estimation (using control chart Options > Estimate). For example. To choose how the standard deviation is estimated 1 2 In the control chart dialog box. Each stage for each variable uses a different value. Click the Estimate tab. Then click OK. • Choose Use the following samples when estimating parameters and enter the subgroups or observation numbers you want to use for the calculations. 1 2 In the control chart dialog box. the omitted or used samples apply to the entire data set. one for each variable.

Place bounds on control limits Lower standard deviation limit bound: Check to set a lower bound for the control limits. If the calculated lower control limit is less than the lower bound. a horizontal line labeled UB will be drawn at the upper bound instead. see Example of an X-bar chart with tests and customized control limits. Each test. You can also force the control limits to be constant when sample sizes are unequal. Only the control limits are affected. each control limit is not a single straight line. detects a specific pattern in the plotted data. Moving Range. Dialog box items Display control limits at These multiples of the standard deviation: Enter one or more values. calculate control limits Using actual sizes of the subgroups: Choose to use the subgroups defined in the main dialog box to estimate σ. 77 . enter one or more values in These multiples of the standard deviation. one that should be investigated. then enter a value. as shown below. When subgroup sizes are unequal. If the sizes do not vary much.. 3 Note Options − Tests . If you specify more than one value. Each value is the number of standard deviations the lines should be drawn at. but varies with the subgroup size. See [1] and [10] for guidance on using these tests. 3 To force control limits and center line to be constant 1 2 In the control chart dialog box. S.. Each value represents the number of standard deviations below and above the mean. Click the S Limits tab. you can perform the 4 tests for special causes. if the calculated lower control limit is less than the lower bound. enter a value of 6 to calculate the control limits and center line as if all subgroup sizes were 6. R. a horizontal line labeled LB is drawn at the lower bound instead. Under When subgroup sizes are unequal. Do any of the following: • To specify where control limits are drawn: under Display control limits at. EWMA Allows you to draw control limits above and below the mean at the multiples of any standard deviation. You can use this method because the size difference is 25% (125 / 100 = 25%). calculate control limits.. choose Assuming all subgroups have size. Similarly. For example.. Click OK in each dialog box. suppose the largest subgroup is size 125 and the smallest is 100. For example. To customize the control limits 1 2 In the control chart dialog box. you may want to force the control limits to be constant by entering a sample size in this text box. Upper standard deviation limit bound: Check to set an upper bound for the control limits. and below the mean. U. a horizontal line labeled LB will be drawn at the lower bound instead. When subgroup sizes are not equal. C With any of the attributes control charts. > control chart Options > S Limits Charts X-bar. Click the S Limits tab. The values entered are the number of standard deviations above and below the center line. P. Click OK. P. Assuming all subgroups have size: Choose to estimate the control limits using a specified subgroup size. click control chart Options. You should force the control limits and center line to be constant only when the difference in size between the largest and smallest subgroups is no more than 25%. • To set bounds on the control limits: check Lower standard deviation limit bound (and/or Upper standard deviation limit bound) and enter a value. a horizontal line labeled UB is drawn at the upper bound instead. then type the subgroup size. Tip You can also modify the control limits using Minitab's graph editing features.Control Charts S Limits . 2003 Minitab Inc. If the calculated upper control limit is greater than the upper bound. You can set bounds on the upper and/or lower control limits. When the calculated upper control limit is greater than the upper bound. NP. S. above. click control chart Options. a pair of control limits is drawn for each value in the list. NP. Moving Average. The occurrence of a pattern suggests a special cause for the variation. the plotted data are not changed. > control chart Options > Tests Charts R. For an example. U.

Set your preferences for the sensitivity of tests with Tools > Options > Control Charts and Quality Tools > Define Tests. You can choose to change the color of a control chart when the last row of data causes a new test failure for any point. • To select certain tests.. To perform the tests for special causes 1 2 Choose Stat > Control Charts > control chart type > control chart > control chart Options > Tests. choose Perform all tests for special causes. Click OK. If a point fails more than one test. 78 2003 Minitab Inc. • To perform no tests. alternating up and down Dialog box items Perform all tests for special causes: Choose to perform all tests. then check any combination of the tests listed below. Do one of the following: • To do all tests. alternating up and down Perform no tests: Choose to perform no tests for special causes. choose Perform the following tests for special causes and check the tests to perform. Minitab marks it with the test number on the plot. Minitab will perform tests when sample sizes are unequal. choose Perform no tests. Test 1 1 point more than 3σ from center line Test 2 9 points in a row on same side of center line Test 3 6 points in a row. > control chart Options > Stages Charts All charts except Z-MR You can display stages in your process by drawing a "historical chart" − a control chart in which the control limits and center line are estimated independently for different groups in your data. 3 Options − Stages .Quality Control When a point fails a test. the test results are not updated in the Session window. Historical charts are particularly useful for comparing data before and after a process improvement. . all increasing or all decreasing Test 4 14 points in a row. See Updating Graphs for more information. Perform the following tests for special causes: Choose to specify a subset of the four tests. all increasing or all decreasing 14 points in a row. The test results are displayed in the Session window.. If the data change and the control chart is live-linked. Minitab marks it by the lowest numbered test. 1 point > 3 standard deviations from center line 9 points in a row on same side of center line 6 points in a row. See Tools > Options > Control Charts and Quality Tools > Other.

when subgroups are across rows). Minitab does not accept any values you enter in the Parameters tab for historical means or historical standard deviations. The Options subdialog box lists the common transformations natural log (λ = 0) and square root (λ = 0. do one of the following: • To start a new stage each time the value in the column changes. then enter the values. You can also choose any value between − 5 and 5 for λ. S. X-bar and R. To display a historical chart To define stages in your process. You can enter the same value more than once. choose With each new value. When executing the command. (λ is lambda. doing so will double-transform the data. In Define stages (historical groups) with this variable. enter the column which contains the stage indicators. Under When to start a new value. Dialog box items Define stages (historical groups) with this variable: Enter the column that contains the stage indicators. you must set up a column of grouping indicators. choose With the first occurrence of these values. Use overall standard deviation: Check to use the overall standard deviation in the estimation of the optimal lambda. then click OK..) To use this option. When you include or exclude rows using control chart Options > Estimate. 2003 Minitab Inc. If you like. do not select the Box-Cox option when you make a control char. Caution If you use Stat > Control Charts > Box-Cox Transformation to find the optimal λ value and choose to store the transformed data with that command. Enclose date/time or text entries in double quotes. Optimal lambda for each stage (in a chart with stages): Choose to have Minitab search for optimal values for each stage. Enter the values.5 (square root): Choose to use the square root of the data. With the first occurrence of these values: Choose to start a new stage at the first occurrence of a certain value. Minitab treats each repeated value as a separate occurrence. or text. The indicators can be numbers. in which case the natural log is taken. When you use this transformation. Click the Stages tab. Lambda = 0 (natural log): Choose to use the natural log of the data. When to start a new stage With each new value: Choose to start a new stage each time the value in the column changes.. use any dialog box options. click control chart Options.5). Dialog box items Box-Cox power transformation (W = Y**Lambda): Choose to use when your data are very skewed or when the withinsubgroup variation is unstable. You may want to first run the command Stat > Control Charts > Box-Cox Transformation to help you find the optimal transformation value. you should not choose a λ outside the range of −2 and 2. the data must be positive. 79 . unless λ = 0. Minitab treats each repeated value as a separate occurrence." The transformation takes the original data to the power λ. • To start a new stage at the first occurrence of a certain value. you must have at least one subgroup with two or more observations: R. In most cases. dates. You can enter the same value more than once. Minitab only uses the non-omitted data to find lambda. > control chart Options > Box-Cox Charts All except the Attributes charts You can use the Box-Cox power transformation when your data are very skewed or when the within-subgroup variation is unstable to make the data "more normal. and X-bar and S. Enclose date/time or text entries in double quotes. Lambda=0.Control Charts Note With the following charts. you can tell Minitab to start a new stage in one of two ways: • • 1 2 3 4 Each time the value in the column changes At the first occurrence of one or more values In the control chart dialog box. Optimal lambda: Choose to have Minitab search for an optimal value. 5 Options − Box-Cox . The column must be the same length as the data column (or columns.

3 To split a control chart 1 2 In the control chart dialog box. choose Display all subgroups. • Choose Lambda=0. • To display a specific number of the last subgroups. Test results Display test results in Session window: Check to display the test results in the Session window. choose Each segment contains __ subgroups. Click the Box-Cox tab. To do the Box-Cox power transformation 1 2 In the control chart dialog box. choose Each segment corresponds to a stage (if chart has stages). Check Box-Cox power transformation (W = Y**Lambda). • To split the chart by stage. 3 80 2003 Minitab Inc.Quality Control Other (enter value(s) between -5 and 5): Choose to transform the data using another lambda value or values. under Subgroups to display. • Choose Optimal lambda to have Minitab search for an optimal value for all stages. • Choose Other (enter value(s) between -5 and 5) to transform the data using another lambda value or values. 3 Click OK.. then enter one or more values. Click OK. Click OK. Click the Display tab. Options − Display . under Subgroups to display. click control chart Options. Each segment contains __ subgroups: Choose to split the control chart into subgroups of a specified number of points. • Choose Optimal lambda for each stage (in a chart with stages) to have Minitab search for optimal values for each stage. see the independent Box-Cox transformation command Stat > Control Charts > Box-Cox Transformation. and enter the number of subgroups for each segment. choose Display last __ subgroups and enter the number of subgroups to display. To display subgroups 1 2 In the control chart dialog box. and enter the number of subgroups for each segment. Split chart into series of segments for display purposes Do not split: Choose not to split the chart. click control chart Options. then enter the lambda values.. Click the Display tab. . Do one of the following: • To display all subgroups in the control chart. Display last __ subgroups: Choose to display a specific number of subgroups. For help choosing a lambda value. Subgroups to display Display all subgroups: Choose to display all subgroups. click control chart Options. Each segment corresponds to a stage (if chart has stages): Choose to split the control chart by stage or stage when you have specified stages in control chart Options > Stage. then do one of the following: • Choose Lambda = 0 (natural log) to use the natural log of the data. and enter the number of subgroups. Do one of the following: • To split the chart into a specific number of subgroups. > control chart Options > Display Charts All charts except Z-MR Use to display the control chart by stage or by number of plotted points and to display the test results in the Session window.5 (square root) to use the square root of the data.

In Subgroup sizes. Choose Stat > Control Charts > Variables Charts for Subgroups > S. 1 2 3 4 5 Open the file EXH_QC. one row for each plotted point. then click OK. one row for each plotted point.Control Charts Options − Storage Stat > Control Charts > control chart type > control chart > control chart Options > Storage Charts R. one row for each plotted point. Click OK in each dialog box. Choose All observations for a chart are in one column. To store statistics in the worksheet 1 2 Choose Stat > Control Charts > control chart type > control chart > control chart Options > Storage. Stage: Check to store the stage. Stores these values for each point Point plotted: Check to store the plotted points. Control limit values: Check to store the control limit values. Example of S chart Suppose you want to display an S chart to monitor the variability of faults during three shifts over the course of a week. 81 . Test results: Check to store the results of any test performed.MTW. Graph window output 2003 Minitab Inc. S Use to store your choice of statistics in the worksheet. Center line value: Check to store the center line value. You want to estimate σ using a pooled standard deviation. Dialog box items Store these estimates for each chart Standard deviations: Check to store the estimates of the standard deviations. one row for each plotted point. The subgroup sizes in your data are unequal because the number of measures taken during the shifts varies. Minitab stores one column for the lower control limits and one column for the upper control limits. one row for each plotted point. Check the statistics to store in the worksheet. Subgroup size: Check to store the subgroup sizes. one row for each plotted point. one row for each stage. then enter Faults. enter Shift.

σ. For more information. Thus. Zone Chart Variables Control Charts for Subgroups Overview data Variables control charts for subgroups plot statistics from measurement data. Dialog box items All observations for a chart are in one column: Choose if data are in one or more columns. 82 2003 Minitab Inc. The moving range is of length 2. An I-MR-R/S chart consists of an I chart. If the subgroup sizes do not vary much. by default. then plotted on the chart. and Comparing a Zone Chart with a Shewart Chart. Then you may want to change the weights when the process is in control and it is more important to detect drift. such as length or pressure. Variables control charts for individuals. Subgroup sizes: Enter a number or a column of subscripts. You can modify the zone chart weighting scheme to provide the sensitivity needed for a specific process. for subgroup data. since consecutive values have the greatest chance of being alike. the process standard deviation. based on "zones" at 1. and 3 standard deviations from the center line. each control limit is not a single straight line but varies with the subgroup size. You can also estimate σ using the median of the moving range. X-bar − a chart of subgroup means R − a chart of subgroup ranges S − a chart of subgroup standard deviations Zone − a chart of the cumulative scores based on each point's distance from the center line The I-MR-R/S (Between/Within) chart requires that you have two or more observations in at least one subgroup. you may want one set of weights during the initial SPC phase for detecting special causes. By default. which are plotted on the chart and used to estimate process parameters. To limit the number of plotted points. the mean of the observations in each subgroup is calculated. With data in subgroups. is estimated using a pooled standard deviation. This method is equivalent to four of the standard tests for special causes in an X-bar or Individuals chart. time-weighted charts. change the length of the moving range. then enter the columns. σ is estimated using MRbar / d2. Subgroups do not need to be the same size. if its score is greater than or equal to 8. and a R or S chart. You can also base the estimate on the average of subgroup ranges or subgroup standard deviations. you do not need to recognize the patterns associated with non-random behavior as on a Shewhart chart. and multivariate charts also plot measurement data. see Control Charts Overview. a MR chart. you may want to force the control limits to be constant by specifying a fixed subgroup size using Zone Options > Estimate. or enter a historical value for σ. For more information about control charts. Minitab generates a standardized zone chart. Choosing a variables control chart for subgroups Minitab offers five variables control charts: • • • • • • • X-bar and R − an X-bar chart and R chart displayed in one window X-bar and S − an X-bar chart and S chart displayed in one window I-MR-R/S (Between/Within) − a three-way control chart that uses both between-subgroup and within-subgroup variation. use Zone Options > Display. Zone charts are usually preferred over X-bar or individuals charts because of their utter simplicity: a point is out of control simply. It plots a cumulative score. If the subgroups are not equal. See [5] for discussions of the various weighting schemes. such as the number of defects or defective units. Minitab calculates summary statistics for each subgroup.Quality Control Interpreting the results The sample standard deviations fall randomly within the control chart limits. see Control Charts Overview. Variables Control Charts for Subgroups Overview. Zone Chart Stat > Control Charts > Variables Charts for Subgroups > Zone A hybrid between an X-bar (or individuals) chart and a CUSUM chart. For example. 2. You can conclude that the variability of faults among shifts is due to common causes. . Moving ranges are artificial subgroups created from the individual measurements. a point is plotted for each observation. You can display a chart based on subgroup means or individual observations. the average of the moving range divided by an unbiasing constant. By default. When subgroup sizes are unequal. With individual observations. or enter a historical value for σ. Attributes control charts plot count data.

83 . Variables charts include: • • • • Variables charts for subgroups Variables charts for individuals Time-weighted charts Multivariate charts Structure your data for these charts using the guidelines below. then enter the columns. <Scale> <Labels> <Multiple Graphs> <Data Options> <Zone Options> Data − Variables Control Chart Organize the data for all variables control charts in the same way. with one column for each variable. using the following table as a guide. C1 contains the process data and C2 contains subgroup indicators: 2003 Minitab Inc. Here is the same data set. and so on. In the following example. you must enter your data in one column. Multivariate data must be entered down columns. structured both ways. the second 5 observations are the second row. Note that the first five observations in the left data set (subgroup 1) are the first row of the right-side data set. When subgroups are of unequal size. then create a second column of subscripts which serve as subgroup indicators.Control Charts Observations for a subgroup are in one row of columns: Choose if subgroups are arranged in rows across several columns. with subgroups of size 5. Subgroups are equal size Univariate (one variable) Multivariate (more than one variable) Down columns or across rows Down columns Subgroups are unequal size Down columns with subgroup indicator column Down columns with subgroup indicator column Structure subgroup data down a column or across rows. Worksheet Structure Structure your data down a column or across rows.

Options − Parameters .Quality Control Each time a subscript changes in C2. which signals an out-of-control situation. Missing data See Missing data in control charts for information on how to handle missing data for different types of control charts. The X-bar. If you like. This is equivalent to a Shewhart chart Rule 5 − two out of three points in a row more than two standard deviations from the center line. 3 Comparing a Zone Chart with a Shewhart Chart The zone control chart procedure incorporates some statistical tests for detecting process shift used with conventional Shewhart control charts. a new subgroup begins in C1.. choose Observations for a subgroup are in one row of columns. I-MR. In this example. • • For further discussions of zone control chart properties. Use the stand alone command as an exploratory tool to help you determine the best lambda value for the transformation. you must enter data that approximate a normal distribution. For example. For information on the stand alone Box-Cox transformation command. Nonnormal data To properly interpret Minitab's control charts. Do one of the following: • If subgroups are in one or more columns. X-bar.. . and Individuals charts produce Shewhart control charts for process data. You can access the Box-Cox transformation two ways: by using the Box-Cox transformation option provided with the control chart commands. A second point in the same zone gives another score of 4. enter a number or a column of subscripts. > control chart Options > Parameters Charts X-bar and R. This is equivalent to a Shewhart chart Rule 6 − four out of five points in a row more than one standard deviation from the center line. then click OK. which signals an out-of-control situation. and so on. subgroup 2 has six observations. In Subgroup sizes. X-bar and S. see Box-Cox Transformation. A point in Zone 2 is given a score of 2. For information on the Box-Cox transformation option. To display a zone chart 1 2 Choose Stat > Control Charts > Variables Charts for Subgroups > Zone. then enter one or more columns. or by using the stand alone Box-Cox command. then enter a series of columns. Three more points in the same zone gives a cumulative score of 8. Individuals. see Options − Box-Cox. This is equivalent to a Shewhart chart Rule 1 − a single value beyond three standard deviations from the center. Then. S. If the data are highly skewed. use any dialog box options. you may want to use the Box-Cox transformation to induce normality. Moving Average. • If subgroups are in rows. The cumulative sum of these two points is 8. R. Zone. refer to [5]. A point in Zone 3 is given a score of 4. subgroup 1 has three observations. using the default weights: • A point in Zone 4 is given a score of 8. EWMA 84 2003 Minitab Inc. Keep in mind that a cumulative score equal to or greater than the weight assigned to Zone 4 signals an out-of-control situation. you can use the transformation option to transform the data at the same time you draw the control chart. choose All observations for a chart are in one column.

not individual observations. Pooled standard deviation: Choose to estimate σ using a pooled standard deviation.. enter one or more values. Median moving range: Choose to estimate σ using the median of the moving range. Dialog box items Omit the following samples when estimating parameters: Choose to specify subgroups of the data to omit from the computations for µ and σ. A single value. click control chart Options. Minitab only uses the non-omitted data to find lambda. A list of values. the omitted or used samples apply to the entire data set.Control Charts Use to enter historical data for estimating µ and σ. one for each stage within each variable. then enter individual subgroup numbers or ranges of subgroups. Square root of MSSD: Choose to estimate σ using the square root of MSSD (1/2 the mean of the squared successive differences). Moving Average. change the length of the moving range. Minitab estimates it from the data. you exclude subgroup 1 from your analysis (using control chart > Data Options > Subset). Subgroup Size = 1 Average moving range: Choose to estimate σ using the average of the moving range. change that value to missing (∗) in the worksheet. Method for estimating standard deviation: See [2] for a discussion of the relative merits of the methods listed below. You can also set preferences for the estimation of σ and the length of the moving range for all Minitab sessions using Tools > Options > Control Charts and Quality Tools > Estimating Standard Deviation. Subgroup Size > 1 Rbar: Choose to estimate σ based on the average of the subgroup ranges. > control chart Options > Estimate Charts Zone. Sbar: Choose to estimate σ based on the average of the subgroup standard deviations. When you set options. all affected dialog box settings automatically reflect your preferences. All stages for each variable use the same value. you can enter these values. • In Standard deviation. For example. Each stage for each variable uses a different value. enter one or more values. You can also select one of several methods to estimate σ. Dialog box items Mean: Enter the historical means as constants. if you have a known process parameter or an estimate obtained from past data. Enter entire subgroups. Use the following samples when estimating parameters: Choose to specify subgroups of data to compute µ and σ. To use historical mean and standard deviation values 1 2 In the control chart dialog box. A list of values. not just the subsetted data. CUSUM Omit or includes certain subgroups to estimate µ and σ. if some subgroups have erratic data due to assignable causes that you have corrected. When you include or exclude rows using control chart Options > Estimate and choose to perform a Box-Cox transformation. EWMA. Because Minitab considers an omitted sample (subgroup 1) as part of the entire data set. which applies to all variables and stages. For example.. 3 • • • You can enter historical data in several ways: Options − Estimate . then decide to omit subgroup 2 from the parameter estimation (using control chart Options > Estimate). you can prevent these subgroups from influencing the estimation of process parameters. You can omit or include certain subgroups when estimating parameters using Estimate. 85 . 2003 Minitab Inc. and use biased estimates of σ. For example. Caution When you include or exclude rows using control chart > Data Options > Subset and also omit or use samples to estimate parameters. Do any of the following: • In Mean. Standard deviation: Enter the historical standard deviations as constants. one for each variable. If you do not specify a value for µ or σ. then enter individual subgroup numbers or a range of subgroups. you must enter subgroup 2 as the subgroup to omit. Click OK. To omit one observation from a sample.

Between 2 and 3 standard deviations: Type the weight for zone 3 (default is 4). > control chart Options > Stages Charts All charts except Z-MR 86 2003 Minitab Inc. Moving Range. Minitab rounds the scores displayed on the chart. But you may want to use or omit certain data if they show abnormal behavior. Zone 2 . See [5] for a discussion of the various weighting schemes. reset signals. Click OK. The span must be < 100. choose a method. reset the cumulative score to zero.. Dialog box items Weights Zone 1 . With these charts. Do one of the following: • Choose Omit the following samples when estimating parameters and enter the subgroups or observation numbers you want to omit from the calculations. To change weights.. Click Zone Options. and store the zone scores. Click the Estimate tab. except with the Individuals. and I-MR charts. When you use Omit the following samples when estimating parameters with Length of moving range. Minitab estimates the process parameters from all the data. Use unbiasing constant: Uncheck to use biased estimates of σ. Zone 3 . then click the Weights/Reset tab. click control chart Options. Minitab excludes any moving ranges with omitted data from the calculations. To omit or use subgroups for the estimates of parameters By default. and store scores 1 2 3 4 Choose Stat > Control Charts > Variables Charts for Subgroups > Zone. If you correct the problem. you should attempt to find and eliminate the cause of the problem. a span of two is used because consecutive values have the greatest chance of being alike.Quality Control Length of moving range: Check to enter the number of observations used to calculate the moving range. Then click OK. Reset cumulative score after each signal: Check to reset the cumulative score to zero following each out-of-control signal. Under Estimating standard deviation. Note 3 Click OK. Check Store zone scores to store the zone scores in the worksheet. Minitab assumes the values you enter are subgroup numbers. Check Reset cumulative score after each signal to reset the cumulative score to zero following each out of control signal. Note Zone Chart Options − Weights/Reset Stat > Control Charts > Variables Charts for Subgroups > Zone > Zone Options > Weights/Reset Controls the weights assigned to each zone. click control chart Options. the values are interpreted as observation numbers. reset the cumulative score when an out-of-control point is encountered. By default. . Weights must be > 0. Options − Stages . type the weight for zones 1 through 4. When a process is out of control. • Choose Use the following samples when estimating parameters and enter the subgroups or observation numbers you want to use for the calculations. 1 2 In the control chart dialog box. To choose how the standard deviation is estimated 1 2 In the control chart dialog box.Within 1 standard deviation of center line: Type the weight for zone 1 (default is 0). Zone 4 . but do not need to be integers. Click the Estimate tab. When you use noninteger weights. Minitab uses the zone 4 weight as the critical value when a process is out of control.More than 3 standard deviations: Type the weight for zone 4 (default is 8). Under Weights.Between 1 and 2 standard deviations: Type the weight for zone 2 (default is 2).

When to start a new stage With each new value: Choose to start a new stage each time the value in the column changes. Click the Stages tab. Dialog box items Box-Cox power transformation (W = Y**Lambda): Choose to use when your data are very skewed or when the withinsubgroup variation is unstable. You may want to first run the command Stat > Control Charts > Box-Cox Transformation to help you find the optimal transformation value. Caution If you use Stat > Control Charts > Box-Cox Transformation to find the optimal λ value and choose to store the transformed data with that command. > control chart Options > Box-Cox Charts All except the Attributes charts You can use the Box-Cox power transformation when your data are very skewed or when the within-subgroup variation is unstable to make the data "more normal.) To use this option. 2003 Minitab Inc. do one of the following: • To start a new stage each time the value in the column changes. When you use this transformation. The indicators can be numbers. Under When to start a new value. When executing the command. choose With the first occurrence of these values. enter the column which contains the stage indicators.5 (square root): Choose to use the square root of the data. Dialog box items Define stages (historical groups) with this variable: Enter the column that contains the stage indicators. you must have at least one subgroup with two or more observations: R. Minitab only uses the non-omitted data to find lambda. doing so will double-transform the data. If you like. choose With each new value. Lambda=0.5). the data must be positive. You can enter the same value more than once. Minitab treats each repeated value as a separate occurrence. you should not choose a λ outside the range of −2 and 2. Optimal lambda: Choose to have Minitab search for an optimal value.. when subgroups are across rows). In most cases. Minitab does not accept any values you enter in the Parameters tab for historical means or historical standard deviations. You can also choose any value between − 5 and 5 for λ. You can enter the same value more than once. Minitab treats each repeated value as a separate occurrence. Enclose date/time or text entries in double quotes. 87 . In Define stages (historical groups) with this variable. • To start a new stage at the first occurrence of a certain value.. X-bar and R. Note With the following charts. click control chart Options. then click OK. dates. and X-bar and S. 5 Options − Box-Cox . in which case the natural log is taken. Lambda = 0 (natural log): Choose to use the natural log of the data. When you include or exclude rows using control chart Options > Estimate.Control Charts You can display stages in your process by drawing a "historical chart" − a control chart in which the control limits and center line are estimated independently for different groups in your data. With the first occurrence of these values: Choose to start a new stage at the first occurrence of a certain value. Enter the values. (λ is lambda. do not select the Box-Cox option when you make a control char. then enter the values. To display a historical chart To define stages in your process. you can tell Minitab to start a new stage in one of two ways: • • 1 2 3 4 Each time the value in the column changes At the first occurrence of one or more values In the control chart dialog box. The column must be the same length as the data column (or columns. The Options subdialog box lists the common transformations natural log (λ = 0) and square root (λ = 0. use any dialog box options. Historical charts are particularly useful for comparing data before and after a process improvement. Enclose date/time or text entries in double quotes. or text." The transformation takes the original data to the power λ. you must set up a column of grouping indicators. S. unless λ = 0.

choose Display all subgroups. • Choose Other (enter value(s) between -5 and 5) to transform the data using another lambda value or values. and enter the number of subgroups. Click the Box-Cox tab. > control chart Options > Display Charts All charts except Z-MR Use to display the control chart by stage or by number of plotted points and to display the test results in the Session window. Subgroups to display Display all subgroups: Choose to display all subgroups. under Subgroups to display. 3 88 2003 Minitab Inc. • Choose Optimal lambda for each stage (in a chart with stages) to have Minitab search for optimal values for each stage. then enter the lambda values. choose Each segment contains __ subgroups. To display subgroups 1 2 In the control chart dialog box. Click the Display tab. Check Box-Cox power transformation (W = Y**Lambda). • To split the chart by stage. • To display a specific number of the last subgroups. For help choosing a lambda value.5 (square root) to use the square root of the data. then do one of the following: • Choose Lambda = 0 (natural log) to use the natural log of the data. and enter the number of subgroups for each segment. • Choose Lambda=0. Test results Display test results in Session window: Check to display the test results in the Session window. Do one of the following: • To split the chart into a specific number of subgroups. and enter the number of subgroups for each segment. see the independent Box-Cox transformation command Stat > Control Charts > Box-Cox Transformation. Other (enter value(s) between -5 and 5): Choose to transform the data using another lambda value or values. 3 To split a control chart 1 2 In the control chart dialog box. choose Display last __ subgroups and enter the number of subgroups to display. Click the Display tab. then enter one or more values. Each segment contains __ subgroups: Choose to split the control chart into subgroups of a specified number of points. • Choose Optimal lambda to have Minitab search for an optimal value for all stages.. Options − Display . choose Each segment corresponds to a stage (if chart has stages). To do the Box-Cox power transformation 1 2 In the control chart dialog box. Do one of the following: • To display all subgroups in the control chart. Click OK. click control chart Options. Each segment corresponds to a stage (if chart has stages): Choose to split the control chart by stage or stage when you have specified stages in control chart Options > Stage.Quality Control Optimal lambda for each stage (in a chart with stages): Choose to have Minitab search for optimal values for each stage. Use overall standard deviation: Check to use the overall standard deviation in the estimation of the optimal lambda. under Subgroups to display. click control chart Options.. Split chart into series of segments for display purposes Do not split: Choose not to split the chart. Display last __ subgroups: Choose to display a specific number of subgroups. 3 Click OK. . click control chart Options. Click OK.

Minitab stores one column for the lower control limits and one column for the upper control limits. 2003 Minitab Inc. Standard deviations: Check to store the estimates of the standard deviations.MTW. then click the Weights/Reset tab. Test Failed at points: 6. Control limit values: Check to store the control limit values. Session window output Zone Chart of Length Test Results for Zone Chart of Length TEST. Check the statistics to store in the worksheet. Choose All observations for a chart are in one column. 10 * WARNING * If graph is updated with new data. You also decide to reset the cumulative score following each out-ofcontrol signal. then enter Length. 1 2 3 4 5 6 Open worksheet EXH_QC. Choose Stat > Control Charts > Variables Charts for Subgroups > Zone. 89 . enter 5. You decide to measure the length of 10 sets of cylinders produced during each of five shifts. Subgroup size: Check to store the subgroup sizes. In Subgroup sizes. one row for each plotted point. then click OK. one row for each plotted point. Store these values for each point Point plotted: Check to store the plotted points. one row for each plotted point. Example of zone chart You work in a manufacturing plant concerned about quality improvement. Because a zone control chart is very easy to interpret. Minitab stores the exact cumulative score for each subgroup.Control Charts Zone Chart − Options − Storage Stat > Control Charts > Variables Charts for Subgroups > Zone > Zone Options > Storage Stores your choice of statistics in the worksheet. you decide to evaluate your data with it. one row for each plotted point. Click Zone Options. one row for each plotted point. one row for each stage. Dialog box items Store these estimates for each chart Means: Check to store the estimates of the means. Check Reset cumulative score after each signal. Click OK in each dialog box. one row for each historical stage. Center line value: Check to store the center line value. Stage: Check to store the stage. To store statistics in the worksheet 1 2 Choose Stat > Control Charts > control chart type > control chart > control chart Options > Storage. the results above may no * longer be correct. Zone score: Check to store the zone scores. Cumulative score greater than or equal to zone 4 score. for a total of 50 samples daily.

Choosing a variables control chart for individuals Minitab offers five variables control charts: • • • • I-MR − an Individuals chart and Moving Range chart displayed in one window Z-MR − a chart of standardized individual observations and moving ranges from short run processes Individuals − a chart of individual observations Moving Range − a chart of moving ranges Individuals−Moving Range Chart Stat > Control Charts > Variables Charts for Individuals > I-MR An I-MR chart is an Individuals chart and Moving Range chart in the same graph window. such as length or pressure. since consecutive values have the greatest chance of being alike. as well as detect the presence of special causes. time-weighted charts. and multivariate charts also plot measurement data. the average of the moving range divided by an unbiasing constant. You find that the operator reset the machine following subgroup 6. See [29] for a discussion of how to interpret joint patterns in the two charts. σ. the zone chart detects the process is out of control again at subgroup 10. Variables Charts for Individuals I-MR Chart Variables Control Charts for Individuals Overview data Variables control charts for individuals plot statistics from measurement data. Variables control charts for subgroups. or enter a historical value for σ. for individuals data. the Moving Range chart in the lower half. as he believed the machine was slipping. For more information. you decide the operator may have overcompensated for the problem identified at subgroup 6. see Control Charts Overview. After seeing the subsequent rise in subgroups 7-10 on the zone chart. The Individuals chart is drawn in the upper half of the screen. By default. Seeing both charts together allows you to track both the process level and process variation at the same time. with MRbar / d2. The moving range is of length 2. see Control Charts Overview and Variables Control Charts for Individuals Overview. You can also estimate σ using the median of the moving range. For more information about control charts. which indicates the process is out of control. However. Attributes control charts plot count data. change the length of the moving range.Quality Control Graph window output Interpreting the results The cumulative score at subgroup 6 equals eight. 90 2003 Minitab Inc. such as the number of defects or defective units. I-MR Chart estimates the process variation. .

Multivariate data must be entered down columns. and so on. When subgroups are of unequal size. structured both ways. Worksheet Structure Structure your data down a column or across rows. Here is the same data set. <Scale> <Labels> <Multiple Graphs> <Data Options> <I-MR Options> Data − Variables Control Chart Organize the data for all variables control charts in the same way. the second 5 observations are the second row. using the following table as a guide. In the following example. Subgroups are equal size Univariate (one variable) Multivariate (more than one variable) Down columns or across rows Down columns Subgroups are unequal size Down columns with subgroup indicator column Down columns with subgroup indicator column Structure subgroup data down a column or across rows. 91 . Variables charts include: • • • • Variables charts for subgroups Variables charts for individuals Time-weighted charts Multivariate charts Structure your data for these charts using the guidelines below. Note that the first five observations in the left data set (subgroup 1) are the first row of the right-side data set.Control Charts Dialog box items Variables: Choose one or more columns containing the observations. with one column for each variable. C1 contains the process data and C2 contains subgroup indicators: 2003 Minitab Inc. with subgroups of size 5. you must enter your data in one column. then create a second column of subscripts which serve as subgroup indicators.

Moving Average. For information on the Box-Cox transformation option. Zone. Each stage for each variable uses a different value. you must enter data that approximate a normal distribution.. A single value. or by using the stand alone Box-Cox command. subgroup 2 has six observations.Quality Control Each time a subscript changes in C2. 3 • • • You can enter historical data in several ways: 92 2003 Minitab Inc. enter one or more values. one for each variable. • In Standard deviation. Use the stand alone command as an exploratory tool to help you determine the best lambda value for the transformation. then click OK. All stages for each variable use the same value. > control chart Options > Parameters Charts X-bar and R. For example. If you do not specify a value for µ or σ. Do any of the following: • In Mean. Minitab estimates it from the data. Nonnormal data To properly interpret Minitab's control charts. which applies to all variables and stages. enter one or more columns. To use historical mean and standard deviation values 1 2 In the control chart dialog box. Then. click control chart Options. use any dialog box options. If you like. Individuals. see Box-Cox Transformation. Click OK. For information on the stand alone Box-Cox transformation command.. X-bar. A list of values. you can enter these values. Options − Parameters . Missing data See Missing data in control charts for information on how to handle missing data for different types of control charts. I-MR. you may want to use the Box-Cox transformation to induce normality. subgroup 1 has three observations. one for each stage within each variable. If the data are highly skewed. You can access the Box-Cox transformation two ways: by using the Box-Cox transformation option provided with the control chart commands. . see Options − Box-Cox. In Variables. if you have a known process parameter or an estimate obtained from past data. you can use the transformation option to transform the data at the same time you draw the control chart. and so on. Dialog box items Mean: Enter the historical means as constants. A list of values. EWMA Use to enter historical data for estimating µ and σ. X-bar and S. a new subgroup begins in C1. In this example. To display an I-MR chart 1 2 3 Choose Stat > Control Charts > Variables Charts for Individuals > I-MR. You can omit or include certain subgroups when estimating parameters using Estimate. enter one or more values. Standard deviation: Enter the historical standard deviations as constants.

You can also force the control limits to be constant when sample sizes are unequal. To omit or use subgroups for the estimates of parameters By default. Minitab excludes any moving ranges with omitted data from the calculations. For example. Moving Range. a horizontal line labeled LB is drawn at the lower bound instead. Use the following samples when estimating parameters: Choose to specify observations to compute µ and σ. Because Minitab considers an omitted sample (subgroup 1) as part of the entire data set. you exclude subgroup 1 from your analysis (using control chart > Data Options > Subset). the omitted or used samples apply to the entire data set. Similarly. not just the subsetted data. With these charts. 2003 Minitab Inc. Then click OK. Minitab estimates the process parameters from all the data.. When you use Omit the following samples when estimating parameters with Length of moving range. Do one of the following: • Choose Omit the following samples when estimating parameters and enter the subgroups or observation numbers you want to omit from the calculations. you must enter subgroup 2 as the subgroup to omit. a horizontal line labeled UB is drawn at the upper bound instead. > control chart Options > Estimate Charts Moving Range. then decide to omit subgroup 2 from the parameter estimation (using control chart Options > Estimate). I-MR Omits or includes certain observations to estimate µ and σ. For example. Median moving range: Choose to estimate σ using the median of the moving range. You can set bounds on the upper and lower control limits. if the calculated lower control limit is less than the lower bound. choose a method. 1 2 In the control chart dialog box. then enter individual observation numbers. then enter individual observation numbers. But you may want to use or omit certain data if they show abnormal behavior. Minitab assumes the values you enter are subgroup numbers. The span must be < 100. Method for estimating standard deviation: See [2] for a discussion of the relative merits of the methods listed below.Control Charts Options − Estimate . and use biased estimates of σ.. When the calculated upper control limit is greater than the upper bound. Minitab only uses the non-omitted data to find lambda. Subgroup size = 1 Average moving range: Choose to estimate σ using the average of the moving range. Minitab uses a span of two because consecutive values have the greatest chance of being alike. Under Estimating standard deviation. you can prevent them from influencing the estimation of µ and σ. Note Individuals-Moving Range Chart − Options − S Limits Stat > Control Charts > Variables Charts for Individuals > I-MR > I-MR Options > S Limits Allows you to draw control limits above and below the mean at the multiples of any standard deviation. the values are interpreted as observation numbers. except with the Individuals. if some observations are erratic due to assignable causes that you have corrected. then enter the number of observations. When you include or exclude rows using control chart Options > Estimate and choose to perform a Box-Cox transformation. You can set preferences for the estimation of σ and the length of the moving range using Tools > Options > Control Charts and Quality Tools > Estimating Standard Deviation. Caution When you include or exclude rows using control chart > Data Options > Subset and also omit or use samples to estimate parameters. click control chart Options. To choose how the standard deviation is estimated 1 2 In the control chart dialog box. 93 . Length of moving range: Check to calculate the moving range. Click the Estimate tab. and I-MR charts. change the length of the moving range. • Choose Use the following samples when estimating parameters and enter the subgroups or observation numbers you want to use for the calculations. You can also select one of two methods to estimate σ. click control chart Options. Dialog box items Omit the following samples when estimating parameters: Choose to specify observations to omit from the computations for µ and σ. Note 3 Click OK. By default. Click the Estimate tab.

See Tools > Options > Control Charts and Quality Tools > Other. The test results are displayed in the Session window. If you specify more than one value. 3 Options − Tests . You can choose to change the color of a control chart when the last row of data causes a new test failure for any point. above. Each value is the number of standard deviations the lines should be drawn at.Quality Control For an example. a horizontal line labeled UB will be drawn at the upper bound instead. I-MR Selects a subset of the tests shown in Eight Tests for Special Causes. If the calculated upper control limit is greater than the upper bound. The values entered are the number of standard deviations above and below the center line. Place bounds on control limits of Individuals chart Lower standard deviation limit bound: Check to set a lower bound for the control limits for the individuals chart.. Dialog box items Display control limits at These multiples of the standard deviation: Enter one or more values. a horizontal line labeled LB will be drawn at the lower bound instead. > control chart Options > Tests Charts X-bar and R. You can set preferences for the sensitivity of these tests using Tools > Options > Control Charts and Quality Tools > Define Tests. one that should be investigated. a pair of control limits is drawn for each number in the list. • To set bounds on the control limits for the moving range chart: under Place bounds on control limits of Moving range chart. Minitab will perform tests when subgroup sizes are unequal. Click the S Limits tab. Each test detects a specific pattern in the data plotted on the chart. The occurrence of a pattern suggests a special cause for the variation. 94 2003 Minitab Inc. a horizontal line labeled UB will be drawn at the upper bound instead. click I-MR Options. If a point fails more than one test. If the calculated upper control limit is greater than the upper bound. and below the mean. Click OK in each dialog box. Individuals. If the calculated lower control limit is less than the lower bound. enter one or more values or a column of values in These multiples of the standard deviation. Upper standard deviation limit bound: Check to set an upper bound for the control limits for the individuals chart. check Lower standard deviation limit bound (and/or Upper standard deviation limit bound) and enter a value. If the data change and the control chart is live-linked. Do any of the following: • To specify where control limits are drawn: under Display control limits at. a horizontal line labeled LB will be drawn at the lower bound instead. See Updating Graphs for more information. Each value represents the number of standard deviations below and above the mean. • To set bounds on the control limits for the individuals chart: under Place bounds on control limits of Individuals chart. I-MR-R/S (Between/Within). Each value represents the number of standard deviations below and above the mean. See [1] and [29] for guidance on using these tests. Minitab marks it with the test number on the plot. X-bar and S. Upper standard deviation limit bound: Check to set an upper bound for the control limits for the moving range chart. the test results are not updated in the Session window.. To customize the control limits for I-MR charts 1 2 In the control chart data source dialog box. Tip You can also modify the control limits using the Minitab graph editing features. If the calculated lower control limit is less than the lower bound. see Example of an X-bar chart with tests and customized control limits. Minitab marks it by the lowest numbered test. . check Lower standard deviation limit bound (and/or Upper standard deviation limit bound) and enter a value. Place bounds on control limits of Moving range chart Lower standard deviation limit bound: Check to set a lower bound for the control limits for the moving range chart. X-bar. When a point fails a test.

Click the Stages tab. X-bar and R. choose With each new value. choose Perform no tests.. choose Perform all tests for special causes. or text. do one of the following: • To start a new stage each time the value in the column changes. Dialog box items Define stages (historical groups) with this variable: Enter the column that contains the stage indicators. • To perform no tests. choose Perform the following tests for special causes and check the tests to perform. alternating up and down 2 out of 3 points > 2 standard deviations from center line (same side) 4 out of 5 points > 1 standard deviation from center line (same side) 15 points in a row within 1 standard deviation of center line (either side) 8 points in a row > 1 standard deviation from center line (either side) Perform no tests: Choose to perform no tests for special causes. With the first occurrence of these values: Choose to start a new stage at the first occurrence of a certain value. 95 . 2003 Minitab Inc. Minitab treats each repeated value as a separate occurrence. 3 Options − Stages .. S. When executing the command. Enclose date/time or text entries in double quotes. To perform the tests for special causes 1 2 Choose Stat > Control Charts > control chart type > control chart > control chart Options > Tests. You can enter the same value more than once. you must set up a column of grouping indicators. dates. • To select certain tests. Do one of the following: • To do all tests. In Define stages (historical groups) with this variable. enter the column which contains the stage indicators. Historical charts are particularly useful for comparing data before and after a process improvement. when subgroups are across rows). then check any combination of the tests listed below.Control Charts Dialog box items Tests For Special Causes (default definitions): Perform all tests for special causes: Choose to perform all eight tests shown below. you must have at least one subgroup with two or more observations: R. Perform the following tests for special causes: Choose to specify a subset of the eight tests. > control chart Options > Stages Charts All charts except Z-MR You can display stages in your process by drawing a "historical chart" − a control chart in which the control limits and center line are estimated independently for different groups in your data. 1 point more than 3 standard deviations from center line 9 points in a row on same side of center line 6 points in a row. then enter the values. and X-bar and S. click control chart Options. Under When to start a new value. you can tell Minitab to start a new stage in one of two ways: • • 1 2 3 4 Each time the value in the column changes At the first occurrence of one or more values In the control chart dialog box. The indicators can be numbers. The column must be the same length as the data column (or columns. Display test results in Session window: Check to display the test results in the Session window. Note With the following charts. all increasing or all decreasing 14 points in a row. When to start a new stage With each new value: Choose to start a new stage each time the value in the column changes. Click OK. To display a historical chart To define stages in your process.

> control chart Options > Display Charts All charts except Z-MR Use to display the control chart by stage or by number of plotted points and to display the test results in the Session window. 5 If you like. Other (enter value(s) between -5 and 5): Choose to transform the data using another lambda value or values.5). Lambda=0. You may want to first run the command Stat > Control Charts > Box-Cox Transformation to help you find the optimal transformation value.5 (square root) to use the square root of the data. unless λ = 0. Minitab does not accept any values you enter in the Parameters tab for historical means or historical standard deviations. 96 2003 Minitab Inc. • Choose Other (enter value(s) between -5 and 5) to transform the data using another lambda value or values. > control chart Options > Box-Cox Charts All except the Attributes charts You can use the Box-Cox power transformation when your data are very skewed or when the within-subgroup variation is unstable to make the data "more normal. Options − Box-Cox .. • Choose Optimal lambda for each stage (in a chart with stages) to have Minitab search for optimal values for each stage. Lambda = 0 (natural log): Choose to use the natural log of the data. The Options subdialog box lists the common transformations natural log (λ = 0) and square root (λ = 0. do not select the Box-Cox option when you make a control char. To do the Box-Cox power transformation 1 2 In the control chart dialog box. 3 Click OK. doing so will double-transform the data. When you use this transformation. then enter the lambda values. Minitab treats each repeated value as a separate occurrence.) To use this option. • Choose Optimal lambda to have Minitab search for an optimal value for all stages.. (λ is lambda. When you include or exclude rows using control chart Options > Estimate. then click OK. you should not choose a λ outside the range of −2 and 2. the data must be positive. Enter the values. Caution If you use Stat > Control Charts > Box-Cox Transformation to find the optimal λ value and choose to store the transformed data with that command. . use any dialog box options. Use overall standard deviation: Check to use the overall standard deviation in the estimation of the optimal lambda. Click the Box-Cox tab. Options − Display . choose With the first occurrence of these values. • Choose Lambda=0. then enter one or more values. Optimal lambda for each stage (in a chart with stages): Choose to have Minitab search for optimal values for each stage. Minitab only uses the non-omitted data to find lambda." The transformation takes the original data to the power λ. then do one of the following: • Choose Lambda = 0 (natural log) to use the natural log of the data. click control chart Options. For help choosing a lambda value. in which case the natural log is taken.5 (square root): Choose to use the square root of the data. Optimal lambda: Choose to have Minitab search for an optimal value. see the independent Box-Cox transformation command Stat > Control Charts > Box-Cox Transformation. In most cases. Dialog box items Box-Cox power transformation (W = Y**Lambda): Choose to use when your data are very skewed or when the withinsubgroup variation is unstable. Enclose date/time or text entries in double quotes..Quality Control • To start a new stage at the first occurrence of a certain value. You can enter the same value more than once. Check Box-Cox power transformation (W = Y**Lambda). You can also choose any value between − 5 and 5 for λ..

choose Display last __ subgroups and enter the number of subgroups to display. Click the Display tab. Click OK. Store these values for each point Point plotted: Check to store the plotted points. Click OK. 97 . Control limit values: Check to store the control limit values. Each segment corresponds to a stage (if chart has stages): Choose to split the control chart by stage or stage when you have specified stages in control chart Options > Stage. one row for each plotted point. and enter the number of subgroups for each segment. choose Each segment corresponds to a stage (if chart has stages). Standard deviations: Check to store the estimates of the standard deviations. Stage: Check to store the stage. Display last __ subgroups: Choose to display a specific number of subgroups. • To display a specific number of the last subgroups. under Subgroups to display. and enter the number of subgroups for each segment. click control chart Options. Do one of the following: • To display all subgroups in the control chart. Minitab stores one column for the lower control limits and one column for the upper control limits. I-MR Use to store your choice of statistics in the worksheet. To store statistics in the worksheet 1 2 Choose Stat > Control Charts > control chart type > control chart > control chart Options > Storage. Test results Display test results in Session window: Check to display the test results in the Session window. one row for each plotted point. 2003 Minitab Inc. one row for each stage. 3 To split a control chart 1 2 In the control chart dialog box. Click the Display tab. Do one of the following: • To split the chart into a specific number of subgroups. Center line value: Check to store the center line value. one row for each plotted point. Each segment contains __ subgroups: Choose to split the control chart into subgroups of a specified number of points. one row for each stage. one row for each plotted point. choose Each segment contains __ subgroups.Control Charts Subgroups to display Display all subgroups: Choose to display all subgroups. under Subgroups to display. Check the statistics to store in the worksheet. Split chart into series of segments for display purposes Do not split: Choose not to split the chart. • To split the chart by stage. and enter the number of subgroups. 3 Options − Storage Stat > Control Charts > control chart type > control chart > control chart Options > Storage Charts Individuals. Dialog box items Store these estimates for each chart Means: Check to store estimates of the means. choose Display all subgroups. then click OK. To display subgroups 1 2 In the control chart dialog box. one row for each plotted point. click control chart Options. Test results: Check to store the results of any test performed.

32. 12. 45 Test Results for MR Chart of Weight TEST 1. 31. Test Failed at points: 5. Choose Perform all tests for special causes. 30. Session window output I-MR Chart of Weight Test Results for I Chart of Weight TEST 1. 23. the results above may no * longer be correct. 35. 19. 36 TEST 5. then click the Tests tab. 6. In Variables. 98 2003 Minitab Inc. 34.Quality Control Example of I-MR chart As the distribution manager at a limestone quarry. 17. 18.MTW. 4 out of 5 points more than 1 standard deviation from center line (on one side of CL).00 standard deviations from center line. 21. 20. 16. you want to monitor the weight (in pounds) and variation in the 45 batches of limestone that are shipped weekly to an important client. Click I-MR Options. 14. Test Failed at points: 25 * WARNING * If graph is updated with new data. 9 points in a row on same side of center line. 31. Test Failed at points: 24. 13. You previously created a Moving Average chart. 10. Test Failed at points: 14. 31. 29. 1 2 3 4 5 Open the worksheet EXH_QC. One point more than 3. 33. enter Weight. 2 out of 3 points more than 2 standard deviations from center line (on one side of CL). then click OK in each dialog box. 30. Test Failed at points: 9. 45 TEST 2. 45 TEST 6. Each batch should weigh approximately 930 pounds.00 standard deviations from center line. . Now you want to examine the same data using an individuals and moving range chart. 11. 15. Choose Stat > Control Charts > Variables Charts for Individuals > I-MR. One point more than 3. 30. 44. 7.

Seeing both charts together lets you track both the process level and process variation at the same time. For more information about control charts. The moving range chart shows one point above the control limit. and multivariate charts also plot measurement data. Short run charts provide a solution to these problems by pooling and standardizing the data in various ways. time-weighted charts. Standard control charting techniques rely upon a sufficiently large amount of data to reliably estimate process parameters. Short run processes often do not have enough data in each run to produce good estimates of the process parameters. or different products. Variables control charts for subgroups. such as the process means (µ) and process standard deviations (σ). 99 . For example. You can use a single machine or process to produce many different parts. you would need a separate control chart for each part made by the process. The chart for individual observations (Z) displays above the chart for moving ranges (MR). 2003 Minitab Inc. Attributes control charts plot count data. suggesting the presence of special causes. Choosing a variables control chart for individuals Minitab offers five variables control charts: • • • • I-MR − an Individuals chart and Moving Range chart displayed in one window Z-MR − a chart of standardized individual observations and moving ranges from short run processes Individuals − a chart of individual observations Moving Range − a chart of moving ranges Z-MR Chart Stat > Control Charts > Variables Charts for Individuals > Z-MR A Z-MR chart is a chart of standardized individual observations (Z) and moving ranges (MR) from a short run process. because it is likely that all parts would not have the same mean and standard deviation. then reset the machine to produce a different part in the next run. Even if the runs are large enough to obtain estimates. you may produce only 20 units of a part. such as length or pressure. You should closely examine the quarry's processes to improve control over the weight of limestone shipments. See [29] for a discussion of how to interpret joint patterns in the two charts.Control Charts Graph window output Interpreting the results The individuals chart shows 6 points outside the control limits and 22 points inside the control limits exhibiting a nonrandom pattern. such as the number of defects or defective units. see Control Charts Overview. Z-MR Chart Variables Control Charts for Individuals Overview data Variables control charts for individuals plot statistics from measurement data. for individuals data.

In Variables. enter a column containing the part/product name or number for each measurement. an upper limit at +3. enter a column containing the historical means. Thus. use any dialog box options. In Part indicator. then enter the column containing the values. enter a column containing the historical standard deviations.Quality Control Several methods are commonly used for short runs. then dividing by the standard deviation. Click OK. To display a Z-MR chart 1 2 3 Choose Stat > Control Charts > Variables Charts for Individuals > Z-MR. enter one or more columns. <Scale> <Labels> <Multiple Graphs> <Data Options> <Z-MR Options> Data − Z-MR Chart Your worksheet should consist of a pair of columns: a data column and column containing the corresponding part/product name or number. each row contains a measurement. The part/product data defines the groupings for estimating process parameters. Standardizing allows you to evaluate data from different runs by interpreting a single control chart. The part column can contain text or numeric data. Part indicator: Enter the column containing the part/product name or number for each measurement. Z-MR Chart − Options – Parameters Stat > Control Charts > Variables Charts for Individuals > Z-MR > Z-MR Options > Parameters Use to estimate process means and standard deviations. To specify parameters for Z-MR chart 1 2 Choose Stat > Control Charts > Variables Charts for Individuals > Z-MR > Z-MR Options > Parameters. If the average and the standard deviation can be obtained. see Control Charts Overview and Variables Control Charts for Individuals Overview. In addition. then the process data can be standardized by subtracting the mean and dividing the result by the standard deviation. The most general method assumes that each part or batch produced by a process has its own unique average and standard deviation. You can estimate the mean and process variation from the data various ways. 3 100 2003 Minitab Inc. and a lower limit at −3. This column must contain a value for each observation in the data column. You may find Calc > Make Patterned Data > Simple Set of Numbers useful for entering patterned data for the part/product number. Use Z-MR Chart with short run processes when there is not enough data in each run to produce good estimates of process parameters. In addition. The resulting control chart has a center line at 0. If you like. each time Minitab encounters a change in the part/product name column a new run is defined. and a part or product name. The standardized data all come from a population with µ = 0 and σ = 1. This column must contain a value for each observation in the data column. each time Minitab encounters a change in the part/product name column. or supply historical values. • In Standard deviations. Standard deviations: Enter a historical standard deviation for each part/product. The part/product data defines the groupings for estimating process parameters. You may find Calc > Make Patterned Data > Simple Set of Numbers useful for entering the part/product number. The part column must have the same number of rows as the data column. then click OK. See [7] and [9] for details. Dialog box items Variables: Enter one or more columns containing the process data. For more information. a new run is defined. Dialog box items Means: Enter a historical mean for each part/product. Z-MR Chart standardizes the measurement data by subtracting the mean to center the data. Do any of the following: • In Means. Now you can use a single plot for the standardized data from different parts or products. then enter the column containing the values. .

a horizontal line labeled LB will be drawn at the lower bound instead. choose a method to estimate the standard deviation. Dialog box items Display control limits at These multiples of the standard deviation: Enter one or more values. Similarly. see Example of an X-bar chart with tests and customized control limits. The values entered are the number of standard deviations above and below the center line. The natural log transformation stabilizes the variation in cases where variation increases as the size of the measurement increases. To estimate the standard deviation for Z-MR chart 1 2 Choose Stat > Control Charts > Variables Charts for Individuals > Z-MR > Z-MR Options > Estimate. use log) Under For each group. You can also force the control limits to be constant when sample sizes are unequal. If the calculated upper control limit is greater than the upper bound. then click OK. a horizontal line labeled UB is drawn at the upper bound instead.Control Charts Z-MR Chart − Options – Estimate Stat > Control Charts > Variables Charts for Individuals > Z-MR > Z-MR Options > Estimate Use to enter historical data for standardizing the data. choose one of the following: • By runs • By parts (combine all observations for the same part) • Constant (combine all observations) • Relative to size (combine all observations. then estimate σ for that part. Constant (combine all observations): Choose to pool all the data − across runs and parts − to obtain a common estimate of σ. Tip You can also modify the control limits using the Minitab graph editing features. For an example. Length of moving range: Enter the length of the moving range. calculate. If you specify more than one value. 2003 Minitab Inc. Median Moving Range: Choose to estimate σ using the median of the moving range. This is repeated to estimate σ for all parts. Upper standard deviation limit bound: Check to set an upper bound for the control limits for the moving range chart. If the calculated lower control limit is less than the lower bound. a pair of control limits is drawn for each number in the list. By default. How to define groups of observations. 3 3 Z-MR Chart − Options − S Limits Stat > Control Charts > Variables Charts for Individuals > Z-MR > Z-MR Options > S Limits Allows you to draw control limits above and below the mean at the multiples of any standard deviation. Dialog box items Method for estimating the standard deviation How to define groups of observations By runs: Choose to estimate σ for each run independently. You can set bounds on the upper and lower control limits. if the calculated lower control limit is less than the lower bound. use log): Choose to take the natural log of the data. When the calculated upper control limit is greater than the upper bound. a horizontal line labeled LB is drawn at the lower bound instead. By parts (combine all observations for same part): Choose to combine the data from runs of the same part. For each group. The span must be < 100. Under Methods for estimating the standard deviation. a horizontal line labeled UB will be drawn at the upper bound instead. If you like. Relative to size (combine all observations. Place bounds on control limits of Moving range chart Lower standard deviation limit bound: Check to set a lower bound for the control limits for the moving range chart. calculate Average Moving Range: Choose to estimate σ using the average of the moving range. then pool the transformed data − across all runs and parts − to obtain a common estimate of σ for the transformed data. See Methods and formulas for information on choosing an estimation method. 101 . use any dialog box options. Minitab uses a span of two because consecutive values have the greatest chance of being alike.

Dialog box items Box-Cox power transformation (W = Y**Lambda): Choose to use when your data are very skewed or when the withinsubgroup variation is unstable. 3 Options − Box-Cox . • Choose Other (enter value(s) between -5 and 5) to transform the data using another lambda value or values. then enter one or more values. In most cases." The transformation takes the original data to the power λ. Click OK in each dialog box. above. (λ is lambda. check Lower standard deviation limit bound (and/or Upper standard deviation limit bound) and enter a value. When you use this transformation. When you include or exclude rows using control chart Options > Estimate. Minitab does not accept any values you enter in the Parameters tab for historical means or historical standard deviations. doing so will double-transform the data. Minitab only uses the non-omitted data to find lambda. • Choose Optimal lambda to have Minitab search for an optimal value for all stages.. the data must be positive. Use overall standard deviation: Check to use the overall standard deviation in the estimation of the optimal lambda. unless λ = 0. Caution If you use Stat > Control Charts > Box-Cox Transformation to find the optimal λ value and choose to store the transformed data with that command. Other (enter value(s) between -5 and 5): Choose to transform the data using another lambda value or values.Quality Control To customize the control limits for Z-MR charts 1 2 In the control chart data source dialog box. click control chart Options. • To set bounds on the control limits for the moving range chart: under Place bounds on control limits of Moving range chart. > control chart Options > Box-Cox Charts All except the Attributes charts You can use the Box-Cox power transformation when your data are very skewed or when the within-subgroup variation is unstable to make the data "more normal. For help choosing a lambda value. enter one or more values in These multiples of the standard deviation. Lambda = 0 (natural log): Choose to use the natural log of the data.5).) To use this option. in which case the natural log is taken. . then do one of the following: • Choose Lambda = 0 (natural log) to use the natural log of the data. Lambda=0. do not select the Box-Cox option when you make a control char. 3 Click OK. You may want to first run the command Stat > Control Charts > Box-Cox Transformation to help you find the optimal transformation value. 102 2003 Minitab Inc. To do the Box-Cox power transformation 1 2 In the control chart dialog box. Optimal lambda: Choose to have Minitab search for an optimal value.. You can also choose any value between − 5 and 5 for λ. and below the mean. • Choose Lambda=0. Each value is the number of standard deviations the lines should be drawn at. you should not choose a λ outside the range of −2 and 2. The Options subdialog box lists the common transformations natural log (λ = 0) and square root (λ = 0. see the independent Box-Cox transformation command Stat > Control Charts > Box-Cox Transformation. then enter the lambda values. • Choose Optimal lambda for each stage (in a chart with stages) to have Minitab search for optimal values for each stage.5 (square root) to use the square root of the data. Optimal lambda for each stage (in a chart with stages): Choose to have Minitab search for optimal values for each stage. Check Box-Cox power transformation (W = Y**Lambda). Each value represents the number of standard deviations below and above the mean.5 (square root): Choose to use the square root of the data. click Z-MR Options. Do any of the following: • To specify where control limits are drawn: under Display control limits at. Click the S Limits tab. Click the Box-Cox tab.

then enter the size.572 1. one row for each plotted point. Standard deviations: Check to store the estimates of the standard deviations.2227 1. use log) option for estimating σ.799 Mean 1.7847 Constant Z −0.5015 1. the standardized values (Z) are obtained by: Z = (X − µ) / σ where µ is the overall mean for a particular part.8800 −0.5015 1.7847 1. • Choose Into subsets of size.7136 −0. and σ is the estimate of the process standard deviation for each X. one row for each historical stage. When you choose the Relative to size (combine all observations. The following table shows how the Z-values vary depending on the method chosen to estimate σ: Run # 1 1 1 2 2 2 Fiber # 134 134 134 221 221 221 Thickness 1. To split a series for Z-MR chart 1 2 Choose Stat > Control Charts > Variables Charts for Individuals > Z-MR > Z-MR Options > Display.9846 −0.715 1. Do any of the following: • Choose By runs to split the series by runs. Center line value: Check to store the center line value. Standardizing the Data In all cases. Into subsets of size: Choose to split the series by size and enter the size.1774 −0.9554 1.8490 0.6240 0. then click OK.Control Charts Z-MR Chart − Options − Display Stat > Control Charts > Variables Charts for Individuals > Z-MR > Z-MR Options > Display Use to split or sort the series.1280 2003 Minitab Inc. X is the natural log of the data. Dialog box items Split series: Check to split the series By runs: Choose to split the series by runs.2131 1. To store statistics in the worksheet 1 2 Choose Stat > Control Charts > control chart type > control chart > control chart Options > Storage.9735 0.1865 By Parts Z −0. one row for each plotted point. one row for each stage.2165 1.1973 −0. Control limit values: Check to store the control limit values.1997 Relative to size Z −0.1569 0. Click OK.0129 −0.3729 −0.486 1. one row for each plotted point.7847 1.8518 0.5015 1.9674 1.883 1.435 1. Check the statistics to store in the worksheet. Minitab stores one column for the lower control limits and one column for the upper control limits.0022 −0. The estimate of σ depends on the method chosen. 103 .1742 By Runs Z −0. Store these values for each point Point plotted: Check to store the plotted points.9288 0. Dialog box items Store these estimates for each chart Means: Check to store the estimates of the means. 3 Z-MR Options − Storage Stat > Control Charts > Variables Charts for Individuals > Z-MR > Z-MR Options > Storage Use to store your choice of statistics in the worksheet.6731 0.

6104 −0.1365 −0.2476 Example of Z-MR chart Suppose you work in a paper manufacturing plant and are concerned about quality improvement. enter Thicknes. Click OK in each dialog box. suggesting that only common causes affect the paper making process.6698 −0. so you plan to use the Relative to size (combine all observations.6187 0.7232 −0.Quality Control 3 3 3 4 4 4 5 5 5 134 134 134 221 221 221 077 077 077 1.9341 0.9024 0. Choose Stat > Control Charts > Variables Charts for Individuals > Z-MR.1477 −0.7847 1.1909 −0. You then produce a Z-MR chart for the individual observations (Z) and the moving ranges (MR) from your short run paper-making process.3883 0.2193 0.2332 −1. enter Grade. Graph window output Interpreting the results All points are in control for all five runs for both the process means and moving ranges.548 1.0293 0. In Part indicator. 1 2 3 4 5 6 Open the worksheet EXH_QC.5015 1.5995 0.832 1. You know that the variation in your process is proportional to the thickness of the paper being produced.2117 −0.1327 −0.7520 0.6987 0.768 1.6215 0.7847 1.6494 −0.1473 −0.6987 0. use log).344 1.6104 −0.457 1.5405 −0.7847 1.5529 −0.6606 0.5015 1. Click Z-MR Options.427 1. You collect data from five runs including three different grades of paper. Because your process makes paper in short runs.3883 1.6394 0.5015 1. . you need to employ standardized control charting techniques to assess quality. In Variables.5812 0.2034 −0. 104 2003 Minitab Inc.6388 0.8977 0.1991 0.511 1. use log) option to estimate σ. then click the Estimate tab. Under How to define groups of observations.MTW.2476 0.6681 −0.6921 0.711 1.404 1.3883 1.5761 0. choose Relative to size (combine all observations.

with MRbar / d2. change the length of the moving range. Subgroups are equal size Univariate (one variable) Multivariate (more than one variable) Down columns or across rows Down columns Subgroups are unequal size Down columns with subgroup indicator column Down columns with subgroup indicator column 2003 Minitab Inc. and multivariate charts also plot measurement data.Control Charts Individuals Chart Variables Control Charts for Individuals Overview data Variables control charts for individuals plot statistics from measurement data. the moving range is of length 2. such as the number of defects or defective units. using the following table as a guide. Multivariate data must be entered down columns. For more information about control charts. or enter historical values of σ. 105 . Dialog box items Variables: Enter one or more columns containing the outcome measures. individuals chart estimates the process variation. since consecutive values have the greatest chance of being alike. By default. time-weighted charts. Attributes control charts plot count data. with one column for each variable. Choosing a variables control chart for individuals Minitab offers five variables control charts: • • • • I-MR − an Individuals chart and Moving Range chart displayed in one window Z-MR − a chart of standardized individual observations and moving ranges from short run processes Individuals − a chart of individual observations Moving Range − a chart of moving ranges Individuals Chart Stat > Control Charts > Variables Charts for Individuals > Individuals An individuals chart is a control chart of individual observations. Variables control charts for subgroups. the average of the moving range divided by an unbiasing constant. By default. Variables charts include: • • • • Variables charts for subgroups Variables charts for individuals Time-weighted charts Multivariate charts Structure your data for these charts using the guidelines below. You can also estimate σ using the median of the moving range. see Control Charts Overview and Variables Control Charts for Individuals Overview. You can use individuals charts to track the process level and detect the presence of special causes when the sample size is 1. for individuals data. such as length or pressure. <Scale> <Labels> <Multiple Graphs> <Data Options> <I Chart Options> Data − Variables Control Chart Organize the data for all variables control charts in the same way. Moving ranges are artificial subgroups created from the individual measurements. see Control Charts Overview. σ. For more information. Worksheet Structure Structure your data down a column or across rows.

with subgroups of size 5. Note that the first five observations in the left data set (subgroup 1) are the first row of the right-side data set. EWMA 106 2003 Minitab Inc. then create a second column of subscripts which serve as subgroup indicators. Options − Parameters . or by using the stand alone Box-Cox command. and so on.. You can access the Box-Cox transformation two ways: by using the Box-Cox transformation option provided with the control chart commands. If the data are highly skewed. subgroup 1 has three observations.Quality Control Structure subgroup data down a column or across rows. see Box-Cox Transformation. see Options − Box-Cox. you may want to use the Box-Cox transformation to induce normality. you must enter your data in one column. In this example. you can use the transformation option to transform the data at the same time you draw the control chart. a new subgroup begins in C1. In the following example. When subgroups are of unequal size. X-bar and S. Missing data See Missing data in control charts for information on how to handle missing data for different types of control charts. In Variables.. Use the stand alone command as an exploratory tool to help you determine the best lambda value for the transformation. subgroup 2 has six observations. enter one or more data columns. . you must enter data that approximate a normal distribution. > control chart Options > Parameters Charts X-bar and R. If you like. then click OK. use any dialog box options. and so on. Individuals. Then. Here is the same data set. For information on the stand alone Box-Cox transformation command. structured both ways. Nonnormal data To properly interpret Minitab's control charts. X-bar. I-MR. Moving Average. Zone. For information on the Box-Cox transformation option. the second 5 observations are the second row. To display an individuals chart 1 2 3 Choose Stat > Control Charts > Variables Charts for Individuals > Individuals. C1 contains the process data and C2 contains subgroup indicators: Each time a subscript changes in C2.

Use the following samples when estimating parameters (eg. the omitted or used samples apply to the entire data set. By default. one for each variable.Control Charts Use to enter historical data for estimating µ and σ. A list of values. 1 In the control chart dialog box. Use unbiasing constant: Uncheck to use unbiased estimates of σ. All stages for each variable use the same value. Each stage for each variable uses a different value. if you have a known process parameter or an estimate obtained from past data. Method for estimating standard deviation: See [2] for a discussion of the relative merits of the methods listed below. For example. Because Minitab considers an omitted sample (subgroup 1) as part of the entire data set. Square root of MSSD: Choose to estimate σ using the square root of MSSD (half the mean of the squared successive differences). change the length of the moving range. Caution When you include or exclude rows using control chart > Data Options > Subset and also omit or use samples to estimate parameters. If you do not specify a value for µ or σ. enter one or more values. then enter individual observation numbers. you can prevent these subgroups from influencing the estimation of process parameters. To omit or use subgroups for the estimates of parameters By default. then decide to omit subgroup 2 from the parameter estimation (using control chart Options > Estimate). For example. To use historical mean and standard deviation values 1 2 In the control chart dialog box. Dialog box items Omit the following samples when estimating parameters (eg. and use biased estimates of σ. Minitab estimates the process parameters from all the data. 3 12:15): Choose to specify observations to compute µ and σ. A single value. Minitab only uses the nonomitted data to find lambda. you exclude subgroup 1 from your analysis (using control chart > Data Options > Subset). Minitab estimates it from the data. 2003 Minitab Inc. Subgroup size = 1 Average moving range: Choose to estimate σ using the average of the moving range. Minitab uses a span of two because consecutive values have the greatest chance of being alike. 3 • • • You can enter historical data in several ways: Individuals Chart − Options − Estimate Stat > Control Charts > Variables Charts > Individuals > I Chart Options > Estimate Allows you to omit certain observations to estimate µ and σ. 3 12:15): Choose to specify observations to omit from the computations for µ and σ. For example. one for each stage within each variable. Standard deviation: Enter the historical standard deviations as constants. 107 . A list of values. Click OK. click control chart Options. if some observations are erratic due to assignable causes that you have corrected. Length of moving range: Type the length of the moving range. You can also select one of three methods to estimate σ. You can omit or include certain subgroups when estimating parameters using Estimate. Do any of the following: • In Mean. not just the subsetted data. • In Standard deviation. You can set preferences for the estimation of σ and the length of the moving range using Tools > Options > Control Charts and Quality Tools > Estimating Standard Deviation. you can enter these values. Median moving range: Choose to estimate σ using the median of the moving range. But you may want to use or omit certain data if they show abnormal behavior. you must enter subgroup 2 as the subgroup to omit. which applies to all variables and stages. enter one or more values. The span must be < 100. click control chart Options. Dialog box items Mean: Enter the historical means as constants. When you include or exclude rows using control chart Options > Estimate and choose to perform a Box-Cox transformation. then enter individual observation numbers. Click the Estimate tab.

Each test detects a specific pattern in the data plotted on the chart.. You can set bounds on the upper and/or lower control limits. With these charts. a horizontal line labeled UB is drawn at the upper bound instead. • To set bounds on the control limits: check Lower standard deviation limit bound (and/or Upper standard deviation limit bound) and enter a value.. 108 2003 Minitab Inc. Click the S Limits tab. and below the mean. > control chart Options > Tests Charts X-bar and R. choose a method. click control chart Options. X-bar and S. and I-MR charts. To choose how the standard deviation is estimated 1 2 In the control chart dialog box. Similarly. If the calculated upper control limit is greater than the upper bound. Tip You can also modify the control limits using Minitab's graph editing features. a horizontal line labeled UB will be drawn at the upper bound instead. The occurrence of a pattern suggests a special cause for the variation. When you use Omit the following samples when estimating parameters with Length of moving range. Dialog box items Display control limits at These multiples of the standard deviation: Enter one or more values. The values entered are the number of standard deviations above and below the center line. Note 3 Click OK. Under Estimating standard deviation. When the calculated upper control limit is greater than the upper bound. enter one or more values in These multiples of the standard deviation.. Minitab excludes any moving ranges with omitted data from the calculations. above. Place bounds on control limits Lower standard deviation limit bound: Check to set a lower bound for the control limits. Click the Estimate tab. except with the Individuals. X-bar. • Choose Use the following samples when estimating parameters and enter the subgroups or observation numbers you want to use for the calculations. if the calculated lower control limit is less than the lower bound. one that should be investigated. Individuals. Do any of the following: • To specify where control limits are drawn: under Display control limits at. Minitab assumes the values you enter are subgroup numbers. Moving Range Allows you to draw control limits above and below the mean at the multiples of any standard deviation. Each value is the number of standard deviations the lines should be drawn at. To customize the control limits 1 2 In the control chart dialog box. Note S Limits . Moving Range. the values are interpreted as observation numbers. a horizontal line labeled LB is drawn at the lower bound instead. Click OK in each dialog box. . Then click OK. I-MR Selects a subset of the tests shown in Eight Tests for Special Causes.Quality Control 2 Do one of the following: • Choose Omit the following samples when estimating parameters and enter the subgroups or observation numbers you want to omit from the calculations. > control chart Options > S Limits Charts Individuals. a pair of control limits is drawn for each value in the list.. If you specify more than one value. 3 Options − Tests . Each value represents the number of standard deviations below and above the mean. Upper standard deviation limit bound: Check to set an upper bound for the control limits. see Example of an X-bar chart with tests and customized control limits. click control chart Options. a horizontal line labeled LB will be drawn at the lower bound instead. For an example. If the calculated lower control limit is less than the lower bound. I-MR-R/S (Between/Within).

Minitab marks it with the test number on the plot. > control chart Options > Stages Charts All charts except Z-MR You can display stages in your process by drawing a "historical chart" − a control chart in which the control limits and center line are estimated independently for different groups in your data. the test results are not updated in the Session window.. X-bar and R. choose Perform no tests. 2003 Minitab Inc. choose Perform the following tests for special causes and check the tests to perform. Do one of the following: • To do all tests. When to start a new stage With each new value: Choose to start a new stage each time the value in the column changes. Dialog box items Tests For Special Causes (default definitions): Perform all tests for special causes: Choose to perform all eight tests shown below. then enter the values. all increasing or all decreasing 14 points in a row. Perform the following tests for special causes: Choose to specify a subset of the eight tests. To perform the tests for special causes 1 2 Choose Stat > Control Charts > control chart type > control chart > control chart Options > Tests. Note With the following charts. Dialog box items Define stages (historical groups) with this variable: Enter the column that contains the stage indicators. Click OK. • To select certain tests. you must have at least one subgroup with two or more observations: R. Enclose date/time or text entries in double quotes. then check any combination of the tests listed below. If a point fails more than one test. You can set preferences for the sensitivity of these tests using Tools > Options > Control Charts and Quality Tools > Define Tests. Minitab marks it by the lowest numbered test.Control Charts When a point fails a test. alternating up and down 2 out of 3 points > 2 standard deviations from center line (same side) 4 out of 5 points > 1 standard deviation from center line (same side) 15 points in a row within 1 standard deviation of center line (either side) 8 points in a row > 1 standard deviation from center line (either side) Perform no tests: Choose to perform no tests for special causes. You can choose to change the color of a control chart when the last row of data causes a new test failure for any point. Display test results in Session window: Check to display the test results in the Session window. If the data change and the control chart is live-linked. See [1] and [29] for guidance on using these tests. 3 Options − Stages . 1 point more than 3 standard deviations from center line 9 points in a row on same side of center line 6 points in a row. Minitab will perform tests when subgroup sizes are unequal. 109 . choose Perform all tests for special causes.. The test results are displayed in the Session window. • To perform no tests. Minitab treats each repeated value as a separate occurrence. Historical charts are particularly useful for comparing data before and after a process improvement. S. and X-bar and S. With the first occurrence of these values: Choose to start a new stage at the first occurrence of a certain value. See Updating Graphs for more information. You can enter the same value more than once. See Tools > Options > Control Charts and Quality Tools > Other.

Minitab only uses the non-omitted data to find lambda. dates. The indicators can be numbers. You can enter the same value more than once. Dialog box items Box-Cox power transformation (W = Y**Lambda): Choose to use when your data are very skewed or when the withinsubgroup variation is unstable." The transformation takes the original data to the power λ.) To use this option. then enter the lambda values. You may want to first run the command Stat > Control Charts > Box-Cox Transformation to help you find the optimal transformation value. enter the column which contains the stage indicators. in which case the natural log is taken. Lambda = 0 (natural log): Choose to use the natural log of the data. Caution If you use Stat > Control Charts > Box-Cox Transformation to find the optimal λ value and choose to store the transformed data with that command. do one of the following: • To start a new stage each time the value in the column changes.5). choose With each new value. unless λ = 0. Enclose date/time or text entries in double quotes.5 (square root) to use the square root of the data. The Options subdialog box lists the common transformations natural log (λ = 0) and square root (λ = 0. In most cases. then do one of the following: • Choose Lambda = 0 (natural log) to use the natural log of the data. you must set up a column of grouping indicators. doing so will double-transform the data. In Define stages (historical groups) with this variable. when subgroups are across rows). • To start a new stage at the first occurrence of a certain value. • Choose Lambda=0. Click the Box-Cox tab. 5 Options − Box-Cox . click control chart Options. Minitab does not accept any values you enter in the Parameters tab for historical means or historical standard deviations. > control chart Options > Box-Cox Charts All except the Attributes charts You can use the Box-Cox power transformation when your data are very skewed or when the within-subgroup variation is unstable to make the data "more normal. then click OK. When you use this transformation. Under When to start a new value. choose With the first occurrence of these values. click control chart Options. Use overall standard deviation: Check to use the overall standard deviation in the estimation of the optimal lambda. If you like. . Enter the values. When executing the command. use any dialog box options. To do the Box-Cox power transformation 1 2 In the control chart dialog box. (λ is lambda.Quality Control To display a historical chart To define stages in your process. Check Box-Cox power transformation (W = Y**Lambda). do not select the Box-Cox option when you make a control char. you can tell Minitab to start a new stage in one of two ways: • • 1 2 3 4 Each time the value in the column changes At the first occurrence of one or more values In the control chart dialog box. or text.5 (square root): Choose to use the square root of the data. 110 2003 Minitab Inc. • Choose Optimal lambda to have Minitab search for an optimal value for all stages. Other (enter value(s) between -5 and 5): Choose to transform the data using another lambda value or values. When you include or exclude rows using control chart Options > Estimate. Lambda=0. the data must be positive... Optimal lambda for each stage (in a chart with stages): Choose to have Minitab search for optimal values for each stage. Click the Stages tab. Minitab treats each repeated value as a separate occurrence. You can also choose any value between − 5 and 5 for λ. you should not choose a λ outside the range of −2 and 2. The column must be the same length as the data column (or columns. Optimal lambda: Choose to have Minitab search for an optimal value.

Standard deviations: Check to store the estimates of the standard deviations. choose Display last __ subgroups and enter the number of subgroups to display. Store these values for each point Point plotted: Check to store the plotted points. • To split the chart by stage. one row for each stage. choose Each segment corresponds to a stage (if chart has stages). For help choosing a lambda value. To display subgroups 1 2 In the control chart dialog box. 3 Options − Storage Stat > Control Charts > control chart type > control chart > control chart Options > Storage Charts Individuals.Control Charts • • Choose Optimal lambda for each stage (in a chart with stages) to have Minitab search for optimal values for each stage. then enter one or more values. and enter the number of subgroups for each segment. under Subgroups to display. 2003 Minitab Inc. click control chart Options. Each segment contains __ subgroups: Choose to split the control chart into subgroups of a specified number of points. > control chart Options > Display Charts All charts except Z-MR Use to display the control chart by stage or by number of plotted points and to display the test results in the Session window. Display last __ subgroups: Choose to display a specific number of subgroups. 3 Click OK. I-MR Use to store your choice of statistics in the worksheet. Options − Display . 3 To split a control chart 1 2 In the control chart dialog box. click control chart Options. 111 . Subgroups to display Display all subgroups: Choose to display all subgroups. Click OK. under Subgroups to display. Dialog box items Store these estimates for each chart Means: Check to store estimates of the means. Click the Display tab. Split chart into series of segments for display purposes Do not split: Choose not to split the chart. and enter the number of subgroups. Test results Display test results in Session window: Check to display the test results in the Session window. Click OK. choose Each segment contains __ subgroups. and enter the number of subgroups for each segment. Do one of the following: • To display all subgroups in the control chart. Do one of the following: • To split the chart into a specific number of subgroups. one row for each plotted point. • To display a specific number of the last subgroups. Choose Other (enter value(s) between -5 and 5) to transform the data using another lambda value or values. Click the Display tab.. choose Display all subgroups. Each segment corresponds to a stage (if chart has stages): Choose to split the control chart by stage or stage when you have specified stages in control chart Options > Stage.. see the independent Box-Cox transformation command Stat > Control Charts > Box-Cox Transformation. one row for each stage.

112 2003 Minitab Inc. . 32. 44. 45 TEST 6. Test results: Check to store the results of any test performed. 30. one row for each plotted point. 45 TEST 2. 21. Test Failed at points: 14. then click OK. 31.Quality Control Center line value: Check to store the center line value. One point more than 3. Session window output I Chart of Weight Test Results for I Chart of Weight TEST 1. one row for each plotted point. Click I Chart Options. 17. Test Failed at points: 9.00 standard deviations from center line. 4 out of 5 points more than 1 standard deviation from center line (on one side of CL). 16. 29. 34. 45 * WARNING * If graph is updated with new data. 9 points in a row on same side of center line. Test Failed at points: 5. Choose Stat > Control Charts > Variables Charts for Individuals > Individuals. 1 2 3 4 5 Open the worksheet EXH_QC. Weight contains the weight in pounds of each batch of raw material. Test Failed at points: 24. In Variables. Choose Perform all tests for special causes. 13. 20. 31. 12. To store statistics in the worksheet 1 2 Choose Stat > Control Charts > control chart type > control chart > control chart Options > Storage. 23. then click OK in each dialog box. Control limit values: Check to store the control limit values. 10. one row for each plotted point. the results above may no * longer be correct. 15. Minitab stores one column for the lower control limits and one column for the upper control limits.MTW. 14. then click the Tests tab. 30. one row for each plotted point. 19. 6. 30. 2 out of 3 points more than 2 standard deviations from center line (on one side of CL). 18. 36 TEST 5. 33. 7. Stage: Check to store the stage. 11. 31. enter Weight. Example of Individuals chart In the following example. 35. Check the statistics to store in the worksheet.

you begin monitoring admission times. In Define stage (historical groups) with this variable. you discover that you can cut down on switchover time by using the same type of IV line used in the operating room. You find that the process is in control. Before making any changes in the process. enter Month. 2003 Minitab Inc. This standardization takes place in July.MTW. Example of a historical chart As manager of a hospital's intensive care unit. you display a historical chart. 1 2 3 4 5 6 Open the worksheet ICU.Control Charts Graph window output Interpreting the results The individuals chart shows six points outside the control limits and 22 points inside the control limits exhibiting a nonrandom pattern. To gain an understanding of the process. 113 . suggesting the presence of special causes. but the variability is large. To share your findings with the staff. While studying the admissions process. Click OK in each dialog box. enter ICUadmit. See the Session window results for a list of the points that failed each test. then click the Stages tab. Choose Stat > Control Charts > Variables Charts for Individuals > Individuals. your team decides to first standardize the admission procedure for all shifts. you are concerned about the length of time it takes to admit patients to your unit. In Variables. You implement this change in August. Click I Chart Options.

The moving range is of length 2. and multivariate charts also plot measurement data. Choosing a variables control chart for individuals Minitab offers five variables control charts: • • • • I-MR − an Individuals chart and Moving Range chart displayed in one window Z-MR − a chart of standardized individual observations and moving ranges from short run processes Individuals − a chart of individual observations Moving Range − a chart of moving ranges Moving Range Chart Stat > Control Charts > Variables Charts for Individuals > Moving Range A moving range chart is a chart of "moving ranges" − ranges calculated from artificial subgroups created from successive observations. time-weighted charts.Quality Control Graph window output Interpreting the results The data in the first part of the Individuals chart are the admission times (in minutes) before any improvements were made. the average of the moving range divided by an unbiasing constant. Attributes control charts plot count data. improvements to standardized procedure further reduced both mean admission time and variation. such as length or pressure. or enter a historical value for σ. see Control Charts Overview and Variables Control Charts for Individuals Overview. Variables control charts for subgroups. You can also estimate σ using the median of the moving range. As you can see. such as the number of defects or defective units. For more information about control charts. Moving Range chart estimates the process variation. σ. the initial standardization in July reduced both the mean admission time and the variation in admission time. for individuals data. with MRbar / d2. change the length of the moving range. By default. You can use moving range charts to track the process variation and detect the presence of special causes when your sample size is 1. . since consecutive values have the greatest chance of being alike. In August. Moving Range Chart Variables Control Charts for Individuals Overview data Variables control charts for individuals plot statistics from measurement data. For more information. 114 2003 Minitab Inc. see Control Charts Overview.

Subgroups are equal size Univariate (one variable) Multivariate (more than one variable) Down columns or across rows Down columns Subgroups are unequal size Down columns with subgroup indicator column Down columns with subgroup indicator column Structure subgroup data down a column or across rows. and so on. Note that the first five observations in the left data set (subgroup 1) are the first row of the right-side data set. Minitab calculates the moving ranges of length K for all nonmissing data and plots them on the chart. K is determined by the moving range span specified in Length of moving range in MR Options > Estimate. 115 . In the following example.Control Charts Dialog box items Variables: Choose one or more columns. Here is the same data set. then create a second column of subscripts which serve as subgroup indicators. with one column for each variable. the second 5 observations are the second row. When subgroups are of unequal size. C1 contains the process data and C2 contains subgroup indicators: 2003 Minitab Inc. structured both ways. <Scale> <Labels> <Multiple Graphs> <Data Options> <MR Options> Data − Variables Control Chart Organize the data for all variables control charts in the same way. Multivariate data must be entered down columns. Variables charts include: • • • • Variables charts for subgroups Variables charts for individuals Time-weighted charts Multivariate charts Structure your data for these charts using the guidelines below. with subgroups of size 5. you must enter your data in one column. Worksheet Structure Structure your data down a column or across rows. using the following table as a guide.

you may want to use the Box-Cox transformation to induce normality. subgroup 2 has six observations. For information on the Box-Cox transformation option. > control chart Options > Estimate Charts Moving Range. Missing data See Missing data in control charts for information on how to handle missing data for different types of control charts. CUSUM Enters historical data for estimating σ. Use the stand alone command as an exploratory tool to help you determine the best lambda value for the transformation. You can access the Box-Cox transformation two ways: by using the Box-Cox transformation option provided with the control chart commands.. you must enter data that approximate a normal distribution. In Variables. Dialog box items Standard deviation: Enter the historical standard deviations as constants. then click OK. or by using the stand alone Box-Cox command. You can omit or include certain subgroups when estimating parameters using Estimate. Each stage for each variable uses a different value. subgroup 1 has three observations.Quality Control Each time a subscript changes in C2. For example. All stages for each variable use the same value.. > control chart Options > Parameters Charts R. see Options − Box-Cox. enter a data column or columns. Minitab estimates it from the data. If you like. For information on the stand alone Box-Cox transformation command. Nonnormal data To properly interpret Minitab's control charts. if you have a known process parameter or an estimate obtained from past data. . see Box-Cox Transformation. A single value. Click OK. I-MR 116 2003 Minitab Inc. use any dialog box options. Options − Parameters . If you do not specify a value for σ. click control chart Options. A list of values. To use historical standard deviation values 1 2 3 • • • In the control chart dialog box. and so on. In Standard deviation... a new subgroup begins in C1. If the data are highly skewed. which applies to all variables and stages. S. You can enter historical data in several ways: Options − Estimate . A list of values. you may enter this value or values. one for each stage within each variable. In this example. one for each variable. you can use the transformation option to transform the data at the same time you draw the control chart. Then. Moving Range. enter one or more values. To display a moving range chart 1 2 3 Choose Stat > Control Charts > Variables Charts for Individuals > Moving Range.

a horizontal line labeled UB is drawn at the upper bound instead. • Choose Use the following samples when estimating parameters and enter the subgroups or observation numbers you want to use for the calculations. Method for estimating standard deviation: See [2] for a discussion of the relative merits of the methods listed below. then enter the number of observations. Minitab assumes the values you enter are subgroup numbers. 117 . Median moving range: Choose to estimate σ using the median of the moving range. To omit or use subgroups for the estimates of parameters By default. except with the Individuals.Control Charts Omits or includes certain observations to estimate µ and σ. a horizontal line labeled LB is drawn at the lower bound instead. For example. Note S Limits . and use biased estimates of σ. Minitab only uses the non-omitted data to find lambda. then decide to omit subgroup 2 from the parameter estimation (using control chart Options > Estimate). Moving Range. see Example of an X-bar chart with tests and customized control limits. To choose how the standard deviation is estimated 1 2 In the control chart dialog box. You can also select one of two methods to estimate σ. Similarly. if the calculated lower control limit is less than the lower bound. The span must be < 100. click control chart Options. Note 3 Click OK. Tip You can also modify the control limits using Minitab's graph editing features. Click the Estimate tab. But you may want to use or omit certain data if they show abnormal behavior. When the calculated upper control limit is greater than the upper bound. Under Estimating standard deviation. click control chart Options. When you use Omit the following samples when estimating parameters with Length of moving range. With these charts.. Length of moving range: Check to calculate the moving range. Minitab excludes any moving ranges with omitted data from the calculations. you can prevent them from influencing the estimation of µ and σ. choose a method. Click the Estimate tab. then enter individual observation numbers. Minitab uses a span of two because consecutive values have the greatest chance of being alike.. Do one of the following: • Choose Omit the following samples when estimating parameters and enter the subgroups or observation numbers you want to omit from the calculations. then enter individual observation numbers. Use the following samples when estimating parameters: Choose to specify observations to compute µ and σ. you exclude subgroup 1 from your analysis (using control chart > Data Options > Subset). Caution When you include or exclude rows using control chart > Data Options > Subset and also omit or use samples to estimate parameters. if some observations are erratic due to assignable causes that you have corrected. You can set preferences for the estimation of σ and the length of the moving range using Tools > Options > Control Charts and Quality Tools > Estimating Standard Deviation. For an example. Then click OK. Subgroup size = 1 Average moving range: Choose to estimate σ using the average of the moving range. Because Minitab considers an omitted sample (subgroup 1) as part of the entire data set. > control chart Options > S Limits Charts Individuals. When you include or exclude rows using control chart Options > Estimate and choose to perform a Box-Cox transformation. For example. the values are interpreted as observation numbers. not just the subsetted data. You can set bounds on the upper and/or lower control limits. 1 2 In the control chart dialog box. Minitab estimates the process parameters from all the data. By default. the omitted or used samples apply to the entire data set. Moving Range Allows you to draw control limits above and below the mean at the multiples of any standard deviation. Dialog box items Omit the following samples when estimating parameters: Choose to specify observations to omit from the computations for µ and σ. change the length of the moving range. and I-MR charts. you must enter subgroup 2 as the subgroup to omit. 2003 Minitab Inc.

Minitab marks it with the test number on the plot. NP. you can perform the 4 tests for special causes. the test results are not updated in the Session window. If the data change and the control chart is live-linked. See Tools > Options > Control Charts and Quality Tools > Other.. Place bounds on control limits Lower standard deviation limit bound: Check to set a lower bound for the control limits. S. enter one or more values in These multiples of the standard deviation. a horizontal line labeled LB will be drawn at the lower bound instead. See Updating Graphs for more information. Each value is the number of standard deviations the lines should be drawn at. Click the S Limits tab. above. C With any of the attributes control charts. Minitab marks it by the lowest numbered test. one that should be investigated. U. The occurrence of a pattern suggests a special cause for the variation. The test results are displayed in the Session window. a pair of control limits is drawn for each value in the list. P. click control chart Options. When a point fails a test. > control chart Options > Tests Charts R. If you specify more than one value. You can choose to change the color of a control chart when the last row of data causes a new test failure for any point. If the calculated lower control limit is less than the lower bound. Minitab will perform tests when sample sizes are unequal. a horizontal line labeled UB will be drawn at the upper bound instead. The values entered are the number of standard deviations above and below the center line. Do any of the following: • To specify where control limits are drawn: under Display control limits at. Each value represents the number of standard deviations below and above the mean. 3 Options − Tests . Set your preferences for the sensitivity of tests with Tools > Options > Control Charts and Quality Tools > Define Tests. If the calculated upper control limit is greater than the upper bound. • To set bounds on the control limits: check Lower standard deviation limit bound (and/or Upper standard deviation limit bound) and enter a value. Each test. If a point fails more than one test. .. and below the mean. To customize the control limits 1 2 In the control chart dialog box.Quality Control Dialog box items Display control limits at These multiples of the standard deviation: Enter one or more values. as shown below. See [1] and [10] for guidance on using these tests. Moving Range. Click OK in each dialog box. Test 1 1 point more than 3σ from center line Test 2 9 points in a row on same side of center line 118 2003 Minitab Inc. detects a specific pattern in the plotted data. Upper standard deviation limit bound: Check to set an upper bound for the control limits.

and X-bar and S. click control chart Options. • To select certain tests. then check any combination of the tests listed below. Click the Stages tab. X-bar and R. To perform the tests for special causes 1 2 Choose Stat > Control Charts > control chart type > control chart > control chart Options > Tests. or text. Minitab treats each repeated value as a separate occurrence. S.. 2003 Minitab Inc. when subgroups are across rows). choose Perform no tests. dates. • To perform no tests. alternating up and down Dialog box items Perform all tests for special causes: Choose to perform all tests. Do one of the following: • To do all tests. You can enter the same value more than once. Perform the following tests for special causes: Choose to specify a subset of the four tests. The indicators can be numbers. alternating up and down Perform no tests: Choose to perform no tests for special causes. you must set up a column of grouping indicators. 1 point > 3 standard deviations from center line 9 points in a row on same side of center line 6 points in a row. choose Perform all tests for special causes. When to start a new stage With each new value: Choose to start a new stage each time the value in the column changes. then enter the values. Note With the following charts. Enclose date/time or text entries in double quotes. When executing the command. Historical charts are particularly useful for comparing data before and after a process improvement. With the first occurrence of these values: Choose to start a new stage at the first occurrence of a certain value. Dialog box items Define stages (historical groups) with this variable: Enter the column that contains the stage indicators.Control Charts Test 3 6 points in a row. you can tell Minitab to start a new stage in one of two ways: • • 1 2 Each time the value in the column changes At the first occurrence of one or more values In the control chart dialog box. 3 Options − Stages . all increasing or all decreasing Test 4 14 points in a row. > control chart Options > Stages Charts All charts except Z-MR You can display stages in your process by drawing a "historical chart" − a control chart in which the control limits and center line are estimated independently for different groups in your data. choose Perform the following tests for special causes and check the tests to perform.. you must have at least one subgroup with two or more observations: R. The column must be the same length as the data column (or columns. Click OK. all increasing or all decreasing 14 points in a row. To display a historical chart To define stages in your process. 119 .

) To use this option. You may want to first run the command Stat > Control Charts > Box-Cox Transformation to help you find the optimal transformation value. Minitab does not accept any values you enter in the Parameters tab for historical means or historical standard deviations. You can also choose any value between − 5 and 5 for λ. enter the column which contains the stage indicators. Minitab only uses the non-omitted data to find lambda.. Options − Box-Cox . Lambda = 0 (natural log): Choose to use the natural log of the data. then enter one or more values. Enclose date/time or text entries in double quotes. In most cases. To start a new stage at the first occurrence of a certain value. click control chart Options. then click OK. use any dialog box options. Enter the values. You can enter the same value more than once. . • Choose Optimal lambda for each stage (in a chart with stages) to have Minitab search for optimal values for each stage. 120 2003 Minitab Inc. unless λ = 0. • Choose Lambda=0. do one of the following: • • To start a new stage each time the value in the column changes. For help choosing a lambda value. Dialog box items Box-Cox power transformation (W = Y**Lambda): Choose to use when your data are very skewed or when the withinsubgroup variation is unstable. Use overall standard deviation: Check to use the overall standard deviation in the estimation of the optimal lambda." The transformation takes the original data to the power λ.5 (square root) to use the square root of the data. do not select the Box-Cox option when you make a control char. 3 Click OK. then do one of the following: • Choose Lambda = 0 (natural log) to use the natural log of the data.5 (square root): Choose to use the square root of the data. choose With each new value. Optimal lambda for each stage (in a chart with stages): Choose to have Minitab search for optimal values for each stage. you should not choose a λ outside the range of −2 and 2. Minitab treats each repeated value as a separate occurrence.. Caution If you use Stat > Control Charts > Box-Cox Transformation to find the optimal λ value and choose to store the transformed data with that command. (λ is lambda. To do the Box-Cox power transformation 1 2 In the control chart dialog box. • Choose Optimal lambda to have Minitab search for an optimal value for all stages. choose With the first occurrence of these values. 5 If you like. Under When to start a new value. in which case the natural log is taken. The Options subdialog box lists the common transformations natural log (λ = 0) and square root (λ = 0. see the independent Box-Cox transformation command Stat > Control Charts > Box-Cox Transformation. Click the Box-Cox tab. then enter the lambda values. doing so will double-transform the data.Quality Control 3 4 In Define stages (historical groups) with this variable.5). Check Box-Cox power transformation (W = Y**Lambda). > control chart Options > Box-Cox Charts All except the Attributes charts You can use the Box-Cox power transformation when your data are very skewed or when the within-subgroup variation is unstable to make the data "more normal. • Choose Other (enter value(s) between -5 and 5) to transform the data using another lambda value or values. When you include or exclude rows using control chart Options > Estimate. Lambda=0. When you use this transformation. Other (enter value(s) between -5 and 5): Choose to transform the data using another lambda value or values. Optimal lambda: Choose to have Minitab search for an optimal value. the data must be positive.

Click the Display tab. Dialog box items Store these estimates for each chart Standard deviations: Check to store the estimates of the standard deviations. one row for each plotted point. and enter the number of subgroups. 3 To split a control chart 1 2 In the control chart dialog box. and enter the number of subgroups for each segment. one row for each plotted point. Do one of the following: • To split the chart into a specific number of subgroups. Click the Display tab. Test results: Check to store the results of any test performed. choose Each segment contains __ subgroups. then click OK. Each segment corresponds to a stage (if chart has stages): Choose to split the control chart by stage or stage when you have specified stages in control chart Options > Stage. 121 . one row for each plotted point. 2003 Minitab Inc. Center line value: Check to store the center line value. one row for each plotted point. Control limit values: Check to store the control limit values. choose Each segment corresponds to a stage (if chart has stages). Minitab stores one column for the lower control limit and one column for the upper control limit. Stage: Check to store the stage. click control chart Options.Control Charts Options − Display . To display subgroups 1 2 In the control chart dialog box. Click OK. > control chart Options > Display Charts All charts except Z-MR Use to display the control chart by stage or by number of plotted points and to display the test results in the Session window. under Subgroups to display. Display last __ subgroups: Choose to display a specific number of subgroups. Do one of the following: • To display all subgroups in the control chart. Subgroups to display Display all subgroups: Choose to display all subgroups.. To store statistics in the worksheet 1 2 Choose Stat > Control Charts > control chart type > control chart > control chart Options > Storage. Each segment contains __ subgroups: Choose to split the control chart into subgroups of a specified number of points. 3 Moving Range Chart − Options − Storage Stat > Control Charts > Variables Charts for Individuals > Moving Range > MR Options > Storage Use to store your choice of statistics in the worksheet. and enter the number of subgroups for each segment. Click OK. Check the statistics to store in the worksheet. Test results Display test results in Session window: Check to display the test results in the Session window. • To display a specific number of the last subgroups. one row for each plotted point. one row for each stage. Store these values for each point Point plotted: Check to store the plotted points. Split chart into series of segments for display purposes Do not split: Choose not to split the chart. • To split the chart by stage. choose Display all subgroups. click control chart Options.. under Subgroups to display. choose Display last __ subgroups and enter the number of subgroups to display.

Test Failed at points: 83 * WARNING * If graph is updated with new data. Test Failed at points: 34. Under Scales for Different Variables. a camshaft. . Click OK. Choose Stat > Control Charts > Variables Charts for Individuals > Moving Range. 122 2003 Minitab Inc. the results above may no * longer be correct. which causes poorfitting assemblies.Quality Control Example of moving range chart You work at an automobile engine assembly plant. must be 600 mm +2 mm long to meet engineering specifications. Your supervisor would also like you to create moving range charts for each supplier to track process variation. 1 2 3 4 5 Open the worksheet CAMSHAFT. check Same Y. the results above may no * longer be correct.00 standard deviations from center line. One point more than 3. Session window output Moving Range Chart of Supp1 Test Results for MR Chart of Supp1 TEST 1. 56 * WARNING * If graph is updated with new data. One point more than 3. You have already made an X-bar and R chart for supplier 1. Click Multiple Graphs. One of the parts. A chronic problem exists with camshaft length being out of specification. resulting in high scrap and rework rates. In Variables. You want to make the yscales of the charts the same so that the charts are easier to compare.MTW. You collect a total of 100 observations (20 samples of 5 camshafts each) from each of your suppliers. enter Supp1 Supp2.00 standard deviations from center line. Moving Range Chart of Supp2 Test Results for MR Chart of Supp2 TEST 1.

Products may also be classified by their number of defects. For the most part. 123 . while the moving range chart for Supplier 2 shows one point outside the control limits. products may be compared against a standard and classified as either being defective or not. For instance. 2003 Minitab Inc.Control Charts Graph window output Interpreting the results The moving range chart for supplier 1 shows two points outside the control limits. except that they plot statistics from count data rather than measurement data. Attributes Charts Overview Attributes Control Charts Overview Attributes control charts are similar in structure to variables control charts. the process variation for both suppliers is within the control limits.

Use historical values of mu With C chart and U chart. such as the number of defects. [8]. which draws a chart of the proportion of defectives in each subgroup Stat > Control Charts > Attributes Charts > NP. Use C Chart when the subgroup size is constant. The control charts for defectives are: • • P Chart. Minitab estimates it from the data. [3]. which draws a chart of the number of defects per unit sampled in each subgroup Suppose you want to count the number of flaws on the inner surface of a television screen. an estimate obtained from past data. a process statistic. 7th Edition. For example. If p is not specified. Grant and R. Alternatively. . by default. and [10] for a discussion of these charts. or a goal. Control Charts for Defects Overview You can create control charts for defects with: • • Stat > Control Charts > Attributes Charts > C. if you were counting the number of flaws on the inner surface of a television screen. or a goal. If µ is not specified. Control charts for defectives You can compare a product to a standard and classify it as being defective or not. you can enter a known process mean. p. 124 2003 Minitab Inc. For example. C Chart would chart the actual number of flaws. Southfield. Statistical Quality Control Methods. The control charts for defects are: • • C Chart.W. [8]. McGraw-Hill. and [10] for a discussion of these charts. an estimate obtained from past data. you can enter an actual known process proportion. The parameters of the binomial distribution are the sample size.Quality Control As with variables control charts. I. Statistical Quality Control. Asian Productivity Organization. [8]. [9]. you assume the process produces data from a population that follows a Poisson distribution. which charts the proportion of defectives in each subgroup NP Chart. U Chart. Control Charts for Defectives Overview You can create control charts for defectives with: • • Stat > Control Charts > Attributes Charts > P. Inc. which draws a chart of the number of defectives in each subgroup See [2]. [3]. which draws a chart of the number of defects in each sample Stat > Control Charts > Attributes Charts > U. Alternatively. you can assume the process produces data from a population that follows a binomial distribution. while U Chart would chart the number of flaws per square inch sampled. Minitab also draws two other lines − the upper and lower control limits − 3σ above and below the center line. n. [9]. and [10] for a discussion of these charts.L. MI. a defect does not always result in a defective product. Use U Chart when the subgroup size varies. Minitab estimates it from the data. References − Attributes Control Charts [1] [2] [3] [4] Automotive Industry Action Group (1991). while U charts the number of flaws per square inch sampled. a length of wire either meets the strength requirements or not. you might count the number of scratches on the surface of an appliance.S. K. [9]. Classifying a product by its number of defects is sometimes more convenient. is plotted versus a sample number or time. Burr (1976). C charts the actual number of flaws. [6]. AIAG. See [1]. Minitab draws a center line at the average of the statistic being plotted for the time being charted. [6]. [8]. which charts the number of defects per unit sampled in each subgroup. Marcel Dekker. [9]. Ishikawa (1967). Fundamental Statistical Process Control. which charts the number of defectives in each subgroup See [1]. Use historical values of p With P chart and NP chart. [3]. [3]. See [2]. For example. [6]. E. Guide to Quality Control. The mean and variance of this distribution are both denoted by µ. which charts the number of defects in each subgroup. and [10] for a discussion of these charts. [6]. and the proportion of defectives. Leavenworth (1996). Control charts for defects When a product is complex.

See [1]. Control charts for defectives You can compare a product to a standard and classify it as being defective or not. [10] Western Electric (1956). [9].M. Classifying a product by its number of defects is sometimes more convenient. a defect does not always result in a defective product. 2nd edition. a length of wire either meets the strength requirements or not.S. John Wiley & Sons. Wadsworth. For more information. Dialog box items Variables: Enter columns that contain the number of defectives for each sample. Use C Chart when the subgroup size is constant.C. Use U Chart when the subgroup size varies. you might count the number of scratches on the surface of an appliance. except that they plot statistics from count data rather than measurement data. a process statistic. D. [9]. Inc. Introduction to Statistical Quality Control. Minitab draws a center line at the average of the statistic being plotted for the time being charted. L. and A. P Chart Attributes Control Charts Overview Attributes control charts are similar in structure to variables control charts. Each entry in the worksheet column is the number of defectives for one subgroup. Ryan (2000). Stephens. Western Electric Corporation. [3]. Godfrey. John Wiley & Sons. For example. and [10] for a discussion of these charts. which charts the proportion of defectives in each subgroup NP Chart. For example. [6]. while U Chart would chart the number of flaws per square inch sampled.P. each control limit is not a single straight line but varies with the subgroup size. you may want to force the control limits to be constant by specifying a fixed subgroup size using P Chart Options > Estimate. If the subgroups are not equal. Subgroup sizes: Enter a number or a column of subscripts. [8]. by default. Products may also be classified by their number of defects. Nelson (1984). T. As with variables control charts. Control charts for defects When a product is complex. which charts the number of defectives in each subgroup See [1]. [6]. such as the number of defects. Statistical Quality Control Handbook.B. The control charts for defects are: • • C Chart. 2nd edition." Journal of Quality Technology. If the subgroup sizes do not vary much. 4th edition. see Control Charts Overview and Attributes Control Charts Overview. 4th edition. Defect Prevention: Use of Simple Statistical Tools. For instance. Statistical Methods for Quality Improvement. H. John Wiley & Sons. C Chart would chart the actual number of flaws. 125 . U Chart. The control charts for defectives are: • • P Chart. [3]. Minitab also draws two other lines − the upper and lower control limits − 3σ above and below the center line. (2001). For example. Modern Methods for Quality Control and Improvement. [8]. Kane (1989).S.E. products may be compared against a standard and classified as either being defective or not. Montgomery (2001). <Scale> <Labels> <Multiple Graphs> <Data Options> <P Chart Options> 2003 Minitab Inc. Marcel Dekker. P Chart Stat > Control Charts > Attributes Charts > P Tracks the proportion defective and detects the presence of special causes. if you were counting the number of flaws on the inner surface of a television screen. is plotted versus a sample number or time. assumed to have come from a binomial distribution with parameters n and p. 237−239. and [10] for a discussion of these charts. "The Shewhart Control Chart − Tests for Special Causes. which charts the number of defects in each subgroup. K. which charts the number of defects per unit sampled in each subgroup.Control Charts [5] [6] [7] [8] [9] V. 16.

subgroup sizes that differ affect both the control limits and the center line. from the overall sample proportion. When an observation is missing. enter columns containing the number of defectives. With NP Chart. In Subgroup size. > control chart Options > Parameters Charts P. enter the value(s).. For example. otherwise. You enter the number that failed to pass inspection in one column. You can omit or include certain subgroups when estimating parameters using Estimate. If you like. both numbers may vary. A single value. All stages for each variable use the same value. or defects per unit. By default. In general. then click OK. . click control chart Options. the control limits are a function of the subgroup size. Dialog box item Proportion: Enter the historical proportions as constants.. NP Chart. NP You can enter historical data for estimating p. p. enter the historical proportions as constants. NP Chart and C Chart plot raw data. To display a P chart 1 2 3 4 Choose Stat > Control Charts > Attributes Charts > P. Minitab calculates the center line and control limits using this value. Minitab estimates it from the data. enter a value for each historical group. A list of values. If you do not specify a value for p. and U Chart handle unequal-sized subgroups. You can enter historical data in several ways: 126 2003 Minitab Inc. You can force the control limits and center line to be constant. Suppose you have collected daily data on the number of parts that have been inspected and the number of parts that failed to pass inspection. one for each variable. P Chart.Quality Control Data Each entry in the worksheet column should contain the number of defectives or defects for a subgroup. In Variables. On any given day. With P Chart and U Chart. To use historical values of p 1 2 3 • • In the control chart data source dialog box. if you have a known process parameter or an estimate obtained from past data. Minitab estimates the process proportion defective. the total number inspected varies from day to day. you must also enter a corresponding column of subgroup sizes. In Proportion. all stages use that value. while the center line is always constant. a gap exists in the chart where the summary statistic for that subgroup would have been plotted. In this case. so you enter the subgroup size in another column: Failed 8 13 13 16 14 15 13 10 24 12 Inspect 968 1216 1004 1101 1076 995 1202 1028 1184 992 P Chart (and U Chart) divide the number of defectives or defects by the subgroup size to get the proportion of defectives. Options − Parameters . use any dialog box options. the control limits are further from the center line for smaller subgroups than they are for larger ones. Click OK in each dialog box. If you enter one constant. Then. which applies to all variables and stages. When subgroup sizes are unequal. enter a number or a column of subscripts.

Control Charts

• A list of values, one for each stage within each variable. Each stage for each variable uses a different value.

Options − Estimate

... > control chart Options > Estimate Charts P, NP You can omit or use certain subgroups to estimate p. For example, if some subgroups have erratic data due to assignable causes that you have corrected, you can prevent these subgroups from influencing the estimation of p. You can also force the control limits to be constant when you have unequal sample sizes. Caution When you include or exclude rows using control chart > Data Options > Subset and also omit or use samples to estimate parameters, the omitted or used samples apply to the entire data set, not just the subsetted data. For example, you exclude subgroup 1 from your analysis (using control chart > Data Options > Subset), then decide to omit subgroup 2 from the parameter estimation (using control chart Options > Estimate). Because Minitab considers an omitted sample (subgroup 1) as part of the entire data set, you must enter subgroup 2 as the subgroup to omit. Dialog box items Omit the following samples when estimating parameters (eg, 3 12:15): Choose to specify subgroups of the data to omit from the estimation of p, then type individual subgroup numbers or ranges of subgroups. Use the following samples when estimating parameters (eg, 3 12:15): Choose to specify subgroups of data to compute p, then type individual subgroup numbers or a range of subgroups.

**To omit or use subgroups for the estimates of parameters
**

By default, Minitab estimates the process parameters from all the data. But you may want to use or omit certain data if they show abnormal behavior. 1 2 In the control chart dialog box, click control chart Options. Click the Estimate tab. Do one of the following: • Choose Omit the following samples when estimating parameters and enter the subgroups or observation numbers you want to omit from the calculations. • Choose Use the following samples when estimating parameters and enter the subgroups or observation numbers you want to use for the calculations. Minitab assumes the values you enter are subgroup numbers, except with the Individuals, Moving Range, and I-MR charts. With these charts, the values are interpreted as observation numbers.

Note 3

Click OK.

S Limits

... > control chart Options > S Limits Charts X-bar, R, S, P, NP, U, Moving Average, EWMA Allows you to draw control limits above and below the mean at the multiples of any standard deviation. You can set bounds on the upper and/or lower control limits. When the calculated upper control limit is greater than the upper bound, a horizontal line labeled UB is drawn at the upper bound instead. Similarly, if the calculated lower control limit is less than the lower bound, a horizontal line labeled LB is drawn at the lower bound instead. You can also force the control limits to be constant when sample sizes are unequal. For an example, see Example of an X-bar chart with tests and customized control limits. Tip You can also modify the control limits using Minitab's graph editing features.

Dialog box items Display control limits at These multiples of the standard deviation: Enter one or more values. The values entered are the number of standard deviations above and below the center line. If you specify more than one value, a pair of control limits is drawn for each value in the list. Place bounds on control limits Lower standard deviation limit bound: Check to set a lower bound for the control limits. If the calculated lower control limit is less than the lower bound, a horizontal line labeled LB will be drawn at the lower bound instead. Upper standard deviation limit bound: Check to set an upper bound for the control limits. If the calculated upper control limit is greater than the upper bound, a horizontal line labeled UB will be drawn at the upper bound instead. When subgroup sizes are unequal, calculate control limits

2003 Minitab Inc.

127

Quality Control

Using actual sizes of the subgroups: Choose to use the subgroups defined in the main dialog box to estimate σ. Assuming all subgroups have size: Choose to estimate the control limits using a specified subgroup size, then type the subgroup size. When subgroup sizes are not equal, each control limit is not a single straight line, but varies with the subgroup size. If the sizes do not vary much, you may want to force the control limits to be constant by entering a sample size in this text box. Only the control limits are affected; the plotted data are not changed.

**To customize the control limits
**

1 2 In the control chart dialog box, click control chart Options. Click the S Limits tab. Do any of the following: • To specify where control limits are drawn: under Display control limits at, enter one or more values in These multiples of the standard deviation. Each value is the number of standard deviations the lines should be drawn at, above, and below the mean. • To set bounds on the control limits: check Lower standard deviation limit bound (and/or Upper standard deviation limit bound) and enter a value. Each value represents the number of standard deviations below and above the mean. Click OK in each dialog box.

3

**To force control limits and center line to be constant
**

1 2 In the control chart dialog box, click control chart Options. Click the S Limits tab. Under When subgroup sizes are unequal, calculate control limits, choose Assuming all subgroups have size, then enter a value. For example, enter a value of 6 to calculate the control limits and center line as if all subgroup sizes were 6. Click OK. You should force the control limits and center line to be constant only when the difference in size between the largest and smallest subgroups is no more than 25%. For example, suppose the largest subgroup is size 125 and the smallest is 100. You can use this method because the size difference is 25% (125 / 100 = 25%).

3

Note

Options − Tests

... > control chart Options > Tests Charts R, S, Moving Range, P, NP, U, C With any of the attributes control charts, you can perform the 4 tests for special causes. Each test, as shown below, detects a specific pattern in the plotted data. The occurrence of a pattern suggests a special cause for the variation, one that should be investigated. See [1] and [10] for guidance on using these tests. When a point fails a test, Minitab marks it with the test number on the plot. If a point fails more than one test, Minitab marks it by the lowest numbered test. You can choose to change the color of a control chart when the last row of data causes a new test failure for any point. See Tools > Options > Control Charts and Quality Tools > Other. Set your preferences for the sensitivity of tests with Tools > Options > Control Charts and Quality Tools > Define Tests. The test results are displayed in the Session window. If the data change and the control chart is live-linked, the test results are not updated in the Session window. See Updating Graphs for more information. Minitab will perform tests when sample sizes are unequal. Test 1 1 point more than 3σ from center line Test 2 9 points in a row on same side of center line

128

2003 Minitab Inc.

Control Charts

Test 3

6 points in a row, all increasing or all decreasing

Test 4

14 points in a row, alternating up and down

Dialog box items Perform all tests for special causes: Choose to perform all tests. Perform the following tests for special causes: Choose to specify a subset of the four tests, then check any combination of the tests listed below. 1 point > 3 standard deviations from center line 9 points in a row on same side of center line 6 points in a row, all increasing or all decreasing 14 points in a row, alternating up and down Perform no tests: Choose to perform no tests for special causes.

**To perform the tests for special causes
**

1 2 Choose Stat > Control Charts > control chart type > control chart > control chart Options > Tests. Do one of the following: • To do all tests, choose Perform all tests for special causes. • To select certain tests, choose Perform the following tests for special causes and check the tests to perform. • To perform no tests, choose Perform no tests. Click OK.

3

Options − Stages

... > control chart Options > Stages Charts All charts except Z-MR You can display stages in your process by drawing a "historical chart" − a control chart in which the control limits and center line are estimated independently for different groups in your data. Historical charts are particularly useful for comparing data before and after a process improvement. Note With the following charts, you must have at least one subgroup with two or more observations: R, S, X-bar and R, and X-bar and S.

Dialog box items Define stages (historical groups) with this variable: Enter the column that contains the stage indicators. When to start a new stage With each new value: Choose to start a new stage each time the value in the column changes. With the first occurrence of these values: Choose to start a new stage at the first occurrence of a certain value, then enter the values. Enclose date/time or text entries in double quotes. You can enter the same value more than once; Minitab treats each repeated value as a separate occurrence.

**To display a historical chart
**

To define stages in your process, you must set up a column of grouping indicators. The indicators can be numbers, dates, or text. When executing the command, you can tell Minitab to start a new stage in one of two ways: • • 1 2 Each time the value in the column changes At the first occurrence of one or more values In the control chart dialog box, click control chart Options. Click the Stages tab.

The column must be the same length as the data column (or columns, when subgroups are across rows).

2003 Minitab Inc.

129

Quality Control

3 4 In Define stages (historical groups) with this variable, enter the column which contains the stage indicators. Under When to start a new value, do one of the following: • • To start a new stage each time the value in the column changes, choose With each new value. To start a new stage at the first occurrence of a certain value, choose With the first occurrence of these values. Enter the values. Enclose date/time or text entries in double quotes. You can enter the same value more than once; Minitab treats each repeated value as a separate occurrence.

5

If you like, use any dialog box options, then click OK.

Options − Display

... > control chart Options > Display Charts All charts except Z-MR Use to display the control chart by stage or by number of plotted points and to display the test results in the Session window. Subgroups to display Display all subgroups: Choose to display all subgroups. Display last __ subgroups: Choose to display a specific number of subgroups, and enter the number of subgroups. Split chart into series of segments for display purposes Do not split: Choose not to split the chart. Each segment contains __ subgroups: Choose to split the control chart into subgroups of a specified number of points, and enter the number of subgroups for each segment. Each segment corresponds to a stage (if chart has stages): Choose to split the control chart by stage or stage when you have specified stages in control chart Options > Stage. Test results Display test results in Session window: Check to display the test results in the Session window.

**To display subgroups
**

1 2 In the control chart dialog box, click control chart Options. Click the Display tab. Do one of the following: • To display all subgroups in the control chart, under Subgroups to display, choose Display all subgroups. • To display a specific number of the last subgroups, under Subgroups to display, choose Display last __ subgroups and enter the number of subgroups to display. Click OK.

3

**To split a control chart
**

1 2 In the control chart dialog box, click control chart Options. Click the Display tab. Do one of the following: • To split the chart into a specific number of subgroups, choose Each segment contains __ subgroups, and enter the number of subgroups for each segment. • To split the chart by stage, choose Each segment corresponds to a stage (if chart has stages). Click OK.

3

Options − Storage

Stat > Control Charts > Attributes Charts > P > P Chart Options > Storage Stat > Control Charts > Attributes Charts > NP > NP Chart Options > Storage Use to store your choice of statistics in the worksheet. Dialog box items Store these estimates for each chart Proportions: Check to store the proportions, one row for each historical stage. Store these values for each point Point plotted: Check to store the plotted points, one row for each plotted point. Center line value: Check to store the center line value, one row for each plotted point.

130

2003 Minitab Inc.

Control Charts

Control limit values: Check to store the control limit values, one row for each plotted point. Minitab stores one column for the lower control limits and one column for the upper control limits. Stage: Check to store the stage, one row for each plotted point. Subgroup size: Check to store the subgroup sizes, one row for each plotted point. Test results: Check to store the results of any test performed, one row for each plotted point. Minitab assigns each plotted point one of three values: 1 (beyond upper control limit), − 1 (beyond lower control limit), 0 (in control).

**To store statistics in the worksheet
**

1 2 Choose Stat > Control Charts > control chart type > control chart > control chart Options > Storage. Check the statistics to store in the worksheet, then click OK.

**Example of P Chart with unequal subgroup sizes
**

Suppose you work in a plant that manufactures picture tubes for televisions. For each lot, you pull some of the tubes and do a visual inspection. If a tube has scratches on the inside, you reject it. If a lot has too many rejects, you do a 100% inspection on that lot. A P chart can define when you need to inspect the whole lot. 1 2 3 4 Open the worksheet EXH_QC.MTW. Choose Stat > Control Charts >Attributes Charts > P. In Variables, enter Rejects. In Subgroup sizes, enter Sampled. Click OK.

Session window output P Chart of Rejects

Test Results for P Chart of Rejects

TEST 1. One point more than 3.00 standard deviations from center line. Test Failed at points: 6 * WARNING * If graph is updated with new data, the results above may no * longer be correct.

Graph window output

2003 Minitab Inc.

131

Quality Control

Interpreting the results Sample 6 is outside the upper control limit. Consider inspecting the lot.

NP Chart

Attributes Control Charts Overview

Attributes control charts are similar in structure to variables control charts, except that they plot statistics from count data rather than measurement data. For instance, products may be compared against a standard and classified as either being defective or not. Products may also be classified by their number of defects. As with variables control charts, a process statistic, such as the number of defects, is plotted versus a sample number or time. Minitab draws a center line at the average of the statistic being plotted for the time being charted. Minitab also draws two other lines − the upper and lower control limits − 3σ above and below the center line, by default. Control charts for defectives You can compare a product to a standard and classify it as being defective or not. For example, a length of wire either meets the strength requirements or not. The control charts for defectives are: • • P Chart, which charts the proportion of defectives in each subgroup NP Chart, which charts the number of defectives in each subgroup

See [1], [3], [6], [8], [9], and [10] for a discussion of these charts. Control charts for defects When a product is complex, a defect does not always result in a defective product. Classifying a product by its number of defects is sometimes more convenient. For example, you might count the number of scratches on the surface of an appliance. The control charts for defects are: • • C Chart, which charts the number of defects in each subgroup. Use C Chart when the subgroup size is constant. U Chart, which charts the number of defects per unit sampled in each subgroup. Use U Chart when the subgroup size varies.

For example, if you were counting the number of flaws on the inner surface of a television screen, C Chart would chart the actual number of flaws, while U Chart would chart the number of flaws per square inch sampled. See [1], [3], [6], [8], [9], and [10] for a discussion of these charts.

NP Chart

Stat > Control Charts > Attributes Charts > NP Tracks the number of defectives and detects the presence of special causes. Each entry in the worksheet column is the number of defectives for one subgroup, assumed to have come from a binomial distribution with parameters n and p. For more information, see Control Charts Overview and Attributes Control Charts Overview. Dialog box items Variables: Enter columns that contain the number of defectives for each sample. Subgroup sizes: Enter a number or a column of subscripts. If the subgroups are not equal, each control limit is not a single straight line but varies with the subgroup size. If the subgroup sizes do not vary much, you may want to force the control limits to be constant by specifying a fixed subgroup size using NP Chart Options > Estimate. <Scale> <Labels> <Multiple Graphs> <Data Options> <NP Chart Options>

Data

Each entry in the worksheet column should contain the number of defectives or defects for a subgroup. When subgroup sizes are unequal, you must also enter a corresponding column of subgroup sizes. Suppose you have collected daily data on the number of parts that have been inspected and the number of parts that failed to pass inspection. On any given day, both numbers may vary. You enter the number that failed to pass inspection in one column. In this case, the total number inspected varies from day to day, so you enter the subgroup size in another column:

132

2003 Minitab Inc.

from the overall sample proportion.. For example. Minitab calculates the center line and control limits using this value. while the center line is always constant. In Variables. In Subgroup sizes.. if you have a known process parameter or an estimate obtained from past data. With NP Chart. Then. 133 . otherwise. the control limits are further from the center line for smaller subgroups than they are for larger ones. All stages for each variable use the same value. If you do not specify a value for p. one for each stage within each variable. With P Chart and U Chart. NP 2003 Minitab Inc. > control chart Options > Parameters Charts P. enter the value(s). Each stage for each variable uses a different value. or defects per unit. If you enter one constant. subgroup sizes that differ affect both the control limits and the center line. In general. NP Chart and C Chart plot raw data. all stages use that value. NP Chart. enter a number or a column of subscripts. A single value. P Chart. which applies to all variables and stages. NP You can enter historical data for estimating p. p. use any dialog box options. You can force the control limits and center line to be constant. If you like. click control chart Options. By default.Control Charts Failed 8 13 13 16 14 15 13 10 24 12 Inspect 968 1216 1004 1101 1076 995 1202 1028 1184 992 P Chart (and U Chart) divide the number of defectives or defects by the subgroup size to get the proportion of defectives. Options − Parameters . the control limits are a function of the subgroup size. enter a value for each historical group. Minitab estimates it from the data. enter the historical proportions as constants. and U Chart handle unequal-sized subgroups. To use historical values of p 1 2 3 • • • In the control chart data source dialog box. A list of values. one for each variable. Dialog box item Proportion: Enter the historical proportions as constants. To display a NP chart 1 2 3 4 Choose Stat > Control Charts > Attributes Charts > NP. Click OK in each dialog box. You can enter historical data in several ways: Options − Estimate .. You can omit or include certain subgroups when estimating parameters using Estimate. A list of values. Minitab estimates the process proportion defective.. then click OK. In Proportion. a gap exists in the chart where the summary statistic for that subgroup would have been plotted. enter columns containing the number of defectives. > control chart Options > Estimate Charts P. When an observation is missing.

When subgroup sizes are not equal. Caution When you include or exclude rows using control chart > Data Options > Subset and also omit or use samples to estimate parameters. EWMA Allows you to draw control limits above and below the mean at the multiples of any standard deviation. S Limits .. Click the Estimate tab. you may want to force the control limits to be constant by entering a sample size in this text box. Dialog box items Display control limits at These multiples of the standard deviation: Enter one or more values. a horizontal line labeled UB will be drawn at the upper bound instead. then type individual subgroup numbers or a range of subgroups. you can prevent these subgroups from influencing the estimation of p. For example. Place bounds on control limits Lower standard deviation limit bound: Check to set a lower bound for the control limits. If the sizes do not vary much. Assuming all subgroups have size: Choose to estimate the control limits using a specified subgroup size. • Choose Use the following samples when estimating parameters and enter the subgroups or observation numbers you want to use for the calculations. then type the subgroup size. and I-MR charts. if the calculated lower control limit is less than the lower bound. except with the Individuals. if some subgroups have erratic data due to assignable causes that you have corrected. see Example of an X-bar chart with tests and customized control limits. 3 12:15): Choose to specify subgroups of data to compute p. a horizontal line labeled LB will be drawn at the lower bound instead. you must enter subgroup 2 as the subgroup to omit. When the calculated upper control limit is greater than the upper bound. a horizontal line labeled LB is drawn at the lower bound instead. > control chart Options > S Limits Charts X-bar. but varies with the subgroup size. To omit or use subgroups for the estimates of parameters By default. You can set bounds on the upper and/or lower control limits. each control limit is not a single straight line. Minitab assumes the values you enter are subgroup numbers. S. With these charts. Dialog box items Omit the following samples when estimating parameters (eg. then type individual subgroup numbers or ranges of subgroups. If the calculated lower control limit is less than the lower bound. calculate control limits Using actual sizes of the subgroups: Choose to use the subgroups defined in the main dialog box to estimate σ. Moving Range. the plotted data are not changed. Do one of the following: • Choose Omit the following samples when estimating parameters and enter the subgroups or observation numbers you want to omit from the calculations. But you may want to use or omit certain data if they show abnormal behavior. Moving Average. P. not just the subsetted data. . the values are interpreted as observation numbers. 134 2003 Minitab Inc. the omitted or used samples apply to the entire data set.Quality Control You can omit or use certain subgroups to estimate p. then decide to omit subgroup 2 from the parameter estimation (using control chart Options > Estimate).. a pair of control limits is drawn for each value in the list. Similarly. Upper standard deviation limit bound: Check to set an upper bound for the control limits. If the calculated upper control limit is greater than the upper bound. a horizontal line labeled UB is drawn at the upper bound instead. Note 3 Click OK. Tip You can also modify the control limits using Minitab's graph editing features. You can also force the control limits to be constant when sample sizes are unequal. U. When subgroup sizes are unequal. NP. R. click control chart Options. Because Minitab considers an omitted sample (subgroup 1) as part of the entire data set. Use the following samples when estimating parameters (eg. you exclude subgroup 1 from your analysis (using control chart > Data Options > Subset). If you specify more than one value. The values entered are the number of standard deviations above and below the center line. For an example. 1 2 In the control chart dialog box. 3 12:15): Choose to specify subgroups of the data to omit from the estimation of p. For example. Only the control limits are affected. You can also force the control limits to be constant when you have unequal sample sizes. Minitab estimates the process parameters from all the data.

Minitab marks it by the lowest numbered test. The occurrence of a pattern suggests a special cause for the variation. you can perform the 4 tests for special causes. P. Each value represents the number of standard deviations below and above the mean. See Updating Graphs for more information. Set your preferences for the sensitivity of tests with Tools > Options > Control Charts and Quality Tools > Define Tests. calculate control limits. When a point fails a test. If a point fails more than one test. Click OK. If the data change and the control chart is live-linked. click control chart Options. You can use this method because the size difference is 25% (125 / 100 = 25%). See [1] and [10] for guidance on using these tests. 3 To force control limits and center line to be constant 1 2 In the control chart dialog box.. > control chart Options > Tests Charts R. Minitab will perform tests when sample sizes are unequal. above. the test results are not updated in the Session window. S. and below the mean. Each test. 3 Note Options − Tests . then enter a value. Moving Range. as shown below. choose Assuming all subgroups have size. detects a specific pattern in the plotted data. For example. enter one or more values in These multiples of the standard deviation. 135 . Click the S Limits tab. Each value is the number of standard deviations the lines should be drawn at. one that should be investigated. Click OK in each dialog box. See Tools > Options > Control Charts and Quality Tools > Other. C With any of the attributes control charts. You should force the control limits and center line to be constant only when the difference in size between the largest and smallest subgroups is no more than 25%. Minitab marks it with the test number on the plot. NP.Control Charts To customize the control limits 1 2 In the control chart dialog box.. You can choose to change the color of a control chart when the last row of data causes a new test failure for any point. For example. click control chart Options. Test 1 1 point more than 3σ from center line Test 2 9 points in a row on same side of center line 2003 Minitab Inc. Do any of the following: • To specify where control limits are drawn: under Display control limits at. The test results are displayed in the Session window. suppose the largest subgroup is size 125 and the smallest is 100. Click the S Limits tab. enter a value of 6 to calculate the control limits and center line as if all subgroup sizes were 6. U. • To set bounds on the control limits: check Lower standard deviation limit bound (and/or Upper standard deviation limit bound) and enter a value. Under When subgroup sizes are unequal.

To perform the tests for special causes 1 2 Choose Stat > Control Charts > control chart type > control chart > control chart Options > Tests. alternating up and down Dialog box items Perform all tests for special causes: Choose to perform all tests. Perform the following tests for special causes: Choose to specify a subset of the four tests. The indicators can be numbers. Enclose date/time or text entries in double quotes. 1 point > 3 standard deviations from center line 9 points in a row on same side of center line 6 points in a row. alternating up and down Perform no tests: Choose to perform no tests for special causes. Click the Stages tab.Quality Control Test 3 6 points in a row. you must have at least one subgroup with two or more observations: R. or text. Historical charts are particularly useful for comparing data before and after a process improvement. Note With the following charts. all increasing or all decreasing 14 points in a row. You can enter the same value more than once. 136 2003 Minitab Inc. Do one of the following: • To do all tests. 3 Options − Stages . click control chart Options. you can tell Minitab to start a new stage in one of two ways: • • 1 2 Each time the value in the column changes At the first occurrence of one or more values In the control chart dialog box. . then check any combination of the tests listed below.. > control chart Options > Stages Charts All charts except Z-MR You can display stages in your process by drawing a "historical chart" − a control chart in which the control limits and center line are estimated independently for different groups in your data. you must set up a column of grouping indicators. Dialog box items Define stages (historical groups) with this variable: Enter the column that contains the stage indicators. • To select certain tests.. choose Perform all tests for special causes. all increasing or all decreasing Test 4 14 points in a row. Click OK. dates. To display a historical chart To define stages in your process. then enter the values. With the first occurrence of these values: Choose to start a new stage at the first occurrence of a certain value. • To perform no tests. S. and X-bar and S. Minitab treats each repeated value as a separate occurrence. when subgroups are across rows). X-bar and R. The column must be the same length as the data column (or columns. When to start a new stage With each new value: Choose to start a new stage each time the value in the column changes. choose Perform no tests. When executing the command. choose Perform the following tests for special causes and check the tests to perform.

Enclose date/time or text entries in double quotes. • To split the chart by stage. Display last __ subgroups: Choose to display a specific number of subgroups. choose With each new value. Minitab treats each repeated value as a separate occurrence. and enter the number of subgroups for each segment. Click the Display tab. 3 Options − Storage Stat > Control Charts > Attributes Charts > P > P Chart Options > Storage Stat > Control Charts > Attributes Charts > NP > NP Chart Options > Storage Use to store your choice of statistics in the worksheet. choose Each segment corresponds to a stage (if chart has stages). 3 To split a control chart 1 2 In the control chart dialog box. under Subgroups to display. Each segment contains __ subgroups: Choose to split the control chart into subgroups of a specified number of points. choose With the first occurrence of these values. 137 . Dialog box items Store these estimates for each chart Proportions: Check to store the proportions. Do one of the following: • To display all subgroups in the control chart. click control chart Options. Split chart into series of segments for display purposes Do not split: Choose not to split the chart. use any dialog box options. 5 If you like. Under When to start a new value. choose Display all subgroups.. one row for each plotted point. Test results Display test results in Session window: Check to display the test results in the Session window. Subgroups to display Display all subgroups: Choose to display all subgroups.. choose Each segment contains __ subgroups. Each segment corresponds to a stage (if chart has stages): Choose to split the control chart by stage or stage when you have specified stages in control chart Options > Stage. Click OK. and enter the number of subgroups for each segment. click control chart Options. do one of the following: • • To start a new stage each time the value in the column changes. under Subgroups to display. Store these values for each point Point plotted: Check to store the plotted points. To display subgroups 1 2 In the control chart dialog box. • To display a specific number of the last subgroups. You can enter the same value more than once. Click OK. 2003 Minitab Inc. > control chart Options > Display Charts All charts except Z-MR Use to display the control chart by stage or by number of plotted points and to display the test results in the Session window. enter the column which contains the stage indicators. Options − Display . then click OK. one row for each plotted point. and enter the number of subgroups. choose Display last __ subgroups and enter the number of subgroups to display. To start a new stage at the first occurrence of a certain value. Click the Display tab. Enter the values. Do one of the following: • To split the chart into a specific number of subgroups. one row for each historical stage. Center line value: Check to store the center line value.Control Charts 3 4 In Define stages (historical groups) with this variable.

One point more than 3. the results above may no * longer be correct. enter Rejects. Click OK in each dialog box. choose Each segment contains __ subgroups and enter10.MTW. one row for each plotted point. Under Split chart into a series of segments for display purposes. one row for each plotted point. enter Inspected. . To make the NP chart easier to present at the next staff meeting. Example of NP chart You work in a toy manufacturing company and your job is to inspect the number of defective bicycle tires. In Subgroup sizes. you decide to split the chart by every 10 inspection lots. Minitab stores one column for the lower control limits and one column for the upper control limits. 1 2 3 4 5 6 7 Open the worksheet TOYS. Test Failed at points: 9. then click the Display tab. 20 * WARNING * If graph is updated with new data. one row for each plotted point. 0 (in control). 138 2003 Minitab Inc.00 standard deviations from center line.Quality Control Control limit values: Check to store the control limit values. Check the statistics to store in the worksheet. You inspect 200 samples in each lot and then decide to create an NP chart to monitor the number of defectives. Minitab assigns each plotted point one of three values: 1 (beyond upper control limit). Click NP Chart Options. then click OK. Choose Stat > Control Charts > Attributes Charts > NP. To store statistics in the worksheet 1 2 Choose Stat > Control Charts > control chart type > control chart > control chart Options > Storage. Session window output NP Chart of Rejects Test Results for NP Chart of Rejects TEST 1. Subgroup size: Check to store the subgroup sizes. In Variables. Stage: Check to store the stage. one row for each plotted point. Test results: Check to store the results of any test performed. − 1 (beyond lower control limit).

As with variables control charts. For instance. C Chart would chart the actual number of flaws. Control charts for defects When a product is complex. The control charts for defects are: • • C Chart. See [1]. except that they plot statistics from count data rather than measurement data. a defect does not always result in a defective product. For example. 139 . [3]. [8]. [9]. Classifying a product by its number of defects is sometimes more convenient. is plotted versus a sample number or time. and [10] for a discussion of these charts. [6]. [9]. a length of wire either meets the strength requirements or not. if you were counting the number of flaws on the inner surface of a television screen. by default. For example.Control Charts Graph window output Interpreting the results Inspection lots 9 and 20 fall above the upper control limit. Minitab draws a center line at the average of the statistic being plotted for the time being charted. Use U Chart when the subgroup size varies. which charts the number of defects in each subgroup. Products may also be classified by their number of defects. 2003 Minitab Inc. which charts the number of defects per unit sampled in each subgroup. Minitab also draws two other lines − the upper and lower control limits − 3σ above and below the center line. and [10] for a discussion of these charts. indicating that special causes may have affected the number of defectives for these lots. a process statistic. [6]. Control charts for defectives You can compare a product to a standard and classify it as being defective or not. products may be compared against a standard and classified as either being defective or not. C Chart Attributes Control Charts Overview Attributes control charts are similar in structure to variables control charts. U Chart. Use C Chart when the subgroup size is constant. you might count the number of scratches on the surface of an appliance. The control charts for defectives are: • • P Chart. [8]. [3]. while U Chart would chart the number of flaws per square inch sampled. For example. which charts the number of defectives in each subgroup See [1]. You should investigate what special causes may have influenced the out-of-control number of bicycle tire defectives for inspection lots 9 and 20. which charts the proportion of defectives in each subgroup NP Chart. such as the number of defects.

enter columns containing the number of defects. see Control Charts Overview and Attributes Control Charts Overview. enter a value for each stage. By default. then click OK. one for each variable. In U Chart. you simply enter the number of defects for each sample into a data column and specify that in the command. . one for each stage within each variable. the process average number of defects. otherwise. To use historical values for the mean 1 2 3 • • • In the control chart dialog box. If all subgroups are the same size. Dialog box item Mean: Enter the historical means as constants. You can omit or include certain subgroups when estimating parameters using Estimate. This value is both the mean and the variance. <Scale> <Labels> <Multiple Graphs> <Data Options> <C Chart Options> To display a C chart 1 2 3 Choose Stat > Control Charts > Attributes Charts > C. In Variables. Minitab also uses this value to calculate the control limits. m. which represents the number of defects. Click OK in each dialog box. if you have a known process parameter or estimates obtained from past data. assumed to have come from a Poisson distribution with parameter m. you can enter the values. For more information. For C Chart. A single value. All stages for each variable use the same value. If you like. See P Chart and NP Chart Data for data entry examples. If you do not specify a value for the mean. This value is the center line on the C chart. Data C and U require that you have 1 value for each subgroup. In Mean. say 50. you also specify the subgroup sizes with a subgroup indicator. all stages use that value. For example. Each stage for each variable uses a different value. enter the historical means as constants. C Chart − Options − Parameters Stat > Control Charts > Attributes Charts > C > C Chart Options > Parameters You can enter historical data for estimating c. U Chart then divides the number of defects by the subgroup size to get defects per unit. You can enter historical data in several ways: 140 2003 Minitab Inc. If subgroup sizes are not equal. use any dialog box options. is estimated from the data. Each entry in the specified column contains the number of defects for one subgroup. Minitab estimates it from the data. A list of values.Quality Control C Chart Stat > Control Charts > Attributes Charts > C Tracks the number of defects and detects the presence of special causes. enter 50 for the subgroup indicator. which applies to all variables and stages. click control chart Options. enter the size of each subgroup into a column and use that column as the subgroup indicator. Dialog box items Variables: Enter columns that contain the number of defectives for each sample. If you enter one constant. A list of values. The subgroup indicator can be either a constant or a column.

3 12:15): Choose to specify subgroups of data to compute c. you must enter subgroup 2 as the subgroup to omit. except with the Individuals. not just the subsetted data. 141 .Control Charts C Chart Options − Estimate Stat > Control Charts > Attributes Charts > C > C Chart Options > Estimate You can omit or use certain subgroups to estimate c. Note 3 Click OK. Use the following subgroups when estimating parameters (eg. the omitted or used samples apply to the entire data set. 3 12:15): Choose to specify subgroups of the data to omit from the estimation of c. Click the S Limits tab. Minitab assumes the values you enter are subgroup numbers. Each value is the number of standard deviations the lines should be drawn at. and below the mean. Caution When you include or exclude rows using control chart > Data Options > Subset and also omit or use samples to estimate parameters. Because Minitab considers an omitted sample (subgroup 1) as part of the entire data set. a horizontal line labeled LB is drawn at the lower bound instead. then decide to omit subgroup 2 from the parameter estimation (using control chart Options > Estimate). Similarly. you can prevent these subgroups from influencing the estimation of this process parameter. Minitab estimates the process parameters from all the data. if the calculated lower control limit is less than the lower bound. the values are interpreted as observation numbers. If the calculated lower control limit is less than the lower bound. Dialog box items Omit the following subgroups when estimating parameters (eg. For example. then type individual subgroup numbers or a range of subgroups. and I-MR charts. But you may want to use or omit certain data if they show abnormal behavior. click control chart Options. If the calculated upper control limit is greater than the upper bound. click C Chart Options. above. a pair of control limits is drawn for each value in the list. C Chart − Options − S Limits Stat > Control Charts > Attributes Control Charts > C Chart > C Chart Options > S Limits Allows you to draw control limits above and below the mean at the multiples of any standard deviation. For example. Moving Range. Click the Estimate tab. Tip You can also modify the control limits using Minitab's graph editing features. enter one or more values in These multiples of the standard deviation. If you specify more than one value. Do one of the following: • Choose Omit the following samples when estimating parameters and enter the subgroups or observation numbers you want to omit from the calculations. To omit or use subgroups for the estimates of parameters By default. you exclude subgroup 1 from your analysis (using control chart > Data Options > Subset). a horizontal line labeled UB will be drawn at the upper bound instead. To customize control limits for a C chart 1 2 In the C chart dialog box. With these charts. You can set bounds on the upper and/or lower control limits. Dialog box items Display control limits at These multiples of the standard deviation: Enter constants. a horizontal line labeled UB is drawn at the upper bound instead. When the calculated upper control limit is greater than the upper bound. Do any of the following: • To specify where control limits are drawn: under Display control limits at. Upper standard deviation limit bound: Check to set an upper bound for the control limits. a horizontal line labeled LB will be drawn at the lower bound instead. The values entered are the number of standard deviations above and below the center line. • Choose Use the following samples when estimating parameters and enter the subgroups or observation numbers you want to use for the calculations. Place bounds on control limits Lower standard deviation limit bound: Check to set a lower bound for the control limits. 2003 Minitab Inc. then type individual subgroup numbers or ranges of subgroups. if some observations have erratic data due to assignable causes that you have corrected. 1 2 In the control chart dialog box.

See [1] and [10] for guidance on using these tests. U. Each value represents the number of standard deviations below and above the mean. all increasing or all decreasing Test 4 14 points in a row. S. See Tools > Options > Control Charts and Quality Tools > Other. 1 point > 3 standard deviations from center line 9 points in a row on same side of center line 6 points in a row. P. Perform the following tests for special causes: Choose to specify a subset of the four tests. Moving Range. alternating up and down Dialog box items Perform all tests for special causes: Choose to perform all tests. To perform the tests for special causes 1 2 Choose Stat > Control Charts > control chart type > control chart > control chart Options > Tests. The test results are displayed in the Session window. detects a specific pattern in the plotted data. all increasing or all decreasing 14 points in a row. C With any of the attributes control charts. one that should be investigated. Minitab marks it with the test number on the plot. NP. > control chart Options > Tests Charts R. Minitab marks it by the lowest numbered test. Each test. Options − Tests . You can choose to change the color of a control chart when the last row of data causes a new test failure for any point. . you can perform the 4 tests for special causes. Set your preferences for the sensitivity of tests with Tools > Options > Control Charts and Quality Tools > Define Tests. When a point fails a test. If a point fails more than one test. See Updating Graphs for more information. If the data change and the control chart is live-linked.. Test 1 1 point more than 3σ from center line Test 2 9 points in a row on same side of center line Test 3 6 points in a row. alternating up and down Perform no tests: Choose to perform no tests for special causes.. then check any combination of the tests listed below. the test results are not updated in the Session window. 3 Click OK in each dialog box. Minitab will perform tests when sample sizes are unequal. Do one of the following: 142 2003 Minitab Inc. The occurrence of a pattern suggests a special cause for the variation.Quality Control • To set bounds on the control limits: check Lower standard deviation limit bound (and/or Upper standard deviation limit bound) and enter a value. as shown below.

Note With the following charts. 2003 Minitab Inc. you must have at least one subgroup with two or more observations: R.. To display a historical chart To define stages in your process. when subgroups are across rows). To select certain tests. choose Perform all tests for special causes. Minitab treats each repeated value as a separate occurrence. click control chart Options. Click the Stages tab. S. To perform no tests. Enclose date/time or text entries in double quotes. use any dialog box options. Enter the values. Each segment corresponds to a stage (if chart has stages): Choose to split the control chart by stage or stage when you have specified stages in control chart Options > Stage. and X-bar and S. Split chart into series of segments for display purposes Do not split: Choose not to split the chart. You can enter the same value more than once. 5 Options − Display . Minitab treats each repeated value as a separate occurrence.. > control chart Options > Stages Charts All charts except Z-MR You can display stages in your process by drawing a "historical chart" − a control chart in which the control limits and center line are estimated independently for different groups in your data. then enter the values.. you must set up a column of grouping indicators. Click OK. or text. choose Perform the following tests for special causes and check the tests to perform. Options − Stages . > control chart Options > Display Charts All charts except Z-MR Use to display the control chart by stage or by number of plotted points and to display the test results in the Session window. You can enter the same value more than once. and enter the number of subgroups for each segment. • To start a new stage at the first occurrence of a certain value. When executing the command. Historical charts are particularly useful for comparing data before and after a process improvement. The column must be the same length as the data column (or columns. you can tell Minitab to start a new stage in one of two ways: • • 1 2 3 4 Each time the value in the column changes At the first occurrence of one or more values In the control chart dialog box. dates. In Define stages (historical groups) with this variable. enter the column which contains the stage indicators. and enter the number of subgroups.Control Charts • • • 3 To do all tests. Each segment contains __ subgroups: Choose to split the control chart into subgroups of a specified number of points. X-bar and R. Under When to start a new value. If you like. Dialog box items Define stages (historical groups) with this variable: Enter the column that contains the stage indicators. With the first occurrence of these values: Choose to start a new stage at the first occurrence of a certain value. Display last __ subgroups: Choose to display a specific number of subgroups. choose With the first occurrence of these values.. 143 . then click OK. choose With each new value. When to start a new stage With each new value: Choose to start a new stage each time the value in the column changes. choose Perform no tests. Enclose date/time or text entries in double quotes. The indicators can be numbers. Subgroups to display Display all subgroups: Choose to display all subgroups. do one of the following: • To start a new stage each time the value in the column changes.

Under Place bounds on control limits. one row for each plotted point. and 3 standard deviations above and below the center line. − 1 (beyond lower control limit). Minitab assigns each plotted point one of three values: 1 (beyond upper control limit). Under Display control limits at. Do one of the following: • To split the chart into a specific number of subgroups. enter 1 2 3 in These multiples of the standard deviation. under Subgroups to display. Do one of the following: • To display all subgroups in the control chart. • To display a specific number of the last subgroups. . Click the Display tab. To display subgroups 1 2 In the control chart dialog box. 0 (in control). In Variables. Control limit values: Check to store the control limit values. • To split the chart by stage. Click OK. 1 2 3 4 5 6 7 Open the worksheet EXH_QC. Center line value: Check to store the center line value. enter Blemish. click control chart Options. Store these values for each point Point plotted: Check to store the plotted points. Test results: Check to store the results of any test performed. then click OK. 144 2003 Minitab Inc. Check the statistics to store in the worksheet. Minitab stores one column for the lower control limits and one column for the upper control limits. 3 C Chart − Options − Storage Stat > Control Charts > Attributes Charts > C > C Chart Options > Storage Use to store your choice of statistics in the worksheet. one row for each plotted point. Example of C chart with customized control limits Suppose you work for a linen manufacturer. Choose Stat > Control Charts > Attributes Charts > C. one row for each plotted point. check Lower standard deviation limit bound and enter0. Click C Chart Options. choose Each segment corresponds to a stage (if chart has stages). You want the control chart to show control limits at 1. choose Display last __ subgroups and enter the number of subgroups to display. Each 100 square yards of fabric can contain a certain number of blemishes before it is rejected. one row for each historical stage. Stage: Check to store the stage. To store statistics in the worksheet 1 2 Choose Stat > Control Charts > control chart type > control chart > control chart Options > Storage. one row for each plotted point. 3 To split a control chart 1 2 In the control chart dialog box. Click OK. 2. choose Display all subgroups. Dialog box items Store these estimates for each chart Means: Check to store the estimates of the means. choose Each segment contains __ subgroups.Quality Control Test results Display test results in Session window: Check to display the test results in the Session window.MTW. you want to track the number of blemishes per 100 square yards over a period of several days. For quality purposes. under Subgroups to display. to see if your process is behaving predictably. then click the S Limits tab. one row for each plotted point. Click OK in each dialog box. Click the Display tab. and enter the number of subgroups for each segment. click control chart Options.

and [10] for a discussion of these charts. [8]. [3]. [6]. except that they plot statistics from count data rather than measurement data. Minitab also draws two other lines − the upper and lower control limits − 3σ above and below the center line.Control Charts Graph window output Interpreting the results Because the points fall in a random pattern. if you were counting the number of flaws on the inner surface of a television screen. [8]. which charts the proportion of defectives in each subgroup NP Chart. while U Chart would chart the number of flaws per square inch sampled. For example. such as the number of defects. a defect does not always result in a defective product. Use U Chart when the subgroup size varies. Use C Chart when the subgroup size is constant. 2003 Minitab Inc. is plotted versus a sample number or time. For example. a process statistic. by default. and [10] for a discussion of these charts. U Chart Attributes Control Charts Overview Attributes control charts are similar in structure to variables control charts. U Chart. Control charts for defects When a product is complex. As with variables control charts. The control charts for defects are: • • C Chart. which charts the number of defectives in each subgroup See [1]. C Chart would chart the actual number of flaws. [9]. Products may also be classified by their number of defects. For example. Control charts for defectives You can compare a product to a standard and classify it as being defective or not. within the bounds of the 3σ control limits. Minitab draws a center line at the average of the statistic being plotted for the time being charted. you might count the number of scratches on the surface of an appliance. [6]. [3]. [9]. See [1]. Classifying a product by its number of defects is sometimes more convenient. a length of wire either meets the strength requirements or not. The control charts for defectives are: • • P Chart. For instance. which charts the number of defects per unit sampled in each subgroup. products may be compared against a standard and classified as either being defective or not. you conclude the process is behaving predictably and is in control. 145 . which charts the number of defects in each subgroup.

To use historical values for the mean 1 2 3 • • • In the control chart dialog box. click control chart Options. If the subgroups are not equal. which represents the number of defects. Each stage for each variable uses a different value. Minitab also calculates the control limits using this value. say 50. which applies to all variables and stages.Quality Control U Chart Stat > Control Charts > Attributes Charts > U Tracks the number of defects per unit sampled and detects the presence of special causes. if you have a known process parameter or an estimate obtained from past data. For more information. For C Chart. A list of values. assumed to come from a Poisson distribution with the parameter m. In Subgroup sizes. use any dialog box options. Dialog box items Variables: Enter columns that contain the number of defects per unit sampled. A single value. If you like. Minitab estimates the process average number of defects. U Chart − Options − Parameters Stat > Control Charts > Attributes Charts > U > U Chart Options > Parameters You can enter historical data for estimating u. each control limit is not a single straight line but varies with the subgroup size. In U Chart. see Control Charts Overview and Attributes Control Charts Overview. All stages for each variable use the same value. you simply enter the number of defects for each sample into a data column and specify that in the command. If all subgroups are the same size. U Chart then divides the number of defects by the subgroup size to get defects per unit. you also specify the subgroup sizes with a subgroup indicator. enter one or more columns containing the number of defects per unit. The subgroup indicator can be either a constant or a column. Click OK in each dialog box. To display a U chart 1 2 3 4 Choose Stat > Control Charts > Attributes Charts > U. If subgroup sizes are not equal. For example. By default. one for each stage within each variable. enter the size of each subgroup into a column and use that column as the subgroup indicator. You can enter historical data in several ways: 146 2003 Minitab Inc. If you do not specify a value for the mean. you may want to force the control limits to be constant by specifying a fixed subgroup size using U Chart Options > Estimate. then click OK. enter 50 for the subgroup indicator. Subgroup sizes: Enter a number or a column of subscripts. <Scale> <Labels> <Multiple Graphs> <Data Options> <U Chart Options> Data C and U require that you have 1 value for each subgroup. enter a number or a column of subscripts. from the data. You can omit or include certain subgroups when estimating parameters using Estimate. Dialog box item Mean: Enter the historical means as constants. In Mean. m. enter the historical means as constants. See P Chart and NP Chart Data for data entry examples. Minitab estimates it from the data. you can enter the value(s). If the subgroup sizes do not vary much. one for each variable. . which is both the mean and the variance. Each entry in the worksheet column is the number of defects in a sample (or subgroup). In Variables. This value is the center line. A list of values.

except with the Individuals. If you specify more than one value. not just the subsetted data. 3 12:15): Choose to specify subgroups of data to compute u.Control Charts U Chart Options − Estimate Stat > Control Charts > Attributes Charts > U > U Chart Options > Estimate You can omit certain subgroups from the estimation of u. When the calculated upper control limit is greater than the upper bound. You can set bounds on the upper and/or lower control limits. then decide to omit subgroup 2 from the parameter estimation (using control chart Options > Estimate). 1 2 In the control chart dialog box. The values entered are the number of standard deviations above and below the center line. you must enter subgroup 2 as the subgroup to omit. Moving Range. a horizontal line labeled UB is drawn at the upper bound instead. When subgroup sizes are not equal. calculate control limits Using actual sizes of the subgroups: Choose to use the subgroups defined in the main dialog box to estimate σ. a pair of control limits is drawn for each value in the list. R. Dialog box items Omit the following samples when estimating parameters (eg. EWMA Allows you to draw control limits above and below the mean at the multiples of any standard deviation. the values are interpreted as observation numbers. you exclude subgroup 1 from your analysis (using control chart > Data Options > Subset). then type individual subgroup numbers or a range of subgroups. Use the following samples when estimating parameters (eg. a horizontal line labeled LB will be drawn at the lower bound instead. U. For example. NP. • Choose Use the following samples when estimating parameters and enter the subgroups or observation numbers you want to use for the calculations. To omit or use subgroups for the estimates of parameters By default. S. a horizontal line labeled UB will be drawn at the upper bound instead. For an example. if the calculated lower control limit is less than the lower bound. For example. S Limits . but varies with the 2003 Minitab Inc. see Example of an X-bar chart with tests and customized control limits. Upper standard deviation limit bound: Check to set an upper bound for the control limits. Moving Average. Caution When you include or exclude rows using control chart > Data Options > Subset and also omit or use samples to estimate parameters. With these charts. P. When subgroup sizes are unequal. But you may want to use or omit certain data if they show abnormal behavior. You can also force the control limits to be straight if you have unequal sample sizes. Place bounds on control limits Lower standard deviation limit bound: Check to set a lower bound for the control limits. Minitab assumes the values you enter are subgroup numbers. If the calculated lower control limit is less than the lower bound. > control chart Options > S Limits Charts X-bar.. You can also force the control limits to be constant when sample sizes are unequal.. 147 . Click the Estimate tab. if some subgroups have erratic data due to assignable causes that you have corrected. a horizontal line labeled LB is drawn at the lower bound instead. the omitted or used samples apply to the entire data set. Because Minitab considers an omitted sample (subgroup 1) as part of the entire data set. 3 12:15): Choose to specify subgroups of the data to omit from the estimation of u. Tip You can also modify the control limits using Minitab's graph editing features. Assuming all subgroups have size: Choose to estimate the control limits using a specified subgroup size. and I-MR charts. each control limit is not a single straight line. click control chart Options. then type the subgroup size. Note 3 Click OK. If the calculated upper control limit is greater than the upper bound. you can prevent these subgroups from influencing the estimation of process parameters. then type individual subgroup numbers or ranges of subgroups. Dialog box items Display control limits at These multiples of the standard deviation: Enter one or more values. Similarly. Do one of the following: • Choose Omit the following samples when estimating parameters and enter the subgroups or observation numbers you want to omit from the calculations. Minitab estimates the process parameters from all the data.

choose Assuming all subgroups have size. C With any of the attributes control charts. For example. U. suppose the largest subgroup is size 125 and the smallest is 100. you can perform the 4 tests for special causes. To customize the control limits 1 2 In the control chart dialog box. NP. click control chart Options. Each test... • To set bounds on the control limits: check Lower standard deviation limit bound (and/or Upper standard deviation limit bound) and enter a value. For example. Each value represents the number of standard deviations below and above the mean. Each value is the number of standard deviations the lines should be drawn at. Do any of the following: • To specify where control limits are drawn: under Display control limits at. the plotted data are not changed. You can choose to change the color of a control chart when the last row of data causes a new test failure for any point. Minitab will perform tests when sample sizes are unequal. Under When subgroup sizes are unequal. Minitab marks it with the test number on the plot. You should force the control limits and center line to be constant only when the difference in size between the largest and smallest subgroups is no more than 25%. You can use this method because the size difference is 25% (125 / 100 = 25%). S. and below the mean. one that should be investigated. . above. 3 Note Options − Tests . Click the S Limits tab. as shown below. calculate control limits. Test 1 1 point more than 3σ from center line Test 2 9 points in a row on same side of center line 148 2003 Minitab Inc. enter one or more values in These multiples of the standard deviation. See Updating Graphs for more information. The occurrence of a pattern suggests a special cause for the variation. Click OK. If the data change and the control chart is live-linked. The test results are displayed in the Session window. the test results are not updated in the Session window. If a point fails more than one test. Click the S Limits tab. Minitab marks it by the lowest numbered test. click control chart Options. 3 To force control limits and center line to be constant 1 2 In the control chart dialog box. you may want to force the control limits to be constant by entering a sample size in this text box. Set your preferences for the sensitivity of tests with Tools > Options > Control Charts and Quality Tools > Define Tests. Only the control limits are affected. Click OK in each dialog box. > control chart Options > Tests Charts R. enter a value of 6 to calculate the control limits and center line as if all subgroup sizes were 6. then enter a value. If the sizes do not vary much. detects a specific pattern in the plotted data. P.Quality Control subgroup size. See Tools > Options > Control Charts and Quality Tools > Other. See [1] and [10] for guidance on using these tests. Moving Range. When a point fails a test.

choose Perform all tests for special causes. 149 .. The indicators can be numbers. To perform the tests for special causes 1 2 Choose Stat > Control Charts > control chart type > control chart > control chart Options > Tests. then enter the values. 2003 Minitab Inc. Dialog box items Define stages (historical groups) with this variable: Enter the column that contains the stage indicators. The column must be the same length as the data column (or columns. Historical charts are particularly useful for comparing data before and after a process improvement. When to start a new stage With each new value: Choose to start a new stage each time the value in the column changes. • To select certain tests. all increasing or all decreasing 14 points in a row. To display a historical chart To define stages in your process. Click OK. and X-bar and S. when subgroups are across rows).Control Charts Test 3 6 points in a row. or text. Click the Stages tab. > control chart Options > Stages Charts All charts except Z-MR You can display stages in your process by drawing a "historical chart" − a control chart in which the control limits and center line are estimated independently for different groups in your data. S. choose Perform the following tests for special causes and check the tests to perform. dates. click control chart Options. With the first occurrence of these values: Choose to start a new stage at the first occurrence of a certain value. You can enter the same value more than once. alternating up and down Perform no tests: Choose to perform no tests for special causes. alternating up and down Dialog box items Perform all tests for special causes: Choose to perform all tests. all increasing or all decreasing Test 4 14 points in a row. 3 Options − Stages . you can tell Minitab to start a new stage in one of two ways: • • 1 2 Each time the value in the column changes At the first occurrence of one or more values In the control chart dialog box. Do one of the following: • To do all tests. Perform the following tests for special causes: Choose to specify a subset of the four tests. Note With the following charts.. • To perform no tests. choose Perform no tests. Minitab treats each repeated value as a separate occurrence. you must have at least one subgroup with two or more observations: R. X-bar and R. then check any combination of the tests listed below. you must set up a column of grouping indicators. Enclose date/time or text entries in double quotes. 1 point > 3 standard deviations from center line 9 points in a row on same side of center line 6 points in a row. When executing the command.

You can enter the same value more than once. choose With the first occurrence of these values. Do one of the following: • To split the chart into a specific number of subgroups. 150 2003 Minitab Inc. Each segment contains __ subgroups: Choose to split the control chart into subgroups of a specified number of points.. one row for each plotted point. click control chart Options. use any dialog box options. Click the Display tab. choose Each segment contains __ subgroups. To display subgroups 1 2 In the control chart dialog box. enter the column which contains the stage indicators. and enter the number of subgroups.Quality Control 3 4 In Define stages (historical groups) with this variable. choose Display all subgroups.. one row for each historical stage. and enter the number of subgroups for each segment. Enclose date/time or text entries in double quotes. . do one of the following: • • To start a new stage each time the value in the column changes. Enter the values. Store these values for each point Point plotted: Check to store the plotted points. Display last __ subgroups: Choose to display a specific number of subgroups. 5 If you like. choose Display last __ subgroups and enter the number of subgroups to display. choose With each new value. Test results Display test results in Session window: Check to display the test results in the Session window. one row for each plotted point. Options − Display . then click OK. Subgroups to display Display all subgroups: Choose to display all subgroups. 3 U Chart − Options − Storage Stat > Control Charts > Attributes Charts > U > U Chart Options > Storage Use to store your choice of statistics in the worksheet. • To split the chart by stage. under Subgroups to display. and enter the number of subgroups for each segment. 3 To split a control chart 1 2 In the control chart dialog box. Minitab treats each repeated value as a separate occurrence. Under When to start a new value. Click the Display tab. Each segment corresponds to a stage (if chart has stages): Choose to split the control chart by stage or stage when you have specified stages in control chart Options > Stage. under Subgroups to display. Click OK. Center line value: Check to store the center line value. click control chart Options. Dialog box items Store these estimates for each chart Means: Check to store the estimates of the means. To start a new stage at the first occurrence of a certain value. • To display a specific number of the last subgroups. Click OK. > control chart Options > Display Charts All charts except Z-MR Use to display the control chart by stage or by number of plotted points and to display the test results in the Session window. Do one of the following: • To display all subgroups in the control chart. choose Each segment corresponds to a stage (if chart has stages). Split chart into series of segments for display purposes Do not split: Choose not to split the chart.

00 standard deviations from center line. In Variables. 151 . one row for each plotted point. Test results: Check to store the results of any test performed.MTW. one row for each plotted point. You want the U chart to feature straight control limits. Click OK in each dialog box. Check the statistics to store in the worksheet. 6 * WARNING * If graph is updated with new data. 2003 Minitab Inc.Control Charts Control limit values: Check to store the control limit values. one row for each plotted point. One point more than 3. then click OK. you want to monitor the number of defects per unit of motorized toy cars. Test Failed at points: 5. calculate control limits. Minitab assigns each plotted point one of three values: 1 (beyond upper control limit). then click the Display tab. Session window output U Chart of Defects Test Results for U Chart of Defects TEST 1. Under When subgroup sizes are unequal. Subgroup size: Check to store the subgroup sizes. Stage: Check to store the stage. so you fix a subgroup size of 102 (the average number of toys per unit). 1 2 3 4 5 6 7 Open the worksheet TOYS. Click U Chart Options. − 1 (beyond lower control limit). Example of U chart As production manager of a toy manufacturing company. 0 (in control). In Subgroup sizes. choose Assuming all subgroups have size then enter 102. enter Sample. Choose Stat > Control Charts > Attributes Charts > U. You inspect 20 units of toys and create a U chart to examine the number of defects in each unit of toys. enter Defects. one row for each plotted point. To store statistics in the worksheet 1 2 Choose Stat > Control Charts > control chart type > control chart > control chart Options > Storage. Minitab stores one column for the lower control limits and one column for the upper control limits. the results above may no * longer be correct.

Moving Average. However. You should investigate what special causes may have influenced the out-of-control number of motorized toy car defects for these units. When data are in subgroups. is estimated using a pooled standard deviation. 152 2003 Minitab Inc. Typically. See [10] and [23] for a discussion. Time-weighted Charts Moving Average Time-Weighted Control Charts Overview data With the exception of moving average charts. The observations can be either individual measurements or subgroup means. This chart is generally not preferred over an EWMA chart because it does not weight the observations as the EWMA does. the mean of all the observations in each subgroup is calculated. Minitab's time-weighted control charts are weighted either by previous subgroup means or a target value. Choosing a time-weighted control chart Minitab offers three time-weighted control charts: • • • Moving Average − a chart of unweighted moving averages EWMA − a chart of exponentially weighted moving averages CUSUM − a chart of the cumulative sum of the deviations from a nominal specification EWMA.Quality Control Graph window output Interpreting the results Units 5 and 6 are above the upper control limit line. you use these charts to evaluate the process level. you can also use EWMA and CUSUM charts to plot control charts for subgroup ranges or standard deviations to evaluate process variation. Moving Average Chart Stat > Control Charts > Time-weighted Charts > Moving Average A moving average chart is a chart of "moving averages" − averages calculated from artificial subgroups that are created from consecutive observations. By default. For more information about control charts. indicating that special causes may have affected the number of defects in these units. and CUSUM produce control charts for either data in subgroups or individual observations. Moving averages are then formed from these means. The advantage of using time-weighted control charts is the ability to detect small shifts from the target value. see Control Charts Overview. the process standard deviation. . σ.

If the subgroups are not equal. see Control Charts Overview and Time-Weighted Control Charts Overview. since consecutive values have the greatest chance of being alike. you specified a subgroup size of 1). Here is the same data set. using the following table as a guide. then enter the columns. Multivariate data must be entered down columns. Subgroup sizes: Enter a number or a column of subscripts. the average of the moving range divided by an unbiasing constant. <Scale> <Labels> <Multiple Graphs> <Data Options> <MA Options> Data − Variables Control Chart Organize the data for all variables control charts in the same way. then enter the columns. 2003 Minitab Inc. change the length of the moving range. with one column for each variable. Length of MA: Enter the length of the moving averages. and so on. with MRbar / d2. or enter a historical value for σ. Worksheet Structure Structure your data down a column or across rows. If the subgroup sizes do not vary much.Control Charts You can also base the estimate on the average of subgroup ranges or subgroup standard deviations. The value you enter is the number of subgroup means to be included in each average. each control limit is not a single straight line but varies with the subgroup size. For more information. Variables charts include: • • • • Variables charts for subgroups Variables charts for individuals Time-weighted charts Multivariate charts Structure your data for these charts using the guidelines below. structured both ways. If you have individual observations (that is. the second 5 observations are the second row. or enter a historical value for σ. 153 . σ is estimated σ. You can also estimate σ using the median of the moving range. Subgroups are equal size Univariate (one variable) Multivariate (more than one variable) Down columns or across rows Down columns Subgroups are unequal size Down columns with subgroup indicator column Down columns with subgroup indicator column Structure subgroup data down a column or across rows. The moving range is of length 2. Moving ranges are artificial subgroups created from consecutive measurements. Minitab uses them in place of the subgroup means in all calculations. Observations for a subgroup are in one row of columns: Choose if subgroups are arranged in rows across several columns. Note that the first five observations in the left data set (subgroup 1) are the first row of the right-side data set. Dialog box items All observations for a chart are in one column: Choose if data are in one or more columns. with subgroups of size 5. you may want to force the control limits to be constant by specifying a fixed subgroup size using MA Option > Estimate. moving averages are formed from the individual observations. By default. When you have individual observations.

you must enter data that approximate a normal distribution. You can access the Box-Cox transformation two ways: by using the Box-Cox transformation option provided with the control chart commands. . 3 154 2003 Minitab Inc. then enter a series of columns. then click OK. In this example. see Box-Cox Transformation. you may want to use the Box-Cox transformation to induce normality. Nonnormal data To properly interpret Minitab's control charts. In Subgroup sizes. To display a moving average chart 1 2 Choose Stat > Control Charts > Time-weighted Charts > Moving Average. • If subgroups are in rows. For information on the stand alone Box-Cox transformation command. choose Observations for a subgroup are in one row of columns. Then. Use the stand alone command as an exploratory tool to help you determine the best lambda value for the transformation.Quality Control When subgroups are of unequal size. then enter one or more columns. use any dialog box options. If the data are highly skewed. choose All observations for a chart are in one column. C1 contains the process data and C2 contains subgroup indicators: Each time a subscript changes in C2. subgroup 1 has three observations. a new subgroup begins in C1. enter a number or a column of subscripts. Do one of the following: • If subgroups are in one or more columns. you can use the transformation option to transform the data at the same time you draw the control chart. Missing data See Missing data in control charts for information on how to handle missing data for different types of control charts. For information on the Box-Cox transformation option. or by using the stand alone Box-Cox command. and so on. see Options − Box-Cox. then create a second column of subscripts which serve as subgroup indicators. If you like. In the following example. subgroup 2 has six observations. you must enter your data in one column.

I-MR. not individual observations. you exclude subgroup 1 from your analysis (using control chart > Data Options > Subset). 2003 Minitab Inc. X-bar. not just the subsetted data. The default is 3. 155 . A single value. you can prevent these subgroups from influencing the estimation of process parameters. For example. A list of values. CUSUM Omit or includes certain subgroups to estimate µ and σ. > control chart Options > Parameters Charts X-bar and R. For example... Click OK. Do any of the following: • In Mean.. Method for estimating standard deviation: See [2] for a discussion of the relative merits of the methods listed below. change the length of the moving range. Dialog box items Mean: Enter the historical means as constants. change that value to missing (∗) in the worksheet. You can also set preferences for the estimation of σ and the length of the moving range for all Minitab sessions using Tools > Options > Control Charts and Quality Tools > Estimating Standard Deviation. Moving Average. Dialog box items Omit the following samples when estimating parameters: Choose to specify subgroups of the data to omit from the computations for µ and σ. enter one or more values. Because Minitab considers an omitted sample (subgroup 1) as part of the entire data set. Individuals. If you do not specify a value for µ or σ. Standard deviation: Enter the historical standard deviations as constants. enter the number of subgroup means to be included in each average in Length of MA. Minitab only uses the non-omitted data to find lambda. all affected dialog box settings automatically reflect your preferences. these are used in place of the subgroup means in all calculations. Subgroup Size = 1 Average moving range: Choose to estimate σ using the average of the moving range. Moving Average. You can also select one of several methods to estimate σ. To use historical mean and standard deviation values 1 2 In the control chart dialog box. then enter individual subgroup numbers or ranges of subgroups. Minitab estimates it from the data. If you have individual observations (that is you specified a subgroup size of 1). When you set options. All stages for each variable use the same value. then decide to omit subgroup 2 from the parameter estimation (using control chart Options > Estimate).Control Charts To change the length of the moving average In the moving average chart main dialog box. Use the following samples when estimating parameters: Choose to specify subgroups of data to compute µ and σ. > control chart Options > Estimate Charts Zone. EWMA Use to enter historical data for estimating µ and σ. if some subgroups have erratic data due to assignable causes that you have corrected. Options − Parameters . EWMA. enter one or more values. Each stage for each variable uses a different value. When you include or exclude rows using control chart Options > Estimate and choose to perform a Box-Cox transformation. one for each stage within each variable. For example. You can omit or include certain subgroups when estimating parameters using Estimate. if you have a known process parameter or an estimate obtained from past data. A list of values.. and use biased estimates of σ. Enter entire subgroups. the omitted or used samples apply to the entire data set. • In Standard deviation. Zone. you must enter subgroup 2 as the subgroup to omit. To omit one observation from a sample. 3 • • • You can enter historical data in several ways: Options − Estimate . X-bar and S. which applies to all variables and stages. you can enter these values. one for each variable. then enter individual subgroup numbers or a range of subgroups. click control chart Options. Caution When you include or exclude rows using control chart > Data Options > Subset and also omit or use samples to estimate parameters.

156 2003 Minitab Inc. see Example of an X-bar chart with tests and customized control limits. • Choose Use the following samples when estimating parameters and enter the subgroups or observation numbers you want to use for the calculations. NP. a span of two is used because consecutive values have the greatest chance of being alike. To choose how the standard deviation is estimated 1 2 In the control chart dialog box. When subgroup sizes are unequal. Square root of MSSD: Choose to estimate σ using the square root of MSSD (1/2 the mean of the squared successive differences). The values entered are the number of standard deviations above and below the center line. Minitab estimates the process parameters from all the data. Click the Estimate tab. Length of moving range: Check to enter the number of observations used to calculate the moving range. Moving Range.Quality Control Median moving range: Choose to estimate σ using the median of the moving range. click control chart Options. EWMA Allows you to draw control limits above and below the mean at the multiples of any standard deviation. For an example. if the calculated lower control limit is less than the lower bound. But you may want to use or omit certain data if they show abnormal behavior.. The span must be < 100. Do one of the following: • Choose Omit the following samples when estimating parameters and enter the subgroups or observation numbers you want to omit from the calculations. Minitab excludes any moving ranges with omitted data from the calculations. a horizontal line labeled LB will be drawn at the lower bound instead. Click the Estimate tab. Upper standard deviation limit bound: Check to set an upper bound for the control limits. a horizontal line labeled UB is drawn at the upper bound instead. To omit or use subgroups for the estimates of parameters By default. and I-MR charts. except with the Individuals. Subgroup Size > 1 Rbar: Choose to estimate σ based on the average of the subgroup ranges. When you use Omit the following samples when estimating parameters with Length of moving range. If you specify more than one value. . Similarly. the values are interpreted as observation numbers. R. Minitab assumes the values you enter are subgroup numbers. 1 2 In the control chart dialog box. Under Estimating standard deviation. a pair of control limits is drawn for each value in the list. Note S Limits . > control chart Options > S Limits Charts X-bar. click control chart Options. Place bounds on control limits Lower standard deviation limit bound: Check to set a lower bound for the control limits. Sbar: Choose to estimate σ based on the average of the subgroup standard deviations. You can also force the control limits to be constant when sample sizes are unequal. S. choose a method. Use unbiasing constant: Uncheck to use biased estimates of σ. Note 3 Click OK. When the calculated upper control limit is greater than the upper bound.. Dialog box items Display control limits at These multiples of the standard deviation: Enter one or more values. With these charts. U. Pooled standard deviation: Choose to estimate σ using a pooled standard deviation. Moving Average. You can set bounds on the upper and/or lower control limits. P. Tip You can also modify the control limits using Minitab's graph editing features. a horizontal line labeled LB is drawn at the lower bound instead. Then click OK. calculate control limits Using actual sizes of the subgroups: Choose to use the subgroups defined in the main dialog box to estimate σ. If the calculated upper control limit is greater than the upper bound. If the calculated lower control limit is less than the lower bound. a horizontal line labeled UB will be drawn at the upper bound instead. By default.

enter a value of 6 to calculate the control limits and center line as if all subgroup sizes were 6. • To set bounds on the control limits: check Lower standard deviation limit bound (and/or Upper standard deviation limit bound) and enter a value.. and below the mean. Each value is the number of standard deviations the lines should be drawn at. choose Assuming all subgroups have size. you may want to force the control limits to be constant by entering a sample size in this text box. do one of the following: • To start a new stage each time the value in the column changes. when subgroups are across rows). Under When to start a new value. click control chart Options. above. With the first occurrence of these values: Choose to start a new stage at the first occurrence of a certain value. When to start a new stage With each new value: Choose to start a new stage each time the value in the column changes.Control Charts Assuming all subgroups have size: Choose to estimate the control limits using a specified subgroup size. click control chart Options. 157 . X-bar and R. In Define stages (historical groups) with this variable. Click the S Limits tab. Click the S Limits tab. enter the column which contains the stage indicators. You can use this method because the size difference is 25% (125 / 100 = 25%). the plotted data are not changed. Minitab treats each repeated value as a separate occurrence. Each value represents the number of standard deviations below and above the mean. enter one or more values in These multiples of the standard deviation. choose With each new value. Click the Stages tab. but varies with the subgroup size. Click OK. Dialog box items Define stages (historical groups) with this variable: Enter the column that contains the stage indicators. > control chart Options > Stages Charts All charts except Z-MR You can display stages in your process by drawing a "historical chart" − a control chart in which the control limits and center line are estimated independently for different groups in your data. each control limit is not a single straight line. When subgroup sizes are not equal. To display a historical chart To define stages in your process. click control chart Options. Do any of the following: • To specify where control limits are drawn: under Display control limits at. 3 To force control limits and center line to be constant 1 2 In the control chart dialog box. For example. and X-bar and S. calculate control limits. suppose the largest subgroup is size 125 and the smallest is 100. For example. you must have at least one subgroup with two or more observations: R. dates. then type the subgroup size. If the sizes do not vary much. Historical charts are particularly useful for comparing data before and after a process improvement. or text. Under When subgroup sizes are unequal. When executing the command. Enclose date/time or text entries in double quotes. you must set up a column of grouping indicators. Note With the following charts. then enter the values. You can enter the same value more than once.. 3 Note Options − Stages . 2003 Minitab Inc. To customize the control limits 1 2 In the control chart dialog box. The indicators can be numbers. The column must be the same length as the data column (or columns. S. Click OK in each dialog box. then enter a value. you can tell Minitab to start a new stage in one of two ways: • • 1 2 3 4 Each time the value in the column changes At the first occurrence of one or more values In the control chart dialog box. You should force the control limits and center line to be constant only when the difference in size between the largest and smallest subgroups is no more than 25%. Only the control limits are affected.

3 Click OK. unless λ = 0. The Options subdialog box lists the common transformations natural log (λ = 0) and square root (λ = 0. use any dialog box options. 5 If you like. • Choose Lambda=0. For help choosing a lambda value. You can also choose any value between − 5 and 5 for λ. Options − Box-Cox . Minitab treats each repeated value as a separate occurrence. To do the Box-Cox power transformation 1 2 In the control chart dialog box.) To use this option. see the independent Box-Cox transformation command Stat > Control Charts > Box-Cox Transformation. You may want to first run the command Stat > Control Charts > Box-Cox Transformation to help you find the optimal transformation value. in which case the natural log is taken. you should not choose a λ outside the range of −2 and 2. Options − Display . You can enter the same value more than once. Caution If you use Stat > Control Charts > Box-Cox Transformation to find the optimal λ value and choose to store the transformed data with that command. click control chart Options.. • Choose Optimal lambda for each stage (in a chart with stages) to have Minitab search for optimal values for each stage. Enclose date/time or text entries in double quotes. Use overall standard deviation: Check to use the overall standard deviation in the estimation of the optimal lambda. Optimal lambda: Choose to have Minitab search for an optimal value.5 (square root) to use the square root of the data. . Lambda = 0 (natural log): Choose to use the natural log of the data. Lambda=0. When you use this transformation. (λ is lambda.Quality Control • To start a new stage at the first occurrence of a certain value. Enter the values." The transformation takes the original data to the power λ. the data must be positive. > control chart Options > Display Charts All charts except Z-MR Use to display the control chart by stage or by number of plotted points and to display the test results in the Session window. do not select the Box-Cox option when you make a control char. When you include or exclude rows using control chart Options > Estimate. In most cases.5 (square root): Choose to use the square root of the data. Click the Box-Cox tab. then enter the lambda values. then do one of the following: • Choose Lambda = 0 (natural log) to use the natural log of the data. Check Box-Cox power transformation (W = Y**Lambda). Optimal lambda for each stage (in a chart with stages): Choose to have Minitab search for optimal values for each stage. Minitab does not accept any values you enter in the Parameters tab for historical means or historical standard deviations. doing so will double-transform the data.. choose With the first occurrence of these values. Dialog box items Box-Cox power transformation (W = Y**Lambda): Choose to use when your data are very skewed or when the withinsubgroup variation is unstable. Other (enter value(s) between -5 and 5): Choose to transform the data using another lambda value or values. • Choose Optimal lambda to have Minitab search for an optimal value for all stages.. • Choose Other (enter value(s) between -5 and 5) to transform the data using another lambda value or values. then enter one or more values.5). then click OK. 158 2003 Minitab Inc. Minitab only uses the non-omitted data to find lambda.. > control chart Options > Box-Cox Charts All except the Attributes charts You can use the Box-Cox power transformation when your data are very skewed or when the within-subgroup variation is unstable to make the data "more normal.

Check the statistics to store in the worksheet. under Subgroups to display. choose Each segment corresponds to a stage (if chart has stages). choose Display all subgroups. Test results: Check to store the results of any test performed. and enter the number of subgroups for each segment. and enter the number of subgroups. Minitab stores one column for the lower control limit and one column for the upper control limit. • To split the chart by stage. X-bar. To store statistics in the worksheet 1 2 Choose Stat > Control Charts > control chart type > control chart > control chart Options > Storage. Do one of the following: • To split the chart into a specific number of subgroups. choose Display last __ subgroups and enter the number of subgroups to display. • To display a specific number of the last subgroups. one row for each plotted point. Each segment contains __ subgroups: Choose to split the control chart into subgroups of a specified number of points. 3 Options − Storage Stat > Control Charts > control chart type > control chart > control chart Options > Storage Charts X-bar and R. Center line value: Check to store the center line value. Subgroup size: Check to store the subgroup sizes. Stage: Check to store the stage. Test results Display test results in Session window: Check to display the test results in the Session window. one row for each plotted point. Click OK. choose Each segment contains __ subgroups. CUSUM Use to store your choice of statistics in the worksheet. Standard deviations: Check to store the estimates of the standard deviations. 2003 Minitab Inc. then click OK. Moving Average. one row for each plotted point. Do one of the following: • To display all subgroups in the control chart. Click the Display tab. 159 . one row for each historical stage. Split chart into series of segments for display purposes Do not split: Choose not to split the chart. Click OK. Each segment corresponds to a stage (if chart has stages): Choose to split the control chart by stage or stage when you have specified stages in control chart Options > Stage. Control limit values: Check to store the control limit values. Dialog box items Store these estimates for each chart Means: Check to store the estimates of the means. EWMA. and enter the number of subgroups for each segment. X-bar and S. under Subgroups to display.Control Charts Subgroups to display Display all subgroups: Choose to display all subgroups. one row for each plotted point. one row for each plotted point. 3 To split a control chart 1 2 In the control chart dialog box. To display subgroups 1 2 In the control chart dialog box. one row for each plotted point. click control chart Options. Click the Display tab. click control chart Options. one row for each stage. Stores these values for each point Point plotted: Check to store the plotted points. Display last __ subgroups: Choose to display a specific number of subgroups.

33. Choose Stat > Control Charts > Time-weighted charts > Moving Average. Test Failed at points: 5. the results above may no * longer be correct. you want to monitor the weight (in pounds) in 45 batches of limestone that are shipped weekly to an important client. 31. 24. You previously created an I-MR chart. Now you want to examine the same data using a moving average chart. 32. EWMA Chart Time-Weighted Control Charts Overview data With the exception of moving average charts. . Each batch should weight approximately 930 pounds. 44. 30. One point beyond control limits. see Control Charts Overview. Click OK. 7. then enter Weight. 1 2 3 4 Open the worksheet EXH_QC. type 1. Session window output Moving Average Chart of Weight Test Results for Moving Average Chart of Weight TEST.Quality Control Example of moving average chart As the distribution manager at a limestone quarry. For more information about control charts. 45 * WARNING * If graph is updated with new data. Graph window output Interpreting the results The moving average chart shows nine points outside of the control limits. You should closely examine the quarry's processes to improve control over the weight of the limestone shipments. 160 2003 Minitab Inc. The advantage of using time-weighted control charts is the ability to detect small shifts from the target value.MTW. Minitab's time-weighted control charts are weighted either by previous subgroup means or a target value. Choose All observations for a chart are in one column. In Subgroup sizes.

you may want to force the control limits to be constant by specifying a fixed subgroup size using EWMA Options > Estimate. using the following table as a guide. Because of this. and CUSUM produce control charts for either data in subgroups or individual observations. since consecutive values have the greatest chance of being alike. you can also use EWMA and CUSUM charts to plot control charts for subgroup ranges or standard deviations to evaluate process variation. the average of the moving range divided by an unbiasing constant. you can construct a chart with specific properties. For more information. See [10] and [23] for a discussion. 2003 Minitab Inc. or enter an historical value for σ. with one column for each variable. EWMA charts can be custom tailored to detect any size shift in the process. EWMA Chart Stat > Control Charts > Time-weighted Charts > EWMA A chart of exponentially weighted moving averages. You can also base the estimate on the average of subgroup ranges or subgroup standard deviations. When you have individual observations. The moving range is of length 2. However. By default. The value specified must be a number between 0 and 1. Dialog box items All observations for a chart are in one column: Choose if data are in one or more columns. By changing the weight used and the number of standard deviations for the control limits. <Scale> <Labels> <Multiple Graphs> <Data Options> <EWMA Options> Data − Variables Control Chart Organize the data for all variables control charts in the same way. By default. exponentially weighted moving averages are formed from the individual observations. Worksheet Structure Structure your data down a column or across rows. 161 . is estimated using a pooled standard deviation. Each EWMA point incorporates information from all the previous subgroups or observations.Control Charts Choosing a time-weighted control chart Minitab offers three time-weighted control charts: • • • Moving Average − a chart of unweighted moving averages EWMA − a chart of exponentially weighted moving averages CUSUM − a chart of the cumulative sum of the deviations from a nominal specification EWMA. When data are in subgroups. You can choose combinations of these two parameters by using an ARL (Average Run Length) table. the process standard deviation. the mean of all the observations in each subgroup is calculated. Variables charts include: • • • • Variables charts for subgroups Variables charts for individuals Time-weighted charts Multivariate charts Structure your data for these charts using the guidelines below. Observations for a subgroup are in one row of columns: Choose if subgroups are arranged in rows across several columns. Moving ranges are artificial subgroups created from the individual measurements. If the subgroup sizes do not vary much. Subgroup sizes: Enter a number or a column of subscripts. see Control Charts Overview and Time-Weighted Control Charts Overview. you use these charts to evaluate the process level. Multivariate data must be entered down columns. See [17] for an extensive table. You can also estimate σ using the median of the moving range. or enter a historical value for σ. then enter the columns. σ is estimated with MRbar / d2. then enter the columns. change the length of the moving range. they are often used to monitor in-control processes for detecting small shifts away from the target. The plot points can be based on either subgroup means or individual observations. Weight for EWMA: Enter the weight to be used in the exponentially weighted moving average. Moving Average. σ. Exponentially weighted moving averages are then formed from these means. Typically.

Here is the same data set. and so on. you must enter data that approximate a normal distribution. For information on the stand alone Box-Cox transformation command. structured both ways. When subgroups are of unequal size. Missing data See Missing data in control charts for information on how to handle missing data for different types of control charts. you can use the transformation option to transform the data at the same time you draw the control chart. Nonnormal data To properly interpret Minitab's control charts. Then. For information on the Box-Cox transformation option. subgroup 2 has six observations. a new subgroup begins in C1. In this example.Quality Control Subgroups are equal size Univariate (one variable) Multivariate (more than one variable) Down columns or across rows Down columns Subgroups are unequal size Down columns with subgroup indicator column Down columns with subgroup indicator column Structure subgroup data down a column or across rows. Note that the first five observations in the left data set (subgroup 1) are the first row of the right-side data set. and so on. . If the data are highly skewed. the second 5 observations are the second row. Use the stand alone command as an exploratory tool to help you determine the best lambda value for the transformation. C1 contains the process data and C2 contains subgroup indicators: Each time a subscript changes in C2. see Box-Cox Transformation. 162 2003 Minitab Inc. You can access the Box-Cox transformation two ways: by using the Box-Cox transformation option provided with the control chart commands. then create a second column of subscripts which serve as subgroup indicators. see Options − Box-Cox. you may want to use the Box-Cox transformation to induce normality. with subgroups of size 5. you must enter your data in one column. subgroup 1 has three observations. In the following example. or by using the stand alone Box-Cox command.

. and use biased estimates of σ. You can omit or include certain subgroups when estimating parameters using Estimate. You can also set preferences for the estimation of σ and the length of the moving range for all Minitab sessions using Tools > Options > Control Charts and Quality Tools > Estimating Standard Deviation. change that value to missing (∗) in the worksheet. Minitab estimates it from the data. enter a number or a column of subscripts.. then enter a series of columns. Zone. Caution When you include or exclude rows using control chart > Data Options > Subset and also omit or use samples to estimate parameters. you must enter subgroup 2 as the subgroup to omit. X-bar and S. EWMA.Control Charts To make an EWMA chart 1 2 Choose Stat > Control Charts > Time-weighted Charts > EWMA. Dialog box items Mean: Enter the historical means as constants. A single value. one for each variable. I-MR. For example. If you do not specify a value for µ or σ. Standard deviation: Enter the historical standard deviations as constants. enter one or more values. not individual observations. choose All observations for a chart are in one column. Each stage for each variable uses a different value. enter one or more values. All stages for each variable use the same value. Enter entire subgroups. To omit one observation from a sample. use any dialog box options. you exclude subgroup 1 from your analysis (using control chart > Data Options > Subset). 2003 Minitab Inc. choose Observations for a subgroup are in one row of columns. When you set options. Moving Average. one for each stage within each variable. CUSUM Omit or includes certain subgroups to estimate µ and σ. • If subgroups are in rows. you can enter these values. if you have a known process parameter or an estimate obtained from past data. Because Minitab considers an omitted sample (subgroup 1) as part of the entire data set.. > control chart Options > Parameters Charts X-bar and R. the omitted or used samples apply to the entire data set. if some subgroups have erratic data due to assignable causes that you have corrected. Moving Average. Click OK. then click OK. Do any of the following: • In Mean. You can also select one of several methods to estimate σ. A list of values. When you include or exclude rows using control chart Options > Estimate and choose to perform a Box-Cox transformation. change the length of the moving range. EWMA Use to enter historical data for estimating µ and σ. For example. If you like. Do one of the following: • If subgroups are in one or more columns. then enter individual subgroup numbers or ranges of subgroups. then decide to omit subgroup 2 from the parameter estimation (using control chart Options > Estimate). In Subgroup sizes. For example. To use historical mean and standard deviation values 1 2 In the control chart dialog box. 3 • • • You can enter historical data in several ways: Options − Estimate . Individuals. > control chart Options > Estimate Charts Zone. 163 . then enter one or more columns. you can prevent these subgroups from influencing the estimation of process parameters.. all affected dialog box settings automatically reflect your preferences. A list of values. not just the subsetted data. click control chart Options. Minitab only uses the non-omitted data to find lambda. 3 Options − Parameters . which applies to all variables and stages. Dialog box items Omit the following samples when estimating parameters: Choose to specify subgroups of the data to omit from the computations for µ and σ. • In Standard deviation. X-bar.

You can set bounds on the upper and/or lower control limits. Minitab estimates the process parameters from all the data. a span of two is used because consecutive values have the greatest chance of being alike. Square root of MSSD: Choose to estimate σ using the square root of MSSD (1/2 the mean of the squared successive differences). Subgroup Size = 1 Average moving range: Choose to estimate σ using the average of the moving range. the values are interpreted as observation numbers. P. a pair of control limits is drawn for each value in the list. Minitab assumes the values you enter are subgroup numbers. To choose how the standard deviation is estimated 1 2 In the control chart dialog box. Method for estimating standard deviation: See [2] for a discussion of the relative merits of the methods listed below. Pooled standard deviation: Choose to estimate σ using a pooled standard deviation.Quality Control Use the following samples when estimating parameters: Choose to specify subgroups of data to compute µ and σ. Do one of the following: • Choose Omit the following samples when estimating parameters and enter the subgroups or observation numbers you want to omit from the calculations. Note 3 Click OK. 1 2 In the control chart dialog box. Note S Limits . and I-MR charts. The values entered are the number of standard deviations above and below the center line. Moving Average. EWMA Allows you to draw control limits above and below the mean at the multiples of any standard deviation. Click the Estimate tab. R. NP. Then click OK. 164 2003 Minitab Inc. if the calculated lower control limit is less than the lower bound. Minitab excludes any moving ranges with omitted data from the calculations. You can also force the control limits to be constant when sample sizes are unequal. When you use Omit the following samples when estimating parameters with Length of moving range. If you specify more than one value. then enter individual subgroup numbers or a range of subgroups. Length of moving range: Check to enter the number of observations used to calculate the moving range. To omit or use subgroups for the estimates of parameters By default. > control chart Options > S Limits Charts X-bar. Tip You can also modify the control limits using Minitab's graph editing features.. Under Estimating standard deviation. Dialog box items Display control limits at These multiples of the standard deviation: Enter one or more values. But you may want to use or omit certain data if they show abnormal behavior. Click the Estimate tab. click control chart Options. a horizontal line labeled LB is drawn at the lower bound instead.. choose a method. • Choose Use the following samples when estimating parameters and enter the subgroups or observation numbers you want to use for the calculations. Moving Range. a horizontal line labeled UB is drawn at the upper bound instead. When the calculated upper control limit is greater than the upper bound. Use unbiasing constant: Uncheck to use biased estimates of σ. except with the Individuals. see Example of an X-bar chart with tests and customized control limits. Subgroup Size > 1 Rbar: Choose to estimate σ based on the average of the subgroup ranges. Median moving range: Choose to estimate σ using the median of the moving range. Similarly. The span must be < 100. . With these charts. Sbar: Choose to estimate σ based on the average of the subgroup standard deviations. S. click control chart Options. By default. U. For an example.

If the calculated lower control limit is less than the lower bound. Click the S Limits tab. a horizontal line labeled UB will be drawn at the upper bound instead. If the calculated upper control limit is greater than the upper bound. each control limit is not a single straight line. then enter the values. suppose the largest subgroup is size 125 and the smallest is 100. If the sizes do not vary much. Each value represents the number of standard deviations below and above the mean. For example.. and below the mean. you must set up a column of grouping indicators. Minitab treats each repeated value as a separate occurrence. You can use this method because the size difference is 25% (125 / 100 = 25%). choose Assuming all subgroups have size. Under When subgroup sizes are unequal. Click OK. To display a historical chart To define stages in your process. a horizontal line labeled LB will be drawn at the lower bound instead. and X-bar and S. calculate control limits. S. enter one or more values in These multiples of the standard deviation. enter a value of 6 to calculate the control limits and center line as if all subgroup sizes were 6. 3 To force control limits and center line to be constant 1 2 In the control chart dialog box. Enclose date/time or text entries in double quotes. Historical charts are particularly useful for comparing data before and after a process improvement. Dialog box items Define stages (historical groups) with this variable: Enter the column that contains the stage indicators. You should force the control limits and center line to be constant only when the difference in size between the largest and smallest subgroups is no more than 25%. or text. 3 Note Options − Stages . When executing the command. 165 . • To set bounds on the control limits: check Lower standard deviation limit bound (and/or Upper standard deviation limit bound) and enter a value. Assuming all subgroups have size: Choose to estimate the control limits using a specified subgroup size. Do any of the following: • To specify where control limits are drawn: under Display control limits at. then type the subgroup size. To customize the control limits 1 2 In the control chart dialog box. you must have at least one subgroup with two or more observations: R. The indicators can be numbers.Control Charts Place bounds on control limits Lower standard deviation limit bound: Check to set a lower bound for the control limits. When to start a new stage With each new value: Choose to start a new stage each time the value in the column changes. you can tell Minitab to start a new stage in one of two ways: • • Each time the value in the column changes At the first occurrence of one or more values 2003 Minitab Inc. you may want to force the control limits to be constant by entering a sample size in this text box. click control chart Options. above. Click the S Limits tab.. click control chart Options. With the first occurrence of these values: Choose to start a new stage at the first occurrence of a certain value. When subgroup sizes are not equal. X-bar and R. Only the control limits are affected. Upper standard deviation limit bound: Check to set an upper bound for the control limits. > control chart Options > Stages Charts All charts except Z-MR You can display stages in your process by drawing a "historical chart" − a control chart in which the control limits and center line are estimated independently for different groups in your data. For example. You can enter the same value more than once. dates. calculate control limits Using actual sizes of the subgroups: Choose to use the subgroups defined in the main dialog box to estimate σ. Each value is the number of standard deviations the lines should be drawn at. but varies with the subgroup size. the plotted data are not changed. When subgroup sizes are unequal. then enter a value. Click OK in each dialog box. Note With the following charts.

Click the Stages tab. Enter the values. choose With each new value. Click the Box-Cox tab. For help choosing a lambda value. doing so will double-transform the data. Optimal lambda: Choose to have Minitab search for an optimal value. enter the column which contains the stage indicators.Quality Control The column must be the same length as the data column (or columns." The transformation takes the original data to the power λ. When you include or exclude rows using control chart Options > Estimate. The Options subdialog box lists the common transformations natural log (λ = 0) and square root (λ = 0. Options − Box-Cox . when subgroups are across rows). In Define stages (historical groups) with this variable. Lambda=0. Dialog box items Box-Cox power transformation (W = Y**Lambda): Choose to use when your data are very skewed or when the withinsubgroup variation is unstable. • Choose Other (enter value(s) between -5 and 5) to transform the data using another lambda value or values. you should not choose a λ outside the range of −2 and 2. do not select the Box-Cox option when you make a control char. click control chart Options. Minitab treats each repeated value as a separate occurrence. 3 Click OK. unless λ = 0. You may want to first run the command Stat > Control Charts > Box-Cox Transformation to help you find the optimal transformation value.5 (square root) to use the square root of the data. use any dialog box options. Use overall standard deviation: Check to use the overall standard deviation in the estimation of the optimal lambda. In most cases. 166 2003 Minitab Inc. (λ is lambda. 5 If you like. then enter one or more values. do one of the following: • • To start a new stage each time the value in the column changes. Optimal lambda for each stage (in a chart with stages): Choose to have Minitab search for optimal values for each stage. Other (enter value(s) between -5 and 5): Choose to transform the data using another lambda value or values. • Choose Optimal lambda for each stage (in a chart with stages) to have Minitab search for optimal values for each stage. Minitab does not accept any values you enter in the Parameters tab for historical means or historical standard deviations. You can enter the same value more than once.5). To do the Box-Cox power transformation 1 2 In the control chart dialog box. > control chart Options > Box-Cox Charts All except the Attributes charts You can use the Box-Cox power transformation when your data are very skewed or when the within-subgroup variation is unstable to make the data "more normal. Lambda = 0 (natural log): Choose to use the natural log of the data. Enclose date/time or text entries in double quotes.. To start a new stage at the first occurrence of a certain value. then do one of the following: • Choose Lambda = 0 (natural log) to use the natural log of the data.5 (square root): Choose to use the square root of the data. When you use this transformation. . Minitab only uses the non-omitted data to find lambda. • Choose Optimal lambda to have Minitab search for an optimal value for all stages. in which case the natural log is taken. 1 2 3 4 In the control chart dialog box. • Choose Lambda=0. click control chart Options.. see the independent Box-Cox transformation command Stat > Control Charts > Box-Cox Transformation.) To use this option. choose With the first occurrence of these values. Under When to start a new value. then enter the lambda values. the data must be positive. Caution If you use Stat > Control Charts > Box-Cox Transformation to find the optimal λ value and choose to store the transformed data with that command. You can also choose any value between − 5 and 5 for λ. then click OK. Check Box-Cox power transformation (W = Y**Lambda).

click control chart Options. Subgroups to display Display all subgroups: Choose to display all subgroups. Dialog box items Store these estimates for each chart Means: Check to store the estimates of the means. one row for each plotted point. Click OK. Center line value: Check to store the center line value. one row for each plotted point. Do one of the following: • To split the chart into a specific number of subgroups. one row for each plotted point. and enter the number of subgroups. Stage: Check to store the stage. Test results: Check to store the results of any test performed. Click the Display tab. choose Display all subgroups. CUSUM Use to store your choice of statistics in the worksheet. 2003 Minitab Inc. one row for each historical stage. > control chart Options > Display Charts All charts except Z-MR Use to display the control chart by stage or by number of plotted points and to display the test results in the Session window. Moving Average. Each segment corresponds to a stage (if chart has stages): Choose to split the control chart by stage or stage when you have specified stages in control chart Options > Stage.. under Subgroups to display. X-bar. • To display a specific number of the last subgroups. X-bar and S. choose Display last __ subgroups and enter the number of subgroups to display. Display last __ subgroups: Choose to display a specific number of subgroups. one row for each plotted point. EWMA.Control Charts Options − Display . and enter the number of subgroups for each segment. choose Each segment corresponds to a stage (if chart has stages).. choose Each segment contains __ subgroups. under Subgroups to display. 167 . one row for each stage. Minitab stores one column for the lower control limit and one column for the upper control limit. Split chart into series of segments for display purposes Do not split: Choose not to split the chart. 3 Options − Storage Stat > Control Charts > control chart type > control chart > control chart Options > Storage Charts X-bar and R. one row for each plotted point. To display subgroups 1 2 In the control chart dialog box. Each segment contains __ subgroups: Choose to split the control chart into subgroups of a specified number of points. Do one of the following: • To display all subgroups in the control chart. and enter the number of subgroups for each segment. Control limit values: Check to store the control limit values. • To split the chart by stage. Click the Display tab. Standard deviations: Check to store the estimates of the standard deviations. Test results Display test results in Session window: Check to display the test results in the Session window. Subgroup size: Check to store the subgroup sizes. Click OK. one row for each plotted point. 3 To split a control chart 1 2 In the control chart dialog box. click control chart Options. Stores these values for each point Point plotted: Check to store the plotted points.

4. 3. Click OK. . enter 5. Choose Stat > Control Charts >Time-weighted Charts > EWMA. Weight contains the weight in pounds of each batch of raw material. 4. Graph window output Interpreting the results The EWMAs for sample numbers 2. Test Failed at points: 2.MTW. One point beyond control limits. Choose All observations for a chart are in one column. 5 * WARNING * If graph is updated with new data. 168 2003 Minitab Inc. the results above may no * longer be correct. Session window output EWMA Chart of Weight Test Results for EWMA Chart of Weight TEST. Example of EWMA Chart In the following example.Quality Control To store statistics in the worksheet 1 2 Choose Stat > Control Charts > control chart type > control chart > control chart Options > Storage. In Subgroup sizes. 3. 1 2 3 4 Open the worksheet EXH_QC. Check the statistics to store in the worksheet. and 5 fall below the lower control limit. then enter Weight. then click OK. suggesting small shifts away from the target.

change the length of the moving range. Variables charts include: • Variables charts for subgroups 2003 Minitab Inc. both the CUSUM chart and EWMA chart are good at detecting small shifts away from the target. the process standard deviation. see Control Charts Overview. σ is estimated σ. CUSUM Chart Stat > Control Charts > Time-weighted Charts > CUSUM Displays the cumulative sums of the deviations of each sample value from the target value. specification. rather than the usual 3σ control limits. You can display two types of CUSUM charts. or nominal. Target: Enter a number to use as a target. then enter the columns. you use these charts to evaluate the process level. Subgroup sizes: Enter a number or a column of subscripts. By default. For more information. The default chart includes two one-sided CUSUMs. See [10] and [23] for a discussion. and CUSUM produce control charts for either data in subgroups or individual observations. See [24] and [26] for a discussion of onesided CUSUMs.Control Charts CUSUM Chart Time-Weighted Control Charts Overview data With the exception of moving average charts. CUSUM statistics are cumulative deviations from this target. then enter the columns. CUSUM statistics are then formed from these means. σ. with MRbar / d2. Observations for a subgroup are in one row of columns: Choose if subgroups are arranged in rows across several columns. since consecutive values have the greatest chance of being alike. The upper CUSUM detects upward shifts in the level of the process. When data are in subgroups. to determine when a process is out-of-control. Typically. the mean of all the observations in each subgroup is calculated. CUSUM statistics are formed from the individual observations. Minitab's time-weighted control charts are weighted either by previous subgroup means or a target value. You can plot a chart based on subgroup means or individual observations. Dialog box items All observations for a chart are in one column: Choose if data are in one or more columns. This chart uses control limits (UCL and LCL) to determine when an out-of-control situation has occurred. However. With in-control processes. You can also estimate σ using the median of the moving range. see Control Charts Overview and Time-Weighted Control Charts Overview. Moving Average. All subgroups must be the same size. the average of the moving range divided by an unbiasing constant. This chart uses a V-mask. The advantage of using time-weighted control charts is the ability to detect small shifts from the target value. you can also use EWMA and CUSUM charts to plot control charts for subgroup ranges or standard deviations to evaluate process variation. is estimated using a pooled standard deviation. By default. or enter an historical value for σ. <Scale> <Labels> <Multiple Graphs> <Data Options> <CUSUM Options> Data − Variables Control Chart Organize the data for all variables control charts in the same way. For more information about control charts. 169 . Choosing a time-weighted control chart Minitab offers three time-weighted control charts: • • • Moving Average − a chart of unweighted moving averages EWMA − a chart of exponentially weighted moving averages CUSUM − a chart of the cumulative sum of the deviations from a nominal specification EWMA. When you have individual observations. or enter a historical value for σ. The moving range is of length 2. See [15] and [28] for a discussion of the V-mask chart. You can also base the estimate on the average of subgroup ranges or subgroup standard deviations. Moving ranges are artificial subgroups created from the individual measurements. You can also plot one two-sided CUSUM. the lower CUSUM detects downward shifts.

Nonnormal data To properly interpret Minitab's control charts. Subgroups are equal size Univariate (one variable) Multivariate (more than one variable) Down columns or across rows Down columns Subgroups are unequal size Down columns with subgroup indicator column Down columns with subgroup indicator column Structure subgroup data down a column or across rows. Use the stand alone command as an 170 2003 Minitab Inc. with one column for each variable. the second 5 observations are the second row. and so on. In the following example. Here is the same data set. C1 contains the process data and C2 contains subgroup indicators: Each time a subscript changes in C2. Worksheet Structure Structure your data down a column or across rows. then create a second column of subscripts which serve as subgroup indicators. with subgroups of size 5. In this example. Note that the first five observations in the left data set (subgroup 1) are the first row of the right-side data set. If the data are highly skewed. you may want to use the Box-Cox transformation to induce normality. you must enter your data in one column. . Multivariate data must be entered down columns. subgroup 1 has three observations. subgroup 2 has six observations. you must enter data that approximate a normal distribution.Quality Control • • • Variables charts for individuals Time-weighted charts Multivariate charts Structure your data for these charts using the guidelines below. a new subgroup begins in C1. using the following table as a guide. You can access the Box-Cox transformation two ways: by using the Box-Cox transformation option provided with the control chart commands. When subgroups are of unequal size. and so on. structured both ways. or by using the stand alone Box-Cox command.

A single value.. one for each stage within each variable. Under Type of CUSUM. In Standard deviation. then click the Plan/Type tab. All stages for each variable use the same value.Control Charts exploratory tool to help you determine the best lambda value for the transformation. To plot two one-sided CUSUMs 1 2 Choose Stat > Control Charts > Time-weighted Charts > CUSUM.. which applies to all variables and stages. Moving Range. then enter one or more columns. 171 . use any dialog box options. then click OK in each dialog box. you may enter this value or values. For information on the Box-Cox transformation option. > control chart Options > Parameters Charts R. In Subgroup sizes. you can use the transformation option to transform the data at the same time you draw the control chart. Do one of the following: • If subgroups are in one or more columns.. CUSUM 2003 Minitab Inc. then enter a series of columns. Then. Moving Average. In Subgroup sizes. UCL). For information on the stand alone Box-Cox transformation command. see Box-Cox Transformation. then enter a series of columns. If you do not specify a value for σ. To use historical standard deviation values 1 2 3 • • • In the control chart dialog box. Do one of the following: • If subgroups are in one or more columns. • If subgroups are in rows. > control chart Options > Estimate Charts Zone. CUSUM Enters historical data for estimating σ. A list of values. Click CUSUM Options. one for each variable. enter the subgroup number on which to center the V-mask. choose One-sided (LCL. enter a number or a column of subscripts. choose All observations for a chart are in one column. Missing data See Missing data in control charts for information on how to handle missing data for different types of control charts. Dialog box items Standard deviation: Enter the historical standard deviations as constants. Each stage for each variable uses a different value. If you like. You can enter historical data in several ways: Options − Estimate . Click OK. choose Observations for a subgroup are in one row of columns.. choose Two-sided (V-mask). A list of values. then enter one or more columns. enter a number or a column of subscripts. S. For example. 3 4 5 Options − Parameters . click control chart Options. Under Type of CUSUM. then click the Plan/Type tab. choose Observations for a subgroup are in one row of columns. choose All observations for a chart are in one column. 3 4 5 To plot one two-sided (V-mask) CUSUM 1 2 Choose Stat > Control Charts > Time-weighted Charts > CUSUM. enter one or more values. • If subgroups are in rows. You can omit or include certain subgroups when estimating parameters using Estimate. Minitab estimates it from the data. EWMA. see Options − Box-Cox. Click CUSUM Options. In Center on subgroup. if you have a known process parameter or an estimate obtained from past data. then click OK in each dialog box.

not just the subsetted data. Subgroup Size = 1 Average moving range: Choose to estimate σ using the average of the moving range. Caution When you include or exclude rows using control chart > Data Options > Subset and also omit or use samples to estimate parameters. Note 172 2003 Minitab Inc. if some subgroups have erratic data due to assignable causes that you have corrected. then enter individual subgroup numbers or a range of subgroups. For example. click control chart Options. Method for estimating standard deviation: See [2] for a discussion of the relative merits of the methods listed below. When you use Omit the following samples when estimating parameters with Length of moving range. Then click OK. By default. You can also select one of several methods to estimate σ. Click the Estimate tab. With these charts. • Choose Use the following samples when estimating parameters and enter the subgroups or observation numbers you want to use for the calculations. But you may want to use or omit certain data if they show abnormal behavior. the values are interpreted as observation numbers. Use unbiasing constant: Uncheck to use biased estimates of σ. Subgroup Size > 1 Rbar: Choose to estimate σ based on the average of the subgroup ranges. The span must be < 100. and use biased estimates of σ.Quality Control Omit or includes certain subgroups to estimate µ and σ. all affected dialog box settings automatically reflect your preferences. a span of two is used because consecutive values have the greatest chance of being alike. . then decide to omit subgroup 2 from the parameter estimation (using control chart Options > Estimate). When you include or exclude rows using control chart Options > Estimate and choose to perform a Box-Cox transformation. To choose how the standard deviation is estimated 1 2 In the control chart dialog box. not individual observations. choose a method. To omit or use subgroups for the estimates of parameters By default. Under Estimating standard deviation. and I-MR charts. click control chart Options. Note 3 Click OK. Click the Estimate tab. then enter individual subgroup numbers or ranges of subgroups. change that value to missing (∗) in the worksheet. To omit one observation from a sample. you exclude subgroup 1 from your analysis (using control chart > Data Options > Subset). Minitab estimates the process parameters from all the data. Median moving range: Choose to estimate σ using the median of the moving range. When you set options. For example. Dialog box items Omit the following samples when estimating parameters: Choose to specify subgroups of the data to omit from the computations for µ and σ. except with the Individuals. Length of moving range: Check to enter the number of observations used to calculate the moving range. Minitab only uses the non-omitted data to find lambda. Square root of MSSD: Choose to estimate σ using the square root of MSSD (1/2 the mean of the squared successive differences). Pooled standard deviation: Choose to estimate σ using a pooled standard deviation. you can prevent these subgroups from influencing the estimation of process parameters. Minitab assumes the values you enter are subgroup numbers. Minitab excludes any moving ranges with omitted data from the calculations. Sbar: Choose to estimate σ based on the average of the subgroup standard deviations. You can also set preferences for the estimation of σ and the length of the moving range for all Minitab sessions using Tools > Options > Control Charts and Quality Tools > Estimating Standard Deviation. Use the following samples when estimating parameters: Choose to specify subgroups of data to compute µ and σ. Because Minitab considers an omitted sample (subgroup 1) as part of the entire data set. 1 2 In the control chart dialog box. change the length of the moving range. the omitted or used samples apply to the entire data set. Do one of the following: • Choose Omit the following samples when estimating parameters and enter the subgroups or observation numbers you want to omit from the calculations. Enter entire subgroups. Moving Range. you must enter subgroup 2 as the subgroup to omit.

2003 Minitab Inc. k is the slope of the V-mask arms. Number of standard deviations: Enter the number of standard deviations above and below the center line. then click OK: • In h. See [15] and [16]. • In k. This chart uses an UCL and an LCL to determine when a process is out-of-control. Use the upper CUSUM to detect upward shifts in the level of the process and the lower CUSUM to detect downward shifts. For one-sided CUSUMs. but varies with the subgroup size. you must have at least one subgroup with two or more observations: R. Minitab centers the V-mask on the last subgroup. Only the control limits are affected. k is the allowable "slack" in the process. Historical charts are particularly useful for comparing data before and after a process improvement. When to start a new stage With each new value: Choose to start a new stage each time the value in the column changes. k: Type a value greater than 0 for k. an attempt should be made to find and eliminate the cause of the problem. If the sizes do not vary much. Reset after each signal: Check to reset the CUSUMs to their initial values whenever an out-of-control signal is generated. When subgroup sizes are not equal. Dialog box items Define stages (historical groups) with this variable: Enter the column that contains the stage indicators. Center on subgroup: Enter a subgroup number on which to center the V-mask. To specify a different CUSUM plan (h and k) 1 2 3 Choose Stat > Control Charts > Time-weighted Charts > CUSUM. and X-bar and S. You can select k using an ARL table.. the plotted data are not changed. type a value greater than zero. Options − Stages . the CUSUMs should be reset to their initial values. See [15] and [16]. CUSUM Plan: These values are often selected from ARL (Average Run Length) tables. See [24] and [26] for a discussion of one-sided CUSUMs. For two-sided CUSUMs (V-mask). Use FIR: Check to use the FIR (Fast Initial Response) method to initialize the one-sided CUSUM. Two-sided (V-mask): Click to generate a single two-sided CUSUM that uses a V-mask to determine when a process is out-of-control. each control limit is not a single straight line. If no value is entered. the CUSUMs will not detect the situation for several subgroups. This method has been shown by [15] to reduce the number of subgroups needed to detect problems at startup.5. type a value greater than zero. For one-sided CUSUMs. Do one or both of the following. then click the Plan/Type tab. > control chart Options > Stages Charts All charts except Z-MR You can display stages in your process by drawing a "historical chart" − a control chart in which the control limits and center line are estimated independently for different groups in your data.. See [15] and [28] for a discussion of the V-mask chart. Note With the following charts. h is the number of standard deviations between the center line and the control limits. 173 . When a process goes out of control. Assuming all subgroups have size: Choose to estimate the control limits using a specified subgroup size. Minitab calculates the half-width of the V-mask (H) at the point of origination by H = h * σ. The default value is 0. h: Type a value greater than 0 for h. If the problem has been corrected. S. Dialog box items Type of CUSUM One-sided (LCL.Control Charts CUSUM Chart − Options − Plan/Type Stat > Control Charts > Time-weighted Charts > CUSUM > CUSUM Options > Plan/Type Allows you to choose and define the CUSUM. See [15] and [16]. Normally they are initialized at 0. When subgroup sizes are unequal. You can select h using an ARL table. UCL): Click to generate two one-sided CUSUMs. but if the process is out of control at startup. The default value is 4. then type the subgroup size. X-bar and R. Click CUSUM Options. you may want to force the control limits to be constant by entering a sample size in this text box. For two-sided CUSUMs. calculate control limits/V-mask Using actual sizes of the subgroups: Choose to use the subgroups defined in the main dialog box to estimate σ.

In most cases..5).) To use this option. 5 Options − Box-Cox . If you like. You can enter the same value more than once. Under When to start a new value. Optimal lambda for each stage (in a chart with stages): Choose to have Minitab search for optimal values for each stage. dates. Optimal lambda: Choose to have Minitab search for an optimal value. click control chart Options. doing so will double-transform the data. then click OK. . the data must be positive. Minitab treats each repeated value as a separate occurrence. you should not choose a λ outside the range of −2 and 2. To do the Box-Cox power transformation 1 2 In the control chart dialog box. do not select the Box-Cox option when you make a control char. When you use this transformation. Minitab treats each repeated value as a separate occurrence. To display a historical chart To define stages in your process. do one of the following: • To start a new stage each time the value in the column changes. or text. Minitab does not accept any values you enter in the Parameters tab for historical means or historical standard deviations. When you include or exclude rows using control chart Options > Estimate. You can enter the same value more than once. unless λ = 0. Enclose date/time or text entries in double quotes. you can tell Minitab to start a new stage in one of two ways: • • 1 2 3 4 Each time the value in the column changes At the first occurrence of one or more values In the control chart dialog box. In Define stages (historical groups) with this variable. Other (enter value(s) between -5 and 5): Choose to transform the data using another lambda value or values. then do one of the following: 174 2003 Minitab Inc. in which case the natural log is taken. Use overall standard deviation: Check to use the overall standard deviation in the estimation of the optimal lambda. click control chart Options.Quality Control With the first occurrence of these values: Choose to start a new stage at the first occurrence of a certain value. • To start a new stage at the first occurrence of a certain value. then enter the lambda values. You can also choose any value between − 5 and 5 for λ. use any dialog box options. (λ is lambda. then enter the values. Minitab only uses the non-omitted data to find lambda. Enclose date/time or text entries in double quotes. choose With the first occurrence of these values. when subgroups are across rows). > control chart Options > Box-Cox Charts All except the Attributes charts You can use the Box-Cox power transformation when your data are very skewed or when the within-subgroup variation is unstable to make the data "more normal. When executing the command. The Options subdialog box lists the common transformations natural log (λ = 0) and square root (λ = 0. The column must be the same length as the data column (or columns. you must set up a column of grouping indicators. Click the Stages tab. You may want to first run the command Stat > Control Charts > Box-Cox Transformation to help you find the optimal transformation value. Lambda = 0 (natural log): Choose to use the natural log of the data." The transformation takes the original data to the power λ. Caution If you use Stat > Control Charts > Box-Cox Transformation to find the optimal λ value and choose to store the transformed data with that command. Lambda=0. Check Box-Cox power transformation (W = Y**Lambda). Enter the values. Click the Box-Cox tab. enter the column which contains the stage indicators. Dialog box items Box-Cox power transformation (W = Y**Lambda): Choose to use when your data are very skewed or when the withinsubgroup variation is unstable.. The indicators can be numbers. choose With each new value.5 (square root): Choose to use the square root of the data.

and enter the number of subgroups. choose Display all subgroups. Dialog box items Store these estimates for each chart Means: Check to store the estimates of the means. Moving Average.. For help choosing a lambda value. Each segment contains __ subgroups: Choose to split the control chart into subgroups of a specified number of points. Do one of the following: • To split the chart into a specific number of subgroups. Click the Display tab. Choose Lambda=0. To display subgroups 1 2 In the control chart dialog box. Split chart into series of segments for display purposes Do not split: Choose not to split the chart. one row for each historical stage. X-bar and S. and enter the number of subgroups for each segment. Options − Display .. and enter the number of subgroups for each segment. then enter one or more values. Click OK. 175 . see the independent Box-Cox transformation command Stat > Control Charts > Box-Cox Transformation. Choose Optimal lambda for each stage (in a chart with stages) to have Minitab search for optimal values for each stage. X-bar. Do one of the following: • To display all subgroups in the control chart. Click OK. EWMA. 3 Options − Storage Stat > Control Charts > control chart type > control chart > control chart Options > Storage Charts X-bar and R. choose Display last __ subgroups and enter the number of subgroups to display. Each segment corresponds to a stage (if chart has stages): Choose to split the control chart by stage or stage when you have specified stages in control chart Options > Stage. 3 To split a control chart 1 2 In the control chart dialog box. • To display a specific number of the last subgroups. 3 Click OK.Control Charts • • • • • Choose Lambda = 0 (natural log) to use the natural log of the data. under Subgroups to display. CUSUM Use to store your choice of statistics in the worksheet. Choose Other (enter value(s) between -5 and 5) to transform the data using another lambda value or values. • To split the chart by stage.5 (square root) to use the square root of the data. Subgroups to display Display all subgroups: Choose to display all subgroups. click control chart Options. > control chart Options > Display Charts All charts except Z-MR Use to display the control chart by stage or by number of plotted points and to display the test results in the Session window. Test results Display test results in Session window: Check to display the test results in the Session window. Choose Optimal lambda to have Minitab search for an optimal value for all stages. click control chart Options. choose Each segment contains __ subgroups. 2003 Minitab Inc. choose Each segment corresponds to a stage (if chart has stages). Click the Display tab. Display last __ subgroups: Choose to display a specific number of subgroups. under Subgroups to display.

In Subgroup sizes. 8. Choose All observations for a chart are in one column. Session window output CUSUM Chart of AtoBDist Test Results for CUSUM Chart of AtoBDist TEST. parts of the crankshaft move up and down a certain distance from an ideal baseline position. 7. Choose Stat > Control Charts > Time-weighted Charts > CUSUM. 6. Stage: Check to store the stage. one row for each plotted point. 10 * WARNING * If graph is updated with new data. 1 2 3 4 Open the worksheet CRANKSH. To store statistics in the worksheet 1 2 Choose Stat > Control Charts > control chart type > control chart > control chart Options > Storage.MTW. 5. and then ten per day from the 18th through the 25th. Example of two one-sided CUSUM charts Suppose you work at a car assembly plant in a department that assembles engines. the results above may no * longer be correct. Check the statistics to store in the worksheet. then enter AtoBDist. one row for each plotted point. Test results: Check to store the results of any test performed. To ensure production quality. Control limit values: Check to store the control limit values. to look for small shifts away from the target. one row for each plotted point. 176 2003 Minitab Inc. enter 5. 9. You already displayed an X-bar chart and an R chart of this data. one row for each stage. Now. One point beyond control limits. Center line value: Check to store the center line value. On the X-bar chart. from September 28 through October 15. one row for each plotted point. then click OK. Minitab stores one column for the lower control limit and one column for the upper control limit. Click OK. Test Failed at points: 4.Quality Control Standard deviations: Check to store the estimates of the standard deviations. AtoBDist is the distance (in mm) from the actual (A) position of a point on the crankshaft to the baseline (B) position. . take five measurements each working day. one row for each plotted point. you want to plot the CUSUMs. Stores these values for each point Point plotted: Check to store the plotted points. Subgroup size: Check to store the subgroup sizes. one row for each plotted point. subgroup 5 failed a test for special causes. In an operating engine.

A single control limit determines whether the process is in control. Multivariate Charts Tsquared . A multivariate chart shows how several variables jointly influence a process or outcome. For example. the use of separate control charts is misleading because the variables jointly affect the process. However. create a correlation matrix of your variables. The distortion of these values increases with the number of measurement variables. Choosing a multivariate control chart Minitab offers four multivariate control charts: • • • • T2 and Generalized Variance − a T2 chart and Generalized Variance chart displayed in one window T2 − a multivariate chart of means Generalized Variance − a multivariate chart of process variability Multivariate EWMA − a multivariate chart of exponentially weighted moving averages 2003 Minitab Inc. If you use separate univariate control charts in a multivariate situation.Generalized Variance Chart Multivariate Control Charts Overview data Display statistics from two or more related measurement variables. you can assume the variables are correlated. see Control Charts Overview. Multivariate control charting has several advantages over creating multiple univariate charts: • • • The actual control region of the related variables is represented (elliptical for bivariate case). Type I error and the probability of a point correctly plotting in control are not equal to their expected values.Control Charts Graph window output Interpreting the results The CUSUMs for subgroups 4 through 10 fall outside the upper control limit. the scale on multivariate charts is unrelated to the scale of any of the variables. If the data include correlated variables. and out-of-control signals do not reveal which variable (or combination of variables) caused the signal. and it is appropriate to create a multivariate control chart.1. If the correlation coefficients are greater than 0. 177 . you can use multivariate control charts to investigate how the tensile strength and diameter of a fiber affect the quality of fabric. To determine whether you should use a univariate or multivariate chart. multivariate charts are more difficult to interpret than classic Shewhart control charts. For more information about control charts. You can maintain a specific Type I error. suggesting small shifts away from the target. For example.

using the following table as a guide. 178 2003 Minitab Inc. Variables charts include: • • • • Variables charts for subgroups Variables charts for individuals Time-weighted charts Multivariate charts Structure your data for these charts using the guidelines below. see Control Charts Overview and Multivariate Control Charts Overview. <Scale> <Labels> <Data Options> <Tsq-GV Options> Data − Variables Control Chart Organize the data for all variables control charts in the same way. with one column for each variable. Because Minitab displays both charts together. Multivariate data must be entered down columns. If the subgroups are not equal. . Worksheet Structure Structure your data down a column or across rows. structured both ways. with subgroups of size 5. For more information. Dialog box items Variables: Enter at least two columns of data.Quality Control T2 and Generalized Variance Chart Stat > Control Charts > Multivariate Charts > Tsquared-Generalized Variance The T2 and generalized variance chart is the multivariate counterpart to the Xbar-R. the second 5 observations are the second row. Here is the same data set. Note that the first five observations in the left data set (subgroup 1) are the first row of the right-side data set. Subgroups are equal size Univariate (one variable) Multivariate (more than one variable) Down columns or across rows Down columns Subgroups are unequal size Down columns with subgroup indicator column Down columns with subgroup indicator column Structure subgroup data down a column or across rows. the generalized variance chart appears in the bottom half. If the subgroup sizes do not vary much. Subgroup sizes: Enter a number or column of subscripts. Minitab draws the T2 chart in the upper half of the display. and so on. and I-MR charts. Xbar-S. you may want to force the control limits to be constant by specifying a fixed subgroup size using Tsq-GV Options > Estimate. each control limit is not a single straight line but varies with the subgroup size. You can also detect the presence of special causes. You can use this chart to simultaneously assess whether the process mean and variation are in control. you can track the process level and process variation simultaneously.

2003 Minitab Inc. If the data are highly skewed. subgroup 1 has three observations. For information on the Box-Cox transformation option. see Options − Box-Cox. you must enter your data in one column. Minitab uses the F distribution to test the significance of the out-of-control points. If you like. subgroup 2 has six observations. then create a second column of subscripts which serve as subgroup indicators. In Variables. you must enter data that approximate a normal distribution. you can use the transformation option to transform the data at the same time you draw the control chart. In Subgroup sizes. Minitab indicates which points are out of control in the Session window. Use the stand alone command as an exploratory tool to help you determine the best lambda value for the transformation. Then.Control Charts When subgroups are of unequal size. along with the p-value for each significant variable. Minitab can decompose the T2 statistic to find out each variable's contribution to the T2 statistic. However. To display a T2 and generalized variance chart 1 2 3 4 Choose Stat > Control Charts > Multivariate Charts > Tsquared-Generalized Variance. Missing data See Missing data in control charts for information on how to handle missing data for different types of control charts. or by using the stand alone Box-Cox command. enter at least two columns containing variables to control. See Example of T2 chart. 179 . you may want to use the Box-Cox transformation to induce normality. but it does not provide information on which variable or set of variables is out of control. see Box-Cox Transformation. use any dialog box options. You can access the Box-Cox transformation two ways: by using the Box-Cox transformation option provided with the control chart commands. then click OK. and so on. C1 contains the process data and C2 contains subgroup indicators: Each time a subscript changes in C2. enter a number or a column of subscripts. Nonnormal data To properly interpret Minitab's control charts. Decomposed T2 Statistic A T2 chart indicates when a process is out of control. For information on the stand alone Box-Cox transformation command. In the following example. a new subgroup begins in C1. In this example.

Number of samples used to estimate covariance: Enter the number of samples used to estimate the covariance matrix. then type individual subgroup numbers or a range of subgroups. if you have a known process parameter or an estimate obtained from past data. 180 2003 Minitab Inc. When you include or exclude rows using control chart options > Estimate and choose to perform a Box-Cox transformation.. If you enter only one matrix. one for each variable. Use the following samples when estimating parameters: Choose to specify subgroups of data to compute µ and Σ. Each stage for each variable uses a different value. Note When you include or exclude rows using control chart > Data Options > Subset and also omit or use samples to estimate parameters. In Mean. one for each stage within each variable. if some subgroups have erratic data due to assignable causes that you have corrected. The number of samples must be equal to the number of matrices.Quality Control You can store the decomposed T2 values. you can prevent these subgroups from influencing the estimation of process parameters. all stages use that covariance matrix. You can omit certain subgroups when estimating parameters using Estimate. you must enter subgroup 2 as the subgroup to omit. enter one or more values. you can enter this value or values. To omit one observation from a subgroup. then type individual subgroup numbers or ranges of subgroups. For example. Enter entire subgroups. For example. Dialog box items Omit the following samples when estimating parameters: Choose to specify subgroups of the data to omit from the computations for µ and Σ. enter the same number of matrices and samples as stages. otherwise. > control chart Options > Parameters Charts T2 and Generalized Variance. which applies to all variables and stages. Covariance matrix Matrix: Enter one or more matrices for estimating the covariance matrix. If you want each stage to use its own covariance matrix. not just the subsetted data. To use a historical mean and covariance matrix 1 2 3 4 • • • Choose Stat > Control Charts > control chart type > control chart > control chart Options > Parameters. then decide to omit subgroup 2 (using control chart Options > Estimate). not individual observations. In Number of samples used to calculate covariance. Note Minitab can perform decomposition only when there are two or more observations in an stage. If you do not specify a value for the mean or the covariance matrix. otherwise. Options − Parameters .. Click OK. T2. the omitted or used samples apply to the entire data set..j) must equal element (j. You can enter the historical mean values in several ways: Options − Estimate . Minitab only uses the non-omitted data to find lambda. A single value. enter a value for each stage.. Because Minitab considers an omitted sample (subgroup 1) as part of the entire data set. . in Matrix. all stages use that value. enter a matrix for each stage. change that value to missing (∗) in the worksheet. Under Covariance matrix.i). > control chart Options > Estimates Charts T2 and Generalized Variance. For example. you exclude subgroup 1 from your analysis (using control chart > Data Options > Subset). A list of values. Dialog box items Mean: Enter the historical µ's as a single value or a list of values. All stages for each variable use the same value. Minitab estimates it from the data. enter one or more matrices. Element (i. Each matrix must be a positive definite square matrix with the same number of rows and columns as columns of data. type the number of samples. A list of values. The elements on the matrix diagonal must be positive. MEWMA Omits or includes certain subgroups to estimate µ and Σ. T2 Enter historical data for estimating µ and Σ. If you enter one constant.

Upper confidence limit bound: Check to set the upper confidence limit bound. • Choose Use the following samples when estimating parameters and enter the subgroups or observation numbers you want to use for the calculations. Minitab estimates the process parameters from all the data. check Upper confidence limit bound.Control Charts To omit or use subgroups for the estimates of parameters By default. Minitab assumes the values you enter are subgroup numbers. • Under Place bounds on confidence limits of Generalized variance chart. 3 Options − Stages . But you may want to use or omit certain data if they show abnormal behavior. If the calculated lower confidence limit is less than the lower bound. click control chart Options.. 181 . then type the lower limit (must be greater than 0) for the T2 chart. Dialog box items Display confidence limits at These percents: Enter one or more values or enter a column of values. • Under Place bounds on confidence limits of Tsquared chart. If the calculated upper confidence limit is greater than the upper bound. Minitab draws a horizontal line labeled LB at the lower bound instead. Click the Confidence Limits tab. click Tsq-GV Options. With these charts. > control chart Options > Stages Charts All charts except Z-MR 2003 Minitab Inc. Click OK. except with the Individuals. Place bounds on confidence limits of Generalized variance chart Lower confidence limit bound: Check to set the lower confidence limit bound. then type the upper limit (must be greater than 0) for the generalized variance chart. Moving Range. Use proportion values (between 0 and 1) or percent values (between 1 and 100). Minitab draws a horizontal line labeled UB at the upper bound instead. Upper confidence limit bound: Check to set the upper confidence limit bound. T2 and Generalized Variance Chart − Options − Confidence Limits Stat > Control Charts > Multivariate Charts > Tsquared-Generalized Variance > Tsq-GV Options > Confidence Limits Specifies the confidence limit positions for the T2 and generalized variance charts. If the calculated lower confidence limit is less than the lower bound. 1 2 In the control chart dialog box. Minitab draws a horizontal line labeled UB at the upper bound instead. check Lower confidence limit bound. check Upper confidence limit bound. then type the lower confidence limit (greater than 0). then type the lower limit (must be greater than 0) for the generalized variance chart. check Lower confidence limit bound. then type the upper confidence limit (greater than 0). To specify confidence limits 1 2 In control chart data source dialog box. Enter proportion values (between 0 and 1) or percent values (between 1 and 100). Click the Estimate tab. Do one of the following: • Choose Omit the following samples when estimating parameters and enter the subgroups or observation numbers you want to omit from the calculations. enter one or more values or enter a column of values in These percents. Note 3 Click OK. • Under Place bounds on confidence limits of Tsquared chart. • Under Place bounds on confidence limits of Generalized variance chart. then type the upper confidence limit (greater than 0). then type the lower confidence limit (greater than 0). Place bounds on confidence limits of Tsquared chart Lower confidence limit bound: Check to set the lower confidence limit bound. Minitab draws a horizontal line labeled LB at the lower bound instead. the values are interpreted as observation numbers. and I-MR charts. then type the upper limit (must be greater than 0) for the T2 chart. Do any of the following: • Under Display confidence limits at. If the calculated upper confidence limit is greater than the upper bound..

do not select the Box-Cox option when you make a control char. In most cases. To display a historical chart To define stages in your process. or text. then click OK. you can tell Minitab to start a new stage in one of two ways: • • 1 2 3 4 Each time the value in the column changes At the first occurrence of one or more values In the control chart dialog box. Optimal lambda: Choose to have Minitab search for an optimal value. Historical charts are particularly useful for comparing data before and after a process improvement. X-bar and R. Enclose date/time or text entries in double quotes. When you include or exclude rows using control chart Options > Estimate. Caution If you use Stat > Control Charts > Box-Cox Transformation to find the optimal λ value and choose to store the transformed data with that command. then enter the values. Under When to start a new value. Note With the following charts. in which case the natural log is taken. When you use this transformation. The Options subdialog box lists the common transformations natural log (λ = 0) and square root (λ = 0. Enter the values. do one of the following: • To start a new stage each time the value in the column changes. use any dialog box options. In Define stages (historical groups) with this variable.Quality Control You can display stages in your process by drawing a "historical chart" − a control chart in which the control limits and center line are estimated independently for different groups in your data. Click the Stages tab. choose With the first occurrence of these values. choose With each new value. You can enter the same value more than once. Minitab treats each repeated value as a separate occurrence... you should not choose a λ outside the range of −2 and 2. 182 2003 Minitab Inc. and X-bar and S. Minitab treats each repeated value as a separate occurrence. click control chart Options. Dialog box items Box-Cox power transformation (W = Y**Lambda): Choose to use when your data are very skewed or when the withinsubgroup variation is unstable. Lambda = 0 (natural log): Choose to use the natural log of the data. Enclose date/time or text entries in double quotes. The indicators can be numbers. (λ is lambda. Dialog box items Define stages (historical groups) with this variable: Enter the column that contains the stage indicators. Minitab does not accept any values you enter in the Parameters tab for historical means or historical standard deviations.) To use this option. dates. When to start a new stage With each new value: Choose to start a new stage each time the value in the column changes. enter the column which contains the stage indicators. Minitab only uses the non-omitted data to find lambda. you must have at least one subgroup with two or more observations: R. • To start a new stage at the first occurrence of a certain value. S. You can enter the same value more than once. With the first occurrence of these values: Choose to start a new stage at the first occurrence of a certain value. When executing the command. You may want to first run the command Stat > Control Charts > Box-Cox Transformation to help you find the optimal transformation value. you must set up a column of grouping indicators. 5 Options − Box-Cox . You can also choose any value between − 5 and 5 for λ. Lambda=0.5).5 (square root): Choose to use the square root of the data. If you like. unless λ = 0. The column must be the same length as the data column (or columns. when subgroups are across rows). the data must be positive. . > control chart Options > Box-Cox Charts All except the Attributes charts You can use the Box-Cox power transformation when your data are very skewed or when the within-subgroup variation is unstable to make the data "more normal." The transformation takes the original data to the power λ. doing so will double-transform the data.

choose Each segment corresponds to a stage (if chart has stages). click control chart Options. then do one of the following: • Choose Lambda = 0 (natural log) to use the natural log of the data. choose Display last __ subgroups and enter the number of subgroups to display. • To display a specific number of the last subgroups. click control chart Options.. then enter one or more values. Each segment corresponds to a stage (if chart has stages): Choose to split the control chart by stage or stage when you have specified stages in control chart Options > Stage. choose Each segment contains __ subgroups. Click OK. For help choosing a lambda value. Do one of the following: • To split the chart into a specific number of subgroups. Use overall standard deviation: Check to use the overall standard deviation in the estimation of the optimal lambda. Each segment contains __ subgroups: Choose to split the control chart into subgroups of a specified number of points. click control chart Options.. Do one of the following: • To display all subgroups in the control chart. • Choose Lambda=0. 3 Click OK. • Choose Optimal lambda for each stage (in a chart with stages) to have Minitab search for optimal values for each stage. then enter the lambda values. Click OK. Options − Display . Other (enter value(s) between -5 and 5): Choose to transform the data using another lambda value or values. To display subgroups 1 2 In the control chart dialog box.5 (square root) to use the square root of the data. 3 2003 Minitab Inc. choose Display all subgroups. and enter the number of subgroups for each segment. Click the Box-Cox tab. under Subgroups to display. To do the Box-Cox power transformation 1 2 In the control chart dialog box. Display last __ subgroups: Choose to display a specific number of subgroups. • Choose Optimal lambda to have Minitab search for an optimal value for all stages. Click the Display tab. under Subgroups to display. Click the Display tab. Test results Display test results in Session window: Check to display the test results in the Session window. 3 To split a control chart 1 2 In the control chart dialog box. and enter the number of subgroups for each segment. • Choose Other (enter value(s) between -5 and 5) to transform the data using another lambda value or values. see the independent Box-Cox transformation command Stat > Control Charts > Box-Cox Transformation. > control chart Options > Display Charts All charts except Z-MR Use to display the control chart by stage or by number of plotted points and to display the test results in the Session window. Subgroups to display Display all subgroups: Choose to display all subgroups. 183 .Control Charts Optimal lambda for each stage (in a chart with stages): Choose to have Minitab search for optimal values for each stage. • To split the chart by stage. and enter the number of subgroups. Split chart into series of segments for display purposes Do not split: Choose not to split the chart. Check Box-Cox power transformation (W = Y**Lambda).

enter Stay Satisfaction. Because satisfaction and length of stay are correlated.MTW. Click OK. Subgroup size: Check to store the subgroup sizes. Test results: Check to store the results of any test performed. Satisfaction Test Results for Tsquared Chart of Stay. then click OK. Covariance matrices: Check to store the estimated covariance matrices. -1 (beyond lower control limit). . Minitab assigns each plotted point one of three values: 1 (beyond upper control limit). You want to create T2 and generalized variance charts to simultaneously monitor the mean and variability of satisfactions ratings and length of stay. one row for each plotted point.0010 0. To store statistics in the worksheet 1 2 Choose Stat > Control Charts > control chart type > control chart > control chart Options > Storage. Stage: Check to store the stage. Control limit values: Check to store the control limit values. one row for each plotted point. Check the statistics to store in the worksheet. the results above may no * longer be correct. Dialog box items Store these estimates for each chart Means: Check to store the estimates of the means. 184 2003 Minitab Inc. 1 2 3 4 Open the worksheet HOSPITAL. enter Departure. one row for each plotted point. In Variables. one row for each historical stage.0000 * WARNING * If graph is updated with new data. T2 and Generalized Variance Use to store your choice of statistics in the worksheet. Satisfaction Greater Than UCL Point 18 19 Variable Stay Satisfaction Satisfaction P-Value 0. one row for each historical stage. Store these values for each point Point plotted: Check to store the plotted points. Choose Stat > Control Charts > Multivariate Charts > Tsquared-Generalized Variance. you previously created a T2 chart to monitor satisfaction ratings (on scale of 1− 7) and length of stay (in days). Example of T2 and generalized variance chart You are a hospital manager interested in monitoring patient satisfaction ratings through the month of January. You randomly ask 5 patients each day to complete a short questionnaire about their stay at the hospital before they check out. Center line value: Check to store the center line value. Decomposed Tsquared values: Check to store the decomposed T2 values. Session window output Tsquared-Generalized Variance Chart of Stay. In Subgroup sizes. one row for each plotted point.Quality Control Options − Storage Stat > Control Charts > control chart type > control chart > control chart Options > Storage Charts T2.0072 0. 0 (in control). Minitab stores one column for the lower control limit and one column for the upper control limit. one row for each plotted point. one row for each plotted point.

Tsquared Chart Multivariate Control Charts Overview data Display statistics from two or more related measurement variables. The Session window results indicate that each variable contributes significantly to January 18. If the data include correlated variables. you can use multivariate control charts to investigate how the tensile strength and diameter of a fiber affect the quality of fabric. and out-of-control signals do not reveal which variable (or combination of variables) caused the signal. and it is appropriate to create a multivariate control chart. If you use separate univariate control charts in a multivariate situation. multivariate charts are more difficult to interpret than classic Shewhart control charts. see Control Charts Overview. A multivariate chart shows how several variables jointly influence a process or outcome. The distortion of these values increases with the number of measurement variables. create a correlation matrix of your variables. The generalized variance chart shows that the joint variability in length of stay and satisfaction ratings remains in control over the 30-day period. However. To determine whether you should use a univariate or multivariate chart. You can maintain a specific Type I error. However. 185 .Control Charts Graph window output Interpreting the results The T2 chart shows two out-of-control points: January 18 and 19. the scale on multivariate charts is unrelated to the scale of any of the variables. the use of separate control charts is misleading because the variables jointly affect the process. If the correlation coefficients are greater than 0. For example. Multivariate control charting has several advantages over creating multiple univariate charts: • • • The actual control region of the related variables is represented (elliptical for bivariate case). Satisfaction is the only variable that is out of control for January 19. You should investigate what special causes − perhaps hospital heating problems or interrupted food service − may have affected length of stay and satisfaction rating for those two days. A single control limit determines whether the process is in control. Choosing a multivariate control chart Minitab offers four multivariate control charts: • • • T2 and Generalized Variance − a T2 chart and Generalized Variance chart displayed in one window T2 − a multivariate chart of means Generalized Variance − a multivariate chart of process variability 2003 Minitab Inc.1. you can assume the variables are correlated. For more information about control charts. For example. Type I error and the probability of a point correctly plotting in control are not equal to their expected values.

. Before using a multivariate control chart. each control limit is not a single straight line but varies with the subgroup size. Its purpose is to assess whether several variables are simultaneously in control. structured both ways. and so on. temperature. Multivariate data must be entered down columns. If you assume the variables are independent. see Control Charts Overview and Multivariate Control Charts Overview. you can simultaneously measure three variables: weight. If the subgroups are not equal. then you do not need a multivariate control chart. <Scale> <Labels> <Data Options> <Tsquared Options> Data − Variables Control Chart Organize the data for all variables control charts in the same way. with one column for each variable. using the following table as a guide. the second 5 observations are the second row. with subgroups of size 5. and percentage of polyester. For example. Variables charts include: • • • • Variables charts for subgroups Variables charts for individuals Time-weighted charts Multivariate charts Structure your data for these charts using the guidelines below. Here is the same data set. Dialog box items Variables: Enter at least two columns of data. 186 2003 Minitab Inc. you may want to force the control limits to be constant by specifying a fixed subgroup size using Tsquared Options > Estimate. Subgroup sizes: Enter a number or a column of subscripts. For more information. you need to make sure that a dependence exists between the variables. Note that the first five observations in the left data set (subgroup 1) are the first row of the right-side data set. to monitor the quality of automobile tire casings. If the subgroup sizes do not vary much. Subgroups are equal size Univariate (one variable) Multivariate (more than one variable) Down columns or across rows Down columns Subgroups are unequal size Down columns with subgroup indicator column Down columns with subgroup indicator column Structure subgroup data down a column or across rows.Quality Control • Multivariate EWMA − a multivariate chart of exponentially weighted moving averages T2 Chart Stat > Control Charts > Multivariate Charts > Tsquared The T2 control chart is the multivariate version of the X-bar chart (or the Individuals chart for individual observations). Worksheet Structure Structure your data down a column or across rows.

see Box-Cox Transformation. For information on the Box-Cox transformation option. Minitab uses the F distribution to test the significance of the out-of-control points. However. along with the p-value for each significant variable. then click OK. 187 . you must enter your data in one column. In the following example. you may want to use the Box-Cox transformation to induce normality. subgroup 1 has three observations. then create a second column of subscripts which serve as subgroup indicators. In this example. In Variables. enter a number or a column of subscripts. Decomposed T2 Statistic A T2 chart indicates when a process is out of control. You can access the Box-Cox transformation two ways: by using the Box-Cox transformation option provided with the control chart commands. Missing data See Missing data in control charts for information on how to handle missing data for different types of control charts. C1 contains the process data and C2 contains subgroup indicators: Each time a subscript changes in C2. you can use the transformation option to transform the data at the same time you draw the control chart. In Subgroup sizes. Minitab indicates which points are out of control in the Session window. 2003 Minitab Inc. Nonnormal data To properly interpret Minitab's control charts. or by using the stand alone Box-Cox command. a new subgroup begins in C1. Use the stand alone command as an exploratory tool to help you determine the best lambda value for the transformation. See Example of T2 chart. If you like. and so on. you must enter data that approximate a normal distribution.Control Charts When subgroups are of unequal size. Then. Minitab can decompose the T2 statistic to find out each variable's contribution to the T2 statistic. If the data are highly skewed. but it does not provide information on which variable or set of variables is out of control. subgroup 2 has six observations. To display a Tsquared chart 1 2 3 4 Choose Stat > Control Charts > Multivariate Charts > Tsquared. For information on the stand alone Box-Cox transformation command. enter at least two columns containing variables to control. use any dialog box options. see Options − Box-Cox.

Note When you include or exclude rows using control chart > Data Options > Subset and also omit or use samples to estimate parameters. To omit one observation from a subgroup. enter one or more values. T2 Enter historical data for estimating µ and Σ.. one for each variable. 188 2003 Minitab Inc.Quality Control You can store the decomposed T2 values. The elements on the matrix diagonal must be positive.j) must equal element (j. The number of samples must be equal to the number of matrices. For example. not individual observations. T2. not just the subsetted data. otherwise. A list of values. in Matrix. Use the following samples when estimating parameters: Choose to specify subgroups of data to compute µ and Σ. If you enter only one matrix. the omitted or used samples apply to the entire data set. Options − Parameters . A list of values. You can enter the historical mean values in several ways: Options − Estimate . Covariance matrix Matrix: Enter one or more matrices for estimating the covariance matrix. MEWMA Omits or includes certain subgroups to estimate µ and Σ. Click OK. which applies to all variables and stages. you exclude subgroup 1 from your analysis (using control chart > Data Options > Subset). In Mean. enter a value for each stage. otherwise. To use a historical mean and covariance matrix 1 2 3 4 • • • Choose Stat > Control Charts > control chart type > control chart > control chart Options > Parameters. If you want each stage to use its own covariance matrix. all stages use that value. you can prevent these subgroups from influencing the estimation of process parameters. then decide to omit subgroup 2 (using control chart Options > Estimate). Dialog box items Omit the following samples when estimating parameters: Choose to specify subgroups of the data to omit from the computations for µ and Σ.i). Under Covariance matrix. If you enter one constant. You can omit certain subgroups when estimating parameters using Estimate.. you can enter this value or values. Note Minitab can perform decomposition only when there are two or more observations in an stage. > control chart Options > Estimates Charts T2 and Generalized Variance. change that value to missing (∗) in the worksheet. if you have a known process parameter or an estimate obtained from past data. All stages for each variable use the same value. For example. . Number of samples used to estimate covariance: Enter the number of samples used to estimate the covariance matrix. Enter entire subgroups. Dialog box items Mean: Enter the historical µ's as a single value or a list of values. Because Minitab considers an omitted sample (subgroup 1) as part of the entire data set. For example. enter the same number of matrices and samples as stages. If you do not specify a value for the mean or the covariance matrix. A single value. Element (i.. one for each stage within each variable. Minitab estimates it from the data. Each matrix must be a positive definite square matrix with the same number of rows and columns as columns of data. then type individual subgroup numbers or a range of subgroups. In Number of samples used to calculate covariance. type the number of samples. enter a matrix for each stage. you must enter subgroup 2 as the subgroup to omit.. Each stage for each variable uses a different value. Minitab only uses the non-omitted data to find lambda. enter one or more matrices. then type individual subgroup numbers or ranges of subgroups. > control chart Options > Parameters Charts T2 and Generalized Variance. all stages use that covariance matrix. When you include or exclude rows using control chart options > Estimate and choose to perform a Box-Cox transformation. if some subgroups have erratic data due to assignable causes that you have corrected.

Minitab estimates the process parameters from all the data. Click OK. Note 3 Click OK. Options − Confidence Limits . When to start a new stage With each new value: Choose to start a new stage each time the value in the column changes. 2003 Minitab Inc. check Lower confidence limit bound. Place bounds on confidence limits Lower confidence limit bound: Check to set the lower confidence limit bound. • Choose Use the following samples when estimating parameters and enter the subgroups or observation numbers you want to use for the calculations. Minitab assumes the values you enter are subgroup numbers. But you may want to use or omit certain data if they show abnormal behavior... • Under Place bounds on confidence limits. Click the Estimate tab. X-bar and R. Minitab draws a horizontal line labeled LB at the lower bound instead. Do any of the following: • Under Display confidence limits at. Moving Range. > control chart Options > Stages Charts All charts except Z-MR You can display stages in your process by drawing a "historical chart" − a control chart in which the control limits and center line are estimated independently for different groups in your data. and X-bar and S.Control Charts To omit or use subgroups for the estimates of parameters By default. Generalized Variance Specifies the confidence limit positions. 189 . you must have at least one subgroup with two or more observations: R. check Upper confidence limit bound. Note With the following charts. click control chart Options. Minitab draws a horizontal line labeled UB at the upper bound instead. Historical charts are particularly useful for comparing data before and after a process improvement. Use proportion values (between 0 and 1) or percent values (between 1 and 100). then type the lower limit (must be greater than 0) for the T2 chart. Dialog box items Display confidence limits at These percents: Enter one or more values or enter a column of values. • Under Place bounds on confidence limits. If the calculated upper confidence limit is greater than the upper bound. except with the Individuals. > control chart Options > Confidence Limits Charts T2. enter one or more values or enter a column of values in These percents. 3 Options − Stages . then type the lower confidence limit (greater than 0). Do one of the following: • Choose Omit the following samples when estimating parameters and enter the subgroups or observation numbers you want to omit from the calculations. With these charts. If the calculated lower confidence limit is less than the lower bound. and I-MR charts. To specify confidence limits 1 2 In control chart dialog box. Enter proportion values (between 0 and 1) or percent values (between 1 and 100). 1 2 In the control chart dialog box. S. Click the Confidence Limits tab. Upper confidence limit bound: Check to set the upper confidence limit bound.. then type the upper limit (must be greater than 0) for the T2 chart. the values are interpreted as observation numbers.. Dialog box items Define stages (historical groups) with this variable: Enter the column that contains the stage indicators. click control chart Options. then type the upper confidence limit (greater than 0).

Lambda = 0 (natural log): Choose to use the natural log of the data. when subgroups are across rows). enter the column which contains the stage indicators. then do one of the following: 190 2003 Minitab Inc. then click OK. Enter the values. you should not choose a λ outside the range of −2 and 2. The Options subdialog box lists the common transformations natural log (λ = 0) and square root (λ = 0. unless λ = 0. To do the Box-Cox power transformation 1 2 In the control chart dialog box. Minitab treats each repeated value as a separate occurrence. you must set up a column of grouping indicators. When you include or exclude rows using control chart Options > Estimate. Check Box-Cox power transformation (W = Y**Lambda). dates.5 (square root): Choose to use the square root of the data.) To use this option. To display a historical chart To define stages in your process. in which case the natural log is taken. Dialog box items Box-Cox power transformation (W = Y**Lambda): Choose to use when your data are very skewed or when the withinsubgroup variation is unstable. Enclose date/time or text entries in double quotes. doing so will double-transform the data. choose With each new value. (λ is lambda. You may want to first run the command Stat > Control Charts > Box-Cox Transformation to help you find the optimal transformation value. Enclose date/time or text entries in double quotes. Under When to start a new value. Optimal lambda for each stage (in a chart with stages): Choose to have Minitab search for optimal values for each stage. > control chart Options > Box-Cox Charts All except the Attributes charts You can use the Box-Cox power transformation when your data are very skewed or when the within-subgroup variation is unstable to make the data "more normal. The column must be the same length as the data column (or columns. When executing the command. the data must be positive. When you use this transformation. In Define stages (historical groups) with this variable. you can tell Minitab to start a new stage in one of two ways: • • 1 2 3 4 Each time the value in the column changes At the first occurrence of one or more values In the control chart dialog box. or text. Minitab treats each repeated value as a separate occurrence. 5 Options − Box-Cox . The indicators can be numbers. click control chart Options. If you like. • To start a new stage at the first occurrence of a certain value. use any dialog box options." The transformation takes the original data to the power λ. Use overall standard deviation: Check to use the overall standard deviation in the estimation of the optimal lambda. Caution If you use Stat > Control Charts > Box-Cox Transformation to find the optimal λ value and choose to store the transformed data with that command. In most cases. click control chart Options. . Optimal lambda: Choose to have Minitab search for an optimal value. choose With the first occurrence of these values. then enter the values. Other (enter value(s) between -5 and 5): Choose to transform the data using another lambda value or values. Lambda=0.Quality Control With the first occurrence of these values: Choose to start a new stage at the first occurrence of a certain value. You can also choose any value between − 5 and 5 for λ. do not select the Box-Cox option when you make a control char. Minitab does not accept any values you enter in the Parameters tab for historical means or historical standard deviations. You can enter the same value more than once. Click the Box-Cox tab.5). then enter the lambda values. do one of the following: • To start a new stage each time the value in the column changes.. You can enter the same value more than once. Click the Stages tab.. Minitab only uses the non-omitted data to find lambda.

then enter one or more values. Dialog box items Store these estimates for each chart Means: Check to store the estimates of the means. Test results Display test results in Session window: Check to display the test results in the Session window. Click the Display tab. see the independent Box-Cox transformation command Stat > Control Charts > Box-Cox Transformation. To display subgroups 1 2 In the control chart dialog box. Choose Optimal lambda to have Minitab search for an optimal value for all stages. choose Each segment corresponds to a stage (if chart has stages). Do one of the following: • To split the chart into a specific number of subgroups. > control chart Options > Display Charts All charts except Z-MR Use to display the control chart by stage or by number of plotted points and to display the test results in the Session window. click control chart Options. under Subgroups to display. Click OK. Subgroups to display Display all subgroups: Choose to display all subgroups. Each segment corresponds to a stage (if chart has stages): Choose to split the control chart by stage or stage when you have specified stages in control chart Options > Stage. For help choosing a lambda value.Control Charts • • • • • Choose Lambda = 0 (natural log) to use the natural log of the data. Options − Display . choose Display all subgroups. • To split the chart by stage. 3 Click OK. 3 To split a control chart 1 2 In the control chart dialog box. and enter the number of subgroups for each segment. Split chart into series of segments for display purposes Do not split: Choose not to split the chart. choose Display last __ subgroups and enter the number of subgroups to display.5 (square root) to use the square root of the data. click control chart Options.. 3 Options − Storage Stat > Control Charts > control chart type > control chart > control chart Options > Storage Charts T2. and enter the number of subgroups for each segment.. Click the Display tab. choose Each segment contains __ subgroups. 2003 Minitab Inc. Click OK. Each segment contains __ subgroups: Choose to split the control chart into subgroups of a specified number of points. • To display a specific number of the last subgroups. T2 and Generalized Variance Use to store your choice of statistics in the worksheet. Do one of the following: • To display all subgroups in the control chart. Choose Lambda=0. and enter the number of subgroups. under Subgroups to display. Choose Optimal lambda for each stage (in a chart with stages) to have Minitab search for optimal values for each stage. 191 . Choose Other (enter value(s) between -5 and 5) to transform the data using another lambda value or values. one row for each historical stage. Display last __ subgroups: Choose to display a specific number of subgroups.

one row for each plotted point. Center line value: Check to store the center line value. Session window output Tsquared Chart of Stay. Satisfaction Greater Than UCL Point 18 19 Variable Stay Satisfaction Satisfaction P-Value 0. you create a T2 chart to simultaneously monitor satisfactions ratings (on scale of 1−7) and length of stay (in days). the results above may no * longer be correct. 192 2003 Minitab Inc. one row for each plotted point. In Variables. You randomly ask 5 patients each day to complete a short questionnaire about their stay at the hospital before they check out. one row for each historical stage. 0 (in control). Subgroup size: Check to store the subgroup sizes. Minitab stores one column for the lower control limit and one column for the upper control limit. Control limit values: Check to store the control limit values. enter Stay Satisfaction.Quality Control Covariance matrices: Check to store the estimated covariance matrices. Because satisfaction and length of stay are correlated. Test results: Check to store the results of any test performed. In Subgroup sizes. then click OK. Minitab assigns each plotted point one of three values: 1 (beyond upper control limit). Choose Stat > Control Charts > Multivariate Charts > Tsquared. 1 2 3 4 Open the worksheet HOSPITAL. then click OK. Stage: Check to store the stage. Decomposed Tsquared values: Check to store the decomposed T2 values.0010 0. one row for each plotted point. -1 (beyond lower control limit). . Satisfaction Test Results for Tsquared Chart of Stay. Store these values for each point Point plotted: Check to store the plotted points.MTW.0000 * WARNING * If graph is updated with new data. one row for each plotted point. one row for each plotted point. To store statistics in the worksheet 1 2 Choose Stat > Control Charts > control chart type > control chart > control chart Options > Storage. Example of T2 chart You are a hospital manager interested in monitoring patient satisfaction ratings through the month of January.0072 0. one row for each plotted point. Check the statistics to store in the worksheet. enter a number or a column of subscripts.

the use of separate control charts is misleading because the variables jointly affect the process. The distortion of these values increases with the number of measurement variables. However. Multivariate control charting has several advantages over creating multiple univariate charts: • • • The actual control region of the related variables is represented (elliptical for bivariate case). A multivariate chart shows how several variables jointly influence a process or outcome. see Control Charts Overview. You should investigate what special causes − perhaps hospital heating problems or interrupted food service − may have affected length of stay and satisfaction rating for those 2 days. Choosing a multivariate control chart Minitab offers four multivariate control charts: • • • • T2 and Generalized Variance − a T2 chart and Generalized Variance chart displayed in one window T2 − a multivariate chart of means Generalized Variance − a multivariate chart of process variability Multivariate EWMA − a multivariate chart of exponentially weighted moving averages 2003 Minitab Inc. 193 . If you use separate univariate control charts in a multivariate situation. you can use multivariate control charts to investigate how the tensile strength and diameter of a fiber affect the quality of fabric. multivariate charts are more difficult to interpret than classic Shewhart control charts.Control Charts Graph window output Interpreting the results Two points are out of control: January 18 and 19. For more information about control charts. However. Generalized Variance Multivariate Control Charts Overview data Display statistics from two or more related measurement variables. To determine whether you should use a univariate or multivariate chart. and out-of-control signals do not reveal which variable (or combination of variables) caused the signal. you can assume the variables are correlated. The Session window results indicate that each variable contributes significantly to January 18. You can maintain a specific Type I error. the scale on multivariate charts is unrelated to the scale of any of the variables. Type I error and the probability of a point correctly plotting in control are not equal to their expected values. Satisfaction is the only variable that is out of control for January 19. and it is appropriate to create a multivariate control chart. If the data include correlated variables. A single control limit determines whether the process is in control. For example. For example. create a correlation matrix of your variables. If the correlation coefficients are greater than 0.1.

. using the following table as a guide. For more information. Subgroup sizes: Enter a number or a column of subscripts. Note that the first five observations in the left data set (subgroup 1) are the first row of the right-side data set. You can use it to simultaneously monitor the process variability of two or more related process characteristics. <Scale> <Labels> <Data Options> <Gen Var Options> Data − Variables Control Chart Organize the data for all variables control charts in the same way. structured both ways. Minitab plots the subgroups of n = 1 as missing values. Variables charts include: • • • • Variables charts for subgroups Variables charts for individuals Time-weighted charts Multivariate charts Structure your data for these charts using the guidelines below. the second 5 observations are the second row. 194 2003 Minitab Inc. Subgroups are equal size Univariate (one variable) Multivariate (more than one variable) Down columns or across rows Down columns Subgroups are unequal size Down columns with subgroup indicator column Down columns with subgroup indicator column Structure subgroup data down a column or across rows. in a medical trial of a new drug. with one column for each variable. For example. Dialog box items Variables: Enter at least two columns of data. with subgroups of size 5. and so on. you may want to force the control limits to be constant by specifying a fixed subgroup size using Gen Var Options > Estimate. If the subgroup sizes do not vary much. see Control Charts Overview and Multivariate Control Charts Overview. you can monitor heart rate and blood pressure of the participants to see if these variables jointly remain constant over the course of the treatment. Worksheet Structure Structure your data down a column or across rows.Quality Control Generalized Variance Chart Stat > Control Charts > Multivariate Charts > Generalized Variance The multivariate form of the S chart. Here is the same data set. If some subgroups of n = 1 and others > 1. Multivariate data must be entered down columns.

Use the stand alone command as an exploratory tool to help you determine the best lambda value for the transformation. Missing data See Missing data in control charts for information on how to handle missing data for different types of control charts. you must enter your data in one column. If you do not specify a value for the covariance matrix. The elements on the matrix diagonal must be positive. then click OK. If you enter only one matrix. Generalized Variance Chart − Options – Parameters Stat > Control Charts > Multivariate Charts > Generalized Variance > Gen Var Options > Parameters Use to enter historical data for estimating Σ. In the following example.i). Nonnormal data To properly interpret Minitab's control charts. For example. You can omit certain subgroups when estimating parameters using Estimate. then click OK. see Options − Box-Cox. If you like. Each matrix must be a positive definite square matrix with the same number of rows and columns as columns of data. Element (i. For information on the stand alone Box-Cox transformation command. otherwise. you can use the transformation option to transform the data at the same time you draw the control chart.j) must equal element (j. In Covariance matrix. use any dialog box options. 2003 Minitab Inc. To display a generalized variance chart 1 2 3 4 Choose Stat > Control Charts > Multivariate Charts > Generalized Variance In Variables. if you have a known process parameter or an estimate obtained from past data. enter one or more matrices. or by using the stand alone Box-Cox command. you may want to use the Box-Cox transformation to induce normality. enter a number or a column of subscripts. You can access the Box-Cox transformation two ways: by using the Box-Cox transformation option provided with the control chart commands. then create a second column of subscripts which serve as subgroup indicators. For information on the Box-Cox transformation option. Dialog box item Covariance matrix Enter one or more matrices for estimating the covariance matrix. Click Gen Var Options. you must enter data that approximate a normal distribution. you can enter this value or values. enter at least two columns containing variables to control. 195 . a new subgroup begins in C1. all stages use that covariance matrix. Then. enter a matrix for each stage. subgroup 1 has three observations. subgroup 2 has six observations. C1 contains the process data and C2 contains subgroup indicators: Each time a subscript changes in C2. Minitab estimates it from the data. see Box-Cox Transformation. If the data are highly skewed. To use a historical covariance matrix 1 2 Choose Stat > Control Charts > Multivariate Charts > Generalized Variance. and so on. In Subgroup sizes.Control Charts When subgroups are of unequal size. In this example.

. the values are interpreted as observation numbers. Use proportion values (between 0 and 1) or percent values (between 1 and 100). Dialog box items Omit or Use Omit the following samples when estimating parameters (eg.. 1 2 In the control chart dialog box. then type the lower limit (must be greater than 0) for the T2 chart. Because Minitab considers an omitted sample (subgroup 1) as part of the entire data set. then type individual subgroup numbers or a range of subgroups. Minitab only uses the non-omitted data to find lambda. Moving Range. For example. then type individual subgroup numbers or ranges of subgroups. then decide to omit subgroup 2 from the parameter estimation (using control chart Options > Estimate). Place bounds on confidence limits Lower confidence limit bound: Check to set the lower confidence limit bound. Minitab estimates the process parameters from all the data. the omitted or used samples apply to the entire data set. When you include or exclude rows using control chart Options > Estimate and choose to perform a Box-Cox transformation. Minitab draws a horizontal line labeled LB at the lower bound instead. Minitab draws a horizontal line labeled UB at the upper bound instead. If the calculated lower confidence limit is less than the lower bound. > control chart Options > Confidence Limits Charts T2. Upper confidence limit bound: Check to set the upper confidence limit bound. 3 12:15): Choose to specify subgroups of data to compute Σ.Quality Control Generalized Variance Chart – Options – Estimate Stat > Control Charts > Multivariate Charts > Generalized Variance > Gen Var Options > Estimate Omits or includes certain subgroups to estimate Σ. Click the Confidence Limits tab. Do one of the following: • Choose Omit the following samples when estimating parameters and enter the subgroups or observation numbers you want to omit from the calculations. Caution When you include or exclude rows using control chart > Data Options > Subset and also omit or use samples to estimate parameters. To omit or use subgroups for the estimates of parameters By default. To specify confidence limits 1 2 In control chart dialog box. click control chart Options. change that value to missing (∗) in the worksheet. To omit one observation from a subgroup. and I-MR charts. Note 3 Click OK. 3 12:15): Choose to specify subgroups of the data to omit from the computations for Σ.. Minitab assumes the values you enter are subgroup numbers. you must enter subgroup 2 as the subgroup to omit. Enter entire subgroups. If the calculated upper confidence limit is greater than the upper bound. For example. • Choose Use the following samples when estimating parameters and enter the subgroups or observation numbers you want to use for the calculations. not just the subsetted data. Click the Estimate tab. Dialog box items Display confidence limits at These percents: Enter one or more values or enter a column of values. you exclude subgroup 1 from your analysis (using control chart > Data Options > Subset). Do any of the following: 196 2003 Minitab Inc. But you may want to use or omit certain data if they show abnormal behavior. Generalized Variance Specifies the confidence limit positions. Options − Confidence Limits . click control chart Options. Use the following samples when estimating parameters (eg. not individual observations. then type the upper limit (must be greater than 0) for the T2 chart. you can prevent these subgroups from influencing the estimation of process parameters. if some subgroups have erratic data due to assignable causes that you have corrected. With these charts. except with the Individuals.

The column must be the same length as the data column (or columns.. you should not choose a λ outside the range of −2 and 2. you must set up a column of grouping indicators. You may want to first run the command Stat > Control Charts > Box-Cox Transformation to help you find the optimal transformation value. If you like. dates. > control chart Options > Stages Charts All charts except Z-MR You can display stages in your process by drawing a "historical chart" − a control chart in which the control limits and center line are estimated independently for different groups in your data. the data must be positive. you can tell Minitab to start a new stage in one of two ways: • • 1 2 3 4 Each time the value in the column changes At the first occurrence of one or more values In the control chart dialog box. In Define stages (historical groups) with this variable. do one of the following: • To start a new stage each time the value in the column changes. or text.5). You can enter the same value more than once. With the first occurrence of these values: Choose to start a new stage at the first occurrence of a certain value. Minitab only uses the non-omitted data to find lambda. check Lower confidence limit bound. then type the lower confidence limit (greater than 0). 197 . Under Place bounds on confidence limits.. S.Control Charts • • • 3 Under Display confidence limits at. choose With the first occurrence of these values. check Upper confidence limit bound. Minitab treats each repeated value as a separate occurrence. Enter proportion values (between 0 and 1) or percent values (between 1 and 100). Historical charts are particularly useful for comparing data before and after a process improvement. Options − Stages . Enter the values. Under When to start a new value. then enter the values.. Click OK. when subgroups are across rows)." The transformation takes the original data to the power λ.. Note With the following charts. unless λ = 0. and X-bar and S. > control chart Options > Box-Cox Charts All except the Attributes charts You can use the Box-Cox power transformation when your data are very skewed or when the within-subgroup variation is unstable to make the data "more normal. in which case the natural log is taken. 2003 Minitab Inc. The Options subdialog box lists the common transformations natural log (λ = 0) and square root (λ = 0. In most cases. • To start a new stage at the first occurrence of a certain value. X-bar and R. then type the upper confidence limit (greater than 0). When executing the command.) To use this option. 5 Options − Box-Cox . choose With each new value. Minitab treats each repeated value as a separate occurrence. use any dialog box options. Click the Stages tab. To display a historical chart To define stages in your process. enter one or more values or enter a column of values in These percents. The indicators can be numbers. (λ is lambda. you must have at least one subgroup with two or more observations: R. then click OK. Enclose date/time or text entries in double quotes. Enclose date/time or text entries in double quotes. Dialog box items Define stages (historical groups) with this variable: Enter the column that contains the stage indicators. You can also choose any value between − 5 and 5 for λ. When you include or exclude rows using control chart Options > Estimate. When to start a new stage With each new value: Choose to start a new stage each time the value in the column changes. You can enter the same value more than once. enter the column which contains the stage indicators. click control chart Options. Under Place bounds on confidence limits.

Do one of the following: • To display all subgroups in the control chart. doing so will double-transform the data. and enter the number of subgroups. Click the Box-Cox tab. > control chart Options > Display Charts All charts except Z-MR Use to display the control chart by stage or by number of plotted points and to display the test results in the Session window. Optimal lambda for each stage (in a chart with stages): Choose to have Minitab search for optimal values for each stage. Caution If you use Stat > Control Charts > Box-Cox Transformation to find the optimal λ value and choose to store the transformed data with that command. Minitab does not accept any values you enter in the Parameters tab for historical means or historical standard deviations. and enter the number of subgroups for each segment.Quality Control When you use this transformation. click control chart Options. Click the Display tab.5 (square root): Choose to use the square root of the data. Display last __ subgroups: Choose to display a specific number of subgroups. Each segment contains __ subgroups: Choose to split the control chart into subgroups of a specified number of points. click control chart Options. see the independent Box-Cox transformation command Stat > Control Charts > Box-Cox Transformation. . 198 2003 Minitab Inc. • Choose Lambda=0. Lambda = 0 (natural log): Choose to use the natural log of the data. Use overall standard deviation: Check to use the overall standard deviation in the estimation of the optimal lambda. do not select the Box-Cox option when you make a control char. • Choose Optimal lambda for each stage (in a chart with stages) to have Minitab search for optimal values for each stage. Lambda=0. • Choose Optimal lambda to have Minitab search for an optimal value for all stages. under Subgroups to display. 3 Click OK. then enter one or more values. then do one of the following: • Choose Lambda = 0 (natural log) to use the natural log of the data. Options − Display . Dialog box items Box-Cox power transformation (W = Y**Lambda): Choose to use when your data are very skewed or when the withinsubgroup variation is unstable. then enter the lambda values. Each segment corresponds to a stage (if chart has stages): Choose to split the control chart by stage or stage when you have specified stages in control chart Options > Stage.5 (square root) to use the square root of the data. Test results Display test results in Session window: Check to display the test results in the Session window. Optimal lambda: Choose to have Minitab search for an optimal value. Split chart into series of segments for display purposes Do not split: Choose not to split the chart. For help choosing a lambda value. Other (enter value(s) between -5 and 5): Choose to transform the data using another lambda value or values. Check Box-Cox power transformation (W = Y**Lambda).. Subgroups to display Display all subgroups: Choose to display all subgroups.. • Choose Other (enter value(s) between -5 and 5) to transform the data using another lambda value or values. To do the Box-Cox power transformation 1 2 In the control chart dialog box. To display subgroups 1 2 In the control chart dialog box. choose Display all subgroups.

Choose Stat > Control Charts > Multivariate Charts > Generalized Variance. Minitab stores one column for the lower control limit and one column for the upper control limit. Store these values for each point Point plotted: Check to store the computed generalized variance values. To split a control chart 1 2 In the control chart dialog box. 199 . 2003 Minitab Inc. Click OK. blood pressure. Click Labels. choose Each segment corresponds to a stage (if chart has stages).MTW. Click the Display tab. You measure these variables every week for three months. the variables must remain in control. Check the statistics to store in the worksheet. In Subgroup sizes. Subgroup size: Check to store the subgroup sizes. enter Generalized Variance Chart for Effects of Heart Drug. you want to create a generalized variance chart to see if the joint variability is constant. enter HeartRate Systolic Diastolic Weight. Do one of the following: • To split the chart into a specific number of subgroups. choose Display last __ subgroups and enter the number of subgroups to display. Therefore. one row for each plotted point. Example of generalized variance chart You are interested in the effects of a new heart drug on the heart rate. Dialog box items Store these estimates for each chart Covariance matrices: Check to store the estimated covariance matrices. Under Title. For FDA approval of the drug. under Subgroups to display. 3 Generalized Variance Chart − Options − Storage Stat > Control Charts > Multivariate Charts > Generalized Variance >Gen Var Options > Storage Stores your choice of statistics in the worksheet. choose Each segment contains __ subgroups. one row for each plotted point. then click OK. and enter the number of subgroups for each segment. In Variables. • To split the chart by stage. Control limit values: Check to store the control limit values. and weight of five elderly female patients. enter Week. To store statistics in the worksheet 1 2 Choose Stat > Control Charts > control chart type > control chart > control chart Options > Storage.Control Charts • 3 To display a specific number of the last subgroups. Center line value: Check to store the center line value. one row for each stage. Click OK. Stage: Check to store the stage. click control chart Options. one row for each plotted point. Click OK in each dialog box. 1 2 3 4 5 6 Open the worksheet HEARTDRUG. one row for each plotted point.

the use of separate control charts is misleading because the variables jointly affect the process. Multivariate EWMA Chart Multivariate Control Charts Overview data Display statistics from two or more related measurement variables. To determine whether you should use a univariate or multivariate chart. see Control Charts Overview. However. multivariate charts are more difficult to interpret than classic Shewhart control charts. For more information about control charts. and out-of-control signals do not reveal which variable (or combination of variables) caused the signal. For example. Type I error and the probability of a point correctly plotting in control are not equal to their expected values. If the correlation coefficients are greater than 0. you can assume the variables are correlated. Multivariate control charting has several advantages over creating multiple univariate charts: • • • The actual control region of the related variables is represented (elliptical for bivariate case). A single control limit determines whether the process is in control. the scale on multivariate charts is unrelated to the scale of any of the variables. The distortion of these values increases with the number of measurement variables. A multivariate chart shows how several variables jointly influence a process or outcome.Quality Control Graph window output Interpreting the results No out-of-control points appear. blood pressures. You can maintain a specific Type I error. create a correlation matrix of your variables.1. you can use multivariate control charts to investigate how the tensile strength and diameter of a fiber affect the quality of fabric. indicating that the joint variability of the five patients' heart rates. and it is appropriate to create a multivariate control chart. If the data include correlated variables. . For example. and weights is constant. If you use separate univariate control charts in a multivariate situation. Choosing a multivariate control chart Minitab offers four multivariate control charts: • • • • T2 and Generalized Variance − a T2 chart and Generalized Variance chart displayed in one window T2 − a multivariate chart of means Generalized Variance − a multivariate chart of process variability Multivariate EWMA − a multivariate chart of exponentially weighted moving averages 200 2003 Minitab Inc.

with one column for each variable. ARL: Enter the average run length. in a spray drying process. You can use a MEWMA chart to simultaneously monitor two or more related process characteristics in an exponentially weighted control chart. By changing the weight used and the number of standard deviations for the control limits. such as the T2 control chart. you can construct a chart with specific properties. the second 5 observations are the second row. See [17] for an extensive table. to control your product particle size. The minimum allowable ARL is 1.Control Charts Multivariate EWMA Chart Stat > Control Charts > Multivariate Charts > Multivariate EWMA The multivariate form of the EWMA control chart. In the MEWMA chart. with subgroups of size 5. and so on. You can specify sample means. If the subgroup sizes do not vary much. Subgroups are equal size Univariate (one variable) Multivariate (more than one variable) Down columns or across rows Down columns Subgroups are unequal size Down columns with subgroup indicator column Down columns with subgroup indicator column Structure subgroup data down a column or across rows. If the subgroups are not equal. Multivariate data must be entered down columns. variances. and covariances for each variable or have Minitab calculate them. you can monitor feed temperature and chamber temperature. structured both ways. Note that the first five observations in the left data set (subgroup 1) are the first row of the right-side data set. For example. see Control Charts Overview and Multivariate Control Charts Overview. Variables charts include: • • • • Variables charts for subgroups Variables charts for individuals Time-weighted charts Multivariate charts Structure your data for these charts using the guidelines below. allowing you to detect small process shifts quicker than with other multivariate methods. each control limit is not a single straight line but varies with the subgroup size. Dialog box items Variables: Enter at least two columns of data. Weight: Specify the weight to be used in the multivariate exponentially weighted moving average. Here is the same data set. Subgroup sizes: Enter a number or a column of subscripts. For more information. Worksheet Structure Structure your data down a column or across rows. You can choose combinations of these two parameters by using an ARL (Average Run Length) table. <Scale> <Labels> <Data Options> <MEWMA Options> Data − Variables Control Chart Organize the data for all variables control charts in the same way. you may want to force the control limits to be constant by specifying a fixed subgroup size using MEWMA Options > Estimate. each plotted point includes weighting from all previous data. 201 . using the following table as a guide. The value must be a number between 0 and 1. 2003 Minitab Inc.

To display a multivariate EWMA chart 1 2 3 4 Choose Stat > Control Charts > Multivariate Charts > Multivariate EWMA. you must enter data that approximate a normal distribution.. Then. In Variables. enter a number or a column of subscripts. you may want to use the Box-Cox transformation to induce normality. Options − Estimate . . Missing data See Missing data in control charts for information on how to handle missing data for different types of control charts. You can access the Box-Cox transformation two ways: by using the Box-Cox transformation option provided with the control chart commands. you must enter your data in one column. and so on. If you like. MEWMA 202 2003 Minitab Inc. In the following example. a new subgroup begins in C1. T2. enter at least two columns containing variables to control. subgroup 2 has six observations.. subgroup 1 has three observations. Nonnormal data To properly interpret Minitab's control charts. see Box-Cox Transformation. then create a second column of subscripts which serve as subgroup indicators. or by using the stand alone Box-Cox command. use any dialog box options.Quality Control When subgroups are of unequal size. Use the stand alone command as an exploratory tool to help you determine the best lambda value for the transformation. see Options − Box-Cox. In this example. > control chart Options > Estimates Charts T2 and Generalized Variance. then click OK. For information on the stand alone Box-Cox transformation command. C1 contains the process data and C2 contains subgroup indicators: Each time a subscript changes in C2. you can use the transformation option to transform the data at the same time you draw the control chart. In Subgroup sizes. For information on the Box-Cox transformation option. If the data are highly skewed.

Multivariate EWMA Chart − Options – Parameters Stat > Control Charts > Multivariate charts > Multivariate EWMA > MEWMA Options > Parameters Enters historical data for estimating µ and Σ. Enter entire subgroups. except with the Individuals. otherwise. if some subgroups have erratic data due to assignable causes that you have corrected. you can prevent these subgroups from influencing the estimation of process parameters. Click MEWMA Options. For example. all stages use that covariance matrix. A list of values. enter a value for each stage. Because Minitab considers an omitted sample (subgroup 1) as part of the entire data set. With these charts. The elements on the matrix diagonal must be positive. not just the subsetted data. Minitab estimates it from the data. if you have a known process parameter or an estimate obtained from past data. To use a historical mean and covariance matrix 1 2 Choose Stat > Control Charts > Multivariate Charts > Multivariate EWMA. you exclude subgroup 1 from your analysis (using control chart > Data Options > Subset). Do any of the following: • In Mean.j) must equal element (j. You can omit certain subgroups when estimating parameters using MEWMA Options > Estimate. Minitab assumes the values you enter are subgroup numbers. Element (i. Moving Range. click control chart Options. Click the Estimate tab. all stages use that value. you must enter subgroup 2 as the subgroup to omit. you can enter this value or values. enter one or more values. enter a matrix for each historical group. Each stage for each variable uses a different value. A list of values. If you enter only one matrix. which applies to all variables and stages. then type individual subgroup numbers or a range of subgroups. change that value to missing (∗) in the worksheet. Minitab only uses the non-omitted data to find lambda. But you may want to use or omit certain data if they show abnormal behavior. and I-MR charts. otherwise. To omit or use subgroups for the estimates of parameters By default.Control Charts Omits or includes certain subgroups to estimate µ and Σ. 203 . When you include or exclude rows using control chart options > Estimate and choose to perform a Box-Cox transformation. Click OK. • Choose Use the following samples when estimating parameters and enter the subgroups or observation numbers you want to use for the calculations. not individual observations. Minitab estimates the process parameters from all the data. the values are interpreted as observation numbers. one for each variable. 3 • • • You can enter the historical mean values in several ways: 2003 Minitab Inc. Dialog box items Omit the following samples when estimating parameters: Choose to specify subgroups of the data to omit from the computations for µ and Σ. one for each stage within each variable. then decide to omit subgroup 2 (using control chart Options > Estimate). then type individual subgroup numbers or ranges of subgroups. For example. A single value. All stages for each variable use the same value. Covariance matrix: Enter the historical matrices. Note 3 Click OK. Each matrix must be a positive definite square matrix with the same number of rows and columns as columns of data. Note When you include or exclude rows using control chart > Data Options > Subset and also omit or use samples to estimate parameters. To omit one observation from a subgroup. Use the following samples when estimating parameters: Choose to specify subgroups of data to compute µ and Σ. 1 2 In the control chart dialog box. the omitted or used samples apply to the entire data set. Do one of the following: • Choose Omit the following samples when estimating parameters and enter the subgroups or observation numbers you want to omit from the calculations. Dialog box item Mean: Enter the historical µ's as a single value or a list of values. enter one or more matrices.i). For example. • In Covariance matrix. If you enter one constant. If you do not specify a value for the mean or the covariance matrix.

then enter the values.Quality Control Multivariate EWMA − Options − Confidence Limits Stat > Control Charts > Multivariate charts > Multivariate EWMA > MEWMA Options > Confidence Limits Specifies the confidence limit positions. choose With each new value. You can enter the same value more than once.. The column must be the same length as the data column (or columns. then click the Confidence Limits tab. use any dialog box options. When to start a new stage With each new value: Choose to start a new stage each time the value in the column changes. Click OK. or text. and X-bar and S. Enclose date/time or text entries in double quotes.. do one of the following: • To start a new stage each time the value in the column changes. Historical charts are particularly useful for comparing data before and after a process improvement. Click MEWMA Options. Minitab treats each repeated value as a separate occurrence. . Note With the following charts. dates. choose With the first occurrence of these values. Under Place bounds on confidence limits. Options − Stages . Minitab treats each repeated value as a separate occurrence. With the first occurrence of these values: Choose to start a new stage at the first occurrence of a certain value. you can tell Minitab to start a new stage in one of two ways: • • 1 2 3 4 Each time the value in the column changes At the first occurrence of one or more values In the control chart dialog box. click control chart Options. 5 204 2003 Minitab Inc. Dialog box items Place bounds on confidence limits Upper confidence limit bound: Check to set the upper confidence limit bound. Under When to start a new value. When executing the command. S. Enclose date/time or text entries in double quotes. then type the upper limit (must be greater than 0). If you like. Minitab draws a horizontal line labeled UB at the upper bound instead. Click the Stages tab. Enter the values. The indicators can be numbers. check Upper confidence limit bound and type the upper limit. > control chart Options > Stages Charts All charts except Z-MR You can display stages in your process by drawing a "historical chart" − a control chart in which the control limits and center line are estimated independently for different groups in your data. • To start a new stage at the first occurrence of a certain value. To specify the upper confidence limit for a Multivariate EWMA chart 1 2 3 Choose Stat > Control Charts > Multivariate Charts > Multivariate EWMA. You can enter the same value more than once. you must set up a column of grouping indicators. X-bar and R. when subgroups are across rows). If the calculated upper confidence limit is greater than the upper bound. Dialog box items Define stages (historical groups) with this variable: Enter the column that contains the stage indicators. In Define stages (historical groups) with this variable. then click OK. you must have at least one subgroup with two or more observations: R. enter the column which contains the stage indicators. To display a historical chart To define stages in your process.

the data must be positive. (λ is lambda. then enter the lambda values. > control chart Options > Display Charts All charts except Z-MR Use to display the control chart by stage or by number of plotted points and to display the test results in the Session window..5). When you include or exclude rows using control chart Options > Estimate. you should not choose a λ outside the range of −2 and 2.. Minitab does not accept any values you enter in the Parameters tab for historical means or historical standard deviations. Optimal lambda: Choose to have Minitab search for an optimal value. > control chart Options > Box-Cox Charts All except the Attributes charts You can use the Box-Cox power transformation when your data are very skewed or when the within-subgroup variation is unstable to make the data "more normal. then do one of the following: • Choose Lambda = 0 (natural log) to use the natural log of the data. see the independent Box-Cox transformation command Stat > Control Charts > Box-Cox Transformation. in which case the natural log is taken. Minitab only uses the non-omitted data to find lambda. Options − Display . The Options subdialog box lists the common transformations natural log (λ = 0) and square root (λ = 0. and enter the number of subgroups.. 3 Click OK. Subgroups to display Display all subgroups: Choose to display all subgroups. doing so will double-transform the data.5 (square root): Choose to use the square root of the data.) To use this option. You may want to first run the command Stat > Control Charts > Box-Cox Transformation to help you find the optimal transformation value. Use overall standard deviation: Check to use the overall standard deviation in the estimation of the optimal lambda." The transformation takes the original data to the power λ. Check Box-Cox power transformation (W = Y**Lambda). Click the Box-Cox tab. • Choose Optimal lambda for each stage (in a chart with stages) to have Minitab search for optimal values for each stage. You can also choose any value between − 5 and 5 for λ. Other (enter value(s) between -5 and 5): Choose to transform the data using another lambda value or values. • Choose Other (enter value(s) between -5 and 5) to transform the data using another lambda value or values. In most cases. then enter one or more values. When you use this transformation. 205 . unless λ = 0. • Choose Lambda=0.5 (square root) to use the square root of the data.Control Charts Options − Box-Cox .. Lambda = 0 (natural log): Choose to use the natural log of the data. do not select the Box-Cox option when you make a control char. Dialog box items Box-Cox power transformation (W = Y**Lambda): Choose to use when your data are very skewed or when the withinsubgroup variation is unstable. Caution If you use Stat > Control Charts > Box-Cox Transformation to find the optimal λ value and choose to store the transformed data with that command. Split chart into series of segments for display purposes 2003 Minitab Inc. For help choosing a lambda value. • Choose Optimal lambda to have Minitab search for an optimal value for all stages. click control chart Options. Lambda=0. To do the Box-Cox power transformation 1 2 In the control chart dialog box. Display last __ subgroups: Choose to display a specific number of subgroups. Optimal lambda for each stage (in a chart with stages): Choose to have Minitab search for optimal values for each stage.

Click the Display tab. . 3 Multivariate EWMA Chart − Options − Storage Stat > Control Charts > Multivariate Charts > Multivariate EWMA > MEWMA Chart Options > Storage Stores your choice of statistics in the worksheet. choose Display last __ subgroups and enter the number of subgroups to display. Each segment corresponds to a stage (if chart has stages): Choose to split the control chart by stage or stage when you have specified stages in control chart Options > Stage. Stage: Check to store the stage.Quality Control Do not split: Choose not to split the chart. Do one of the following: • To split the chart into a specific number of subgroups. Test results: Check to store the test results. choose Each segment contains __ subgroups. Subgroup size: Check to store the subgroup sizes. • To display a specific number of the last subgroups. Store these values for each point: Point plotted: Check to store the computed MEWMA values. you want to monitor the weight (in grams) and length (in centimeters) of one of your toy parts. one row for each plotted point. Example of multivariate EWMA chart As production manager of a toy manufacturing company. Minitab assigns each plotted point either a 1 (beyond upper control limit) or a 0 (in control). In Subgroup sizes. Click OK. then click OK. and enter the number of subgroups for each segment. one row for each plotted point. one column for each column of data and one row for each stage. To display subgroups 1 2 In the control chart dialog box. Check the statistics to store in the worksheet. 206 2003 Minitab Inc. Dialog box items Store these estimates for each chart Means: Check to store the mean estimates. choose Display all subgroups. You collect 4 samples each day for 20 days. Choose Stat > Control Charts > Multivariate Charts > Multivariate EWMA. click control chart Options. Test results Display test results in Session window: Check to display the test results in the Session window. enter Weight Length. Control limit values: Check to store the upper control limit values. one row for each plotted point. Because weight and length are correlated and you want to detect small shifts in these variables. • To split the chart by stage.MTW. choose Each segment corresponds to a stage (if chart has stages). To store statistics in the worksheet 1 2 Choose Stat > Control Charts > control chart type > control chart > control chart Options > Storage. Each segment contains __ subgroups: Choose to split the control chart into subgroups of a specified number of points. under Subgroups to display. Click OK. Click the Display tab. Covariance matrices: Check to store the estimated covariance matrices. Do one of the following: • To display all subgroups in the control chart. click control chart Options. 3 To split a control chart 1 2 In the control chart dialog box. one row for each plotted point. enter Day. and enter the number of subgroups for each segment. under Subgroups to display. 1 2 3 4 Open the worksheet TOYS. one row for each stage. In Variables. Click OK. you create a multivariate EWMA chart.

2003 Minitab Inc. 207 . indicating that differences in weight and length over time are due to common causes.Control Charts Graph window output Interpreting the results All points fall below the upper control limit.

.

Precision describes the variation you see when you measure the same part repeatedly with the same device. on the same part. Thus. Gage R&R (Nested). Attribute Agreement Analysis was previously called Attribute Gage R&R Study in Minitab Release 13. but not precise. 209 . and reducing that variation as much as possible to get a more consistent product. Note Any time you measure the results of a process you will see variation. To examine your measurement system's accuracy. Gage R&R (Crossed). • • • Gage Run Chart. you may want to check that the variation you observe is not mostly due to errors in your measurement system. It is the difference in the observed accuracy values through the expected range of measurements. Gage Linearity and Bias Study examines gage linearity and accuracy. It is the difference between the observed average measurement and a master value. Stability − a measure of how accurately the system performs over time. a method to examine the bias and repeatability of an attribute measurement system. and Attribute Gage Study (Analytic Method) examine measurement system precision. • • Accuracy describes the difference between the measurement and the part's actual value. You can also have a device that is accurate (the average of the measurements is very close to the accurate value). Accuracy The accuracy of a measurement system is usually broken into three components: • • • Linearity − a measure of how the size of the part affects the accuracy of the measurement system. Reproducibility − the variation due to the measurement system. It is the variation observed when the same operator measures the same part repeatedly with the same device. Within any measurement system. But before you conduct any SPC analyses. Attribute Agreement Analysis. Precision Precision. that is. you can have a device that measures parts precisely (little variation in the measurements) but not accurately. Bias − a measure of the bias in the measurement system. you can have one or both of these problems. Attribute Agreement Analysis helps assess the agreement of subjective ratings or classifications given by multiple appraisers for both nominal and ordinal data. is different from Attribute Gage Study (Analytic Method). can be broken down into two components: • • Repeatability − the variation due to the measuring device. It is the total variation obtained with a particular device. measuring the same part repeatedly does not result in identical measurements. or measurement variation. This variation comes from two sources: differences between parts made by the process and imperfections in taking measurements. 2003 Minitab Inc. see Gage Linearity and Bias Study. It is the variation observed when different operators measure the same parts using the same device. For example.Measurement Systems Analysis Overview Measurement Systems Analysis Overview Minitab offers several commands to help you determine how much of your process variation arises from variation in your measurement system. the measurements have large variance. You can also have a device that is neither accurate nor precise. a technique to assess appraisers' agreement. Statistical Process Control (SPC) is concerned with identifying sources of part-to-part variation. Measurement system error Measurement system errors can be classified into two categories: accuracy and precision. when measuring a single characteristic over time.

If you want to distinguish a higher number of distinct categories. This means that some of those 10 parts are not different enough to be discerned as being different by your measurement system. see Attribute Gage Study (Analytic Method). Measurement Systems Analysis Stat > Quality Tools > Gage Study Gage Run Chart Gage Linearity and Bias Study Gage R&R Study (Nested) Gage R&R Study (Crossed) Attribute Gage Study (Analytic Method) Stat > Quality Tools Attribute Agreement Analysis Examples of Measurement Systems Analysis Minitab Help offers examples of the following measurement systems analyses: Gage Run Chart Gage Linearity and Bias Study Gage R&R Study (Crossed) − ANOVA Method Gage R&R Study (Crossed) − X-bar and R method Gage R&R Study (Nested) Attribute Gage Study (Analytic Method) Attribute Agreement Analysis Number of Distinct Categories Statement Minitab calculates the number in this statement by dividing the standard deviation for Parts by the standard deviation for Gage. see Gage R&R Study. A value of 5 or more denotes an acceptable measurement system. A. Ford. Chrysler. say high and low." Quality Progress. Measurement Systems Analysis Reference Manual. J. middle and high. 3rd edition. 2nd edition. N. When the number of categories is 2. 210 2003 Minitab Inc. Imagine you measured 10 different parts. General Motors Supplier Quality Requirements Task Force. You can also think of it as the number of groups within your process data that your measurement system can discern. This number represents the number of non-overlapping confidence intervals that will span the range of product variation.L. Quality Control and Industrial Statistics. then multiplying by 1. Irwin. the data can be divided into two groups. Kotz (1990). "When Quality is a Matter of Taste. the data can be divided into 3 groups. Futrell (1995). Desu and S. you need a more precise gage. Use Reliability Indexes. . To look at a plot of all of the measurements by operator/part combination. Duncan (1986).Quality Control Choosing a command To look at your measurement system's precision. the measurement system is of no value for controlling the process. and thus visualize the repeatability and reproducibility components of the measurement variation. Statistical Methods for Rates and Proportions. John Wiley & Sons. To look at both the bias and repeatability in an attribute measurement system. since one part cannot be distinguished from another. Fleiss (1981). When the number of categories is 3. D. say low. and Minitab reported that your measurement system could discern 4 distinct categories. References − Measurement Systems [1] [2] [3] [4] [5] Automotive Industry Action Group (AIAG) (2002). The Automobile Industry Action Group (AIAG) [1] suggests that when the number of categories is less than 2. 28 (5). Sample Size Methodology. 5th edition. Academic Press.L. see Gage Run Chart. 81-86.41 and truncating this value.

6 (2). A stable process would give you a random horizontal scattering of points. the reference line is the mean of all observations (optional). [10] S. Measure 1. Measurement data: Enter the column containing the observed measurements. <Gage Info> <Options> Data − Gage Run Chart Worksheet Structure Structure your data so each row contains the part name or number.33 . PartNum 1 1 2 2 3 3 1 1 . or a value you enter from prior knowledge of the process. which can be calculated from the data. Part I: Basic Methods. If you like. Runger (1993-4). 2nd edition. Johnson and S. operator (optional).C.38 1. Parts and operators can be text or numbers. Historical mu: Enter your own value for the horizontal reference line. 2nd edition. Operators: Enter the column containing operator names or numbers." Quality Engineering. Dialog box items Part numbers: Enter the column containing part names or numbers..R." Quality Engineering.43 1. In Operators.J. Gage Run Chart Gage Run Chart Stat > Quality Tools > Gage Study > Gage Run Chart A gage run chart is a plot of all of your observations by operator and part number. [11] S.78 1. an operator or part effect would give you some kind of pattern in the plot. enter the column of part names or numbers. To make a gage run chart 1 2 3 4 5 Choose Stat > Quality Tools > Gage Study > Gage Run Chart. John Wiley & Sons.C. Montgomery and G. By default. Rubin (2002). enter the column of measurements. and the observed measurement. Siegel and N.. "Gauge Capability Analysis and Designed Experiments. D.J. McGraw-Hill.83 1.. D. "Gauge Capability and Designed Experiments.. Discrete Distributions. Variance Components. Operator Daryl Daryl Daryl Daryl Daryl Daryl Beth Beth .53 1. enter the column of operator names or numbers. Little and D. McCulloch (1992). John Wiley & Sons. John Wiley & Sons. Montgomery and G.. Part II: Experimental Design Models and Variance Component Estimation. use one or more of the available dialog box options. R. 6 (1). Statistical Analysis With Missing Data. In Measurement data. Runger (1993-4). 289−305. Casella. 211 . A horizontal reference line is drawn at the mean. and C. In Part numbers.Measurement Systems Analysis [6] [7] [8] [9] N.C. then click OK. Castellan (1988)..E. B. You can use this chart to quickly assess differences in measurements between different operators and parts.C.83 1. Searle. G.48 1. 115−135.A. Nonparametric Statistics for the Behavioral Sciences. Trial numbers: Enter the column containing trial numbers (optional). Kotz (1969). 2003 Minitab Inc.

In Operators. Step 2: Gage Run Chart with GAGE2 data 1 2 3 3 4 Open the file GAGE2. In Part numbers. you draw a gage run chart with two data sets: one in which measurement system variation contributes little to the overall observed variation (GAGEAIAG. For the GAGEAIAG data. Dialog box items Gage name: Type the name of the gage. Gage R&R Study (Nested). three parts were selected that represent the expected range of the process variation. enter Part.MTW. three times per part.MTW. For comparison. enter Response. Choose Stat > Quality Tools > Gage Study > Gage Run Chart. enter Part. Gage Run Chart − Options Stat > Quality Tools > Gage Study > Gage Run Chart > Options Lets you enter a new title to replace the default title. Dialog box items Title: Enter a new title to replace the default title on the graphical output. or Attribute Gage Study (Analytic Method) > Gage Info Use these fields to fill in the label on the Session window and graphical output. Gage Tolerance: Type the gage tolerance. in a random order. In Measurement data. enter Operator. Each field corresponds to a line on the label. Step 1: Gage Run Chart with GAGEAIAG data 1 2 3 4 5 Open the worksheet GAGEAIAG. In Part numbers. ten parts were selected that represent the expected range of the process variation. in a random order. Gage R&R Study (Crossed). Reported by: Type the name of the person reporting the study information.MTW). Click OK. Miscellaneous: Type any miscellaneous notes. Three operators measured the ten parts. and one in which measurement system variation contributes a lot to the overall observed variation (GAGE2. Choose Stat > Quality Tools > Gage Study > Gage Run Chart. For the GAGE2 data. In Measurement data. Date of study: Type the date. If you do not use these fields. see the same data sets analyzed by the Gage R&R Study − ANOVA Method or the Gage R&R Study − Xbar and R method. Click OK. Gage Linearity and Bias Study. two times per part.Quality Control Gage Info Stat > Quality Tools > Gage Study > choose Gage Run Chart. In Operators. 212 2003 Minitab Inc. enter Response. . enter Operator.MTW). Three operators measured the three parts. Two Examples of Gage Run Chart In this example. the label will be printed with blank lines.

2003 Minitab Inc. Some smaller patterns also appear. and differences in measurements between operators. By default the reference line is the mean of all observations. Operator 2's second measurement is consistently (seven times out of ten) smaller than the first measurement. For example. Most of the variation is due to differences between parts.Measurement Systems Analysis Graph window output Gage Run Chart with GAGEAIAG data Gage Run Chart with GAGE2 data Interpreting the results Gage Run Chart Example − GAGEAIAG. 213 . Operator 2's measurements are consistently (eight times out of ten) smaller than Operator 1's measurements. you can compare both the variation between measurements made by each operator. You can also look at the measurements in relationship to the horizontal reference line.MTW For each part.

MTW For each part..1 3. Reference Response 2 2 2 2 2 2 2 2 2 2 2 2 4 4 .3 2. "Does my gage have the same accuracy for all sizes of objects being measured?" Gage bias examines the difference between the observed average measurement and a reference or master value. reference value. You can get the process standard deviation from the Total Variation row of the 6 ∗ SD column of the output from the Gage R&R Study − ANOVA Method.6 2. Reference values: Enter the column containing the reference values. It answers the question.. You can also look at the measurements in relationship to the horizontal reference line. The dominant factor here is repeatability − large differences in measurements when the same operator measures the same part. Oscillations might suggest the operators are "adjusting" how they measure between measurements.5 2. you can enter the process tolerance instead. 2. you can compare both the variation between measurements made by each operator..Quality Control Gage Run Chart Example − GAGE2. If you do not know the value for the process variation.. By default the reference line is the mean of all observations. Parts can be text or numbers. 214 2003 Minitab Inc.9 ..7 2. To do a gage linearity and bias study 1 Choose Stat > Quality Tools > Gage Study > Gage Linearity and Bias Study. and differences in measurements between operators.7 2. PartNum 1 1 1 1 1 1 1 1 1 1 1 1 2 2 . "How biased is my gage when compared to a master value?" Dialog box items Part numbers: Enter the column containing the part name or number. Measurement data: Enter the column containing the observed measurements. It answers the question.5 2.4 2. <Gage Info> <Options> Data − Gage Linearity and Bias Study Worksheet Structure Structure your data so each row contains a part.5 2. and the observed measurement on that part (the response).4 2.5 2. Process Variation (optional): Enter the process standard deviation.4 2.4 5.. Gage Linearity and Bias Study Gage Linearity and Bias Study Stat > Quality Tools > Gage Study > Gage Linearity and Bias Study Gage linearity tells you how accurate your measurements are through the expected range of the measurements. . This is the number that is usually associated with process variation.

type 14. enter Master. 1 2 3 4 5 6 Open the worksheet GAGELIN. If you do not use these fields. one operator randomly measured each part twelve times. In Measurement data. Date of study: Type the date. Each part was measured by layout inspection to determine its reference (master) value. Title: Enter a new title to replace the default title on the graphical output. Ford. Gage R&R Study (Nested). the label will be printed with blank lines. General Motors Supplier Quality Requirements Task Force). The data set used in this example has been reprinted with permission from the Measurement Systems Analysis Reference Manual (Chrysler. If you like.1941. This number is usually associated with process variation. Reported by: Type the name of the person reporting the study information. Gage R&R Study (Crossed). Choose Stat > Quality Tools > Gage Study > Gage Linearity and Bias Study. In Reference values. enter the process tolerance instead. Gage Tolerance: Type the gage tolerance. enter Response. You obtained the process variation (14. Dialog box items Methods of estimating repeatability standard deviation Sample range: Choose to use the sample range to estimate the repeatability standard deviation.Measurement Systems Analysis 2 3 4 5 In Part numbers. In Process Variation. Example of Gage Linearity and Bias Study A plant foreman chose five parts that represented the expected range of the measurements. In Process Variation. enter the column of observed measurements. 2003 Minitab Inc. In Measurement data. enter the column of part names or numbers. If you do not know the process variation value. then click OK. 215 . Each field corresponds to a line on the label. Gage Linearity and Bias − Options Stat > Quality Tools > Gage Study > Gage Linearity and Bias Study > Options Use to specify the method to estimate the repeatability standard deviation or to enter a new title to replace the default title. Dialog box items Gage name: Type the name of the gage. Sample standard deviation: Choose to use the sample standard deviation to estimate the repeatability standard deviation. Gage Linearity and Bias Study. In Master Measurements. enter the column of master measurements. enter Part. Then.MTW. Miscellaneous: Type any miscellaneous notes. Minitab displays the process variation in the Session window (Total Variation row of the Study Var (6 * SD) column). In Part numbers. use any dialog box options. enter a value. 6 Gage Info Stat > Quality Tools > Gage Study > choose Gage Run Chart.1941) from a Gage R&R study using the ANOVA method. Click OK. or Attribute Gage Study (Analytic Method) > Gage Info Use these fields to fill in the label on the Session window and graphical output. You can get this value from the Gage R&R Study − ANOVA method: it is the value in the Total Variation row of the 6 ∗ SD column.

which means that the gage bias accounts for less than 0. Gage R&R Study (Nested) uses the ANOVA method only. and ANOVA.4% of the overall process variation. whenever operators measure unique parts.Quality Control Graph window output Interpreting the results The %Linearity (absolute value of the slope * 100) is 13. If you are unable to make that assumption then part-to-part variation within a batch will mask the measurement system variation. Use Gage R&R Study (Nested) when each part is measured by only one operator. . because it considers the operator by part interaction [8. Minitab allows you to perform either crossed or nested Gage R&R studies. If you need to use destructive testing. then choosing between a crossed or nested Gage R&R Study for destructive testing depends on how your measurement process is set up. and reproducibility. Gage R&R Study (Crossed) allows you to choose between the X-bar and R method and the ANOVA method. Crash testing is an example of destructive testing. and operator-by-part. In fact. The X-bar and R method breaks down the overall variation into three categories: part-to-part. repeatability. you must be able to assume that all parts within a single batch are identical enough to claim that they are the same part. which means that the gage linearity accounts for 13% of the overall process variation. then use Gage R&R Study (Crossed). The ANOVA method is more accurate than the X-bar and R method.2. the measured characteristic is different after the measurement process than it was at the beginning. 216 2003 Minitab Inc. components. The %Bias for the reference average is 0. in part.4. you have a nested design. If all operators measure parts from each batch. 9]. Minitab provides two methods for assessing repeatability and reproducibility: X-bar and R. then you must use Gage R&R Study (Nested). In destructive testing. If each batch is only measured by a single operator. • • Use Gage R&R Study (Crossed) when each part is measured multiple times by each operator. such as in destructive testing. If you can make that assumption. Gage R&R Study (Crossed) Gage R&R Gage repeatability and reproducibility studies determine how much of your observed process variation is due to measurement system variation. The ANOVA method goes one step further and breaks down reproducibility into its operator.

Measurement Systems Analysis If you have attribute data.53 1. 2003 Minitab Inc. Method of Analysis ANOVA: Choose to use the ANOVA method. Measurement data: Enter the column containing the observed measurements. repeatability. you can use Attribute Gage Study (Analytic Method) to analyze the bias and repeatability of your process. components. 217 . and the observed measurement.33 . 9]. operator (optional). or ANOVA.38 1. The Xbar and R method breaks down the overall variation into three categories: part-to-part. Operators: Enter the column containing operator names or numbers (optional). The ANOVA method is more accurate than the X-bar and R method.. in part. Gage R&R Study (Crossed) Stat > Quality Tools > Gage Study > Gage R&R Study (Crossed) Gage repeatability and reproducibility studies determine how much of your observed process variation is due to measurement system variation. Operator Daryl Daryl Daryl Daryl Daryl Daryl Beth Beth . Dialog box items Part numbers: Enter the column containing the part names or numbers. and operator by part..83 1.78 1. Measure 1. Xbar and R: Choose to use the X-bar and R method. PartNum 1 1 2 2 3 3 1 1 ..43 1. because it accounts for the Operator by Part interaction [8.. <Gage Info> <Options> Data − Gage R&R Study (Crossed) Structure your data so that each row contains the part name or number. Parts and operators can be text or numbers..83 1. Minitab provides two methods for Gage R&R Crossed: X-bar and R. The ANOVA method goes one step further and breaks down reproducibility into its operator. and reproducibility..48 1.

If you do not use these fields. and to the mean. in which measurement system variation has a large effect on the overall observed variation GAGEAIAG. Gage R&R Study (Nested). Otherwise. allowing you to compare operators to each other. Gage R&R − Graph Window Output The Gage R&R study produces six graphs: • • Components of Variation is a visualization of the final table in the Session window output. as well as compare results from the various analyses. so you can compare the mean measurement for each part. Gage Linearity and Bias Study. choose ANOVA (the default) or Xbar and R. Gage Info Stat > Quality Tools > Gage Study > choose Gage Run Chart. you may want to estimate the missing observations. General Motors Supplier Quality Requirements Task Force). If you have many replicates. The table below lists the other measurement systems analysis examples. or Attribute Gage Study (Analytic Method) > Gage Info Use these fields to fill in the label on the Session window and graphical output. Operator by Part Interaction displays the Operator by Part effect. Repeatability. the label will be printed with blank lines. Data Sets Used in Examples The same two data sets are used in the Gage R&R Study and the Gage Run Chart examples: • • GAGE2. a S chart is displayed. boxplots are displayed on the By Operator graph. use any dialog box options. so you can compare mean measurement for each operator. In Part numbers. Gage R&R Study (Crossed). and Part-to-Part variation. X-bar Chart by Operator displays the measurements in relation to the overall mean for each operator. Ford. • • • • Missing Data Gage R&R studies require balanced designs (equal numbers of observations per cell) and replicates to calculate the variance components. Once the missing observations have been estimated. By Operator displays the main effect for Operator. the data can be analyzed as a balanced design. R Chart by Operator displays the variation in the measurements made by each operator. If you like.MTW. In Measurement data. You can estimate any missing observations with the methods described in [7]. To do a gage R&R study (crossed) 1 2 3 4 5 Choose Stat > Quality Tools > Gage Study > Gage R&R Study (Crossed).Quality Control The Gage R&R studies require balanced designs (equal numbers of observations per cell) and replicates. See [7] for methods of estimating. If you have missing data.MTW. Each field corresponds to a line on the label. in which measurement system variation has a small effect on the overall observed variation You can compare the output for the two data sets. Under Method of Analysis. Reproducibility (but not Operator and Operator by Part).MTW are reprinted with permission from the Measurement Systems Analysis Reference Manual (Chrysler.MTW and GAGELIN.MTW AUTOGAGE.MTW GAGEAIAG. displaying bars for: Total Gage R&R. Example Gage Linearity Study Attribute Gage Study (Analytic Method) Attribute Agreement Analysis Data set GAGELIN. . If you have many replicates. then click OK.MTW ESSAY. boxplots are displayed on the By Part graph. allowing you to compare operators to each other. By Part displays the main effect for Part. 218 2003 Minitab Inc. The R chart is displayed when the number of replicates is less than 9. enter the column of part names or numbers. enter the column of measurements. so you can see how the relationship between Operator and Part changes depending on the operator.

Dialog box items Study variation: Enter a multiplier to use in the StudyVar column of the Session window output. we analyze the data using both the ANOVA method (below) and the Xbar and R method. Draw figures on separate graphs. in a random order. You can also look at the same data plotted on a Gage Run Chart. In Part numbers. enter Operator. enter Part. Click OK. Gage Tolerance: Type the gage tolerance. Do not display percent study variation: Check to suppress display of percent study variation. enter Part. Date of study: Type the date. Gage R&R Study (Crossed) − Options Stat > Quality Tools > Gage Study > Gage R&R Study (Crossed) > Options Allows you to control the Session and Graph window output. Step 2: Use the ANOVA method with GAGE2 data 1 2 3 4 5 Open the file GAGE2.MTW. For the GAGE2 data. This column shows the percentage of the process tolerance taken up by each variance component. A %Process column will be displayed in the bottom table of the Session window output and in the components of variation graph. which is the number of standard deviations needed to capture 99. This option is only available with the ANOVA method. In Operators. Reported by: Type the name of the person reporting the study information. enter Response.MTW. we do a gage R&R study on two data sets: one in which measurement system variation contributes little to the overall observed variation (GAGEAIAG. Process tolerance: Enter a known tolerance range. Three operators measured the three parts. Choose Stat > Quality Tools > Gage Study > Gage R&R Study (Crossed). Historical standard deviation: Enter a known historical standard deviation. three parts were selected that represent the expected range of the process variation. In Measurement data. The default multiplier is 6. This column shows the percentage of process standard deviation taken up by each variance component. three times per part. In Part numbers.73% of your process measurements. In Measurement data. Miscellaneous: Type any miscellaneous notes. two times per part. enter Operator. By default. ten parts were selected that represent the expected range of the process variation. 219 .MTW). one figure per graph: Check to display each graph separately. enter Response. Example of Gage R&R Study (ANOVA method) In this example. and one in which measurement system variation contributes a lot to the overall observed variation (GAGE2. Title: Enter a new title to replace the default title on the graphical output. For comparison. Choose Stat > Quality Tools > Gage Study > Gage R&R Study (Crossed). Three operators measured the ten parts. Alpha to remove interaction term: Enter the alpha value used to remove the interaction term from the model. 2003 Minitab Inc. choose ANOVA. Under Method of Analysis. For the GAGEAIAG data set. Do not display percent contribution: Check to suppress display of percent contribution.Measurement Systems Analysis Dialog box items Gage name: Type the name of the gage. In Operators. Gage R&R Study produces a layout of 6 graphs in one window. in a random order Step 1: Use the ANOVA method with GAGEAIAG data 1 2 3 4 5 6 7 Open the file GAGEAIAG.MTW). A %Tolerance column will be displayed in the bottom table of the Session window output and in the components of variation graph.

03940 0.37 89.00 Source Total Gage R&R Repeatability Reproducibility Operator Operator*Part Part-To-Part Total Variation StdDev (SD) 0.10 7.17 94.4 7437.7178 4.15668 1.33 100.0009120 0.0031458 0.05871 0.192781 0.21564 0.0044375 0.0022338 0.ANOVA Method Two-Way ANOVA Table With Interaction Source Part Operator Part * Operator Repeatability Total DF 2 2 4 18 26 SS 38990 529 26830 133873 200222 MS 19495. choose ANOVA. Click OK.39969 0.484 Two-Way ANOVA Table Without Interaction Source DF SS MS F P 220 2003 Minitab Inc.24913 MS 0.50 14. Session window output ANOVA method with GAGEAIAG data Gage R&R Study .024000 0.66 17.Quality Control 6 7 Under Method of Analysis.033 0.228745 0.005759 0.066615 0.2 264.4588 P 0.22379 %Study Var (%SV) 32.67 3.203965 Study Var (6 * SD) 0.166 0.90650 0.03875 2.030200 0.3 6707.1672 4.56 2.001292 F 39.035940 0. .52 100.000 0.81 23.19 5.00 Number of Distinct Categories = 4 Gage R&R for Response ANOVA method with GAGE2 data Gage R&R Study .33653 0.04800 0.056088 0.000 Gage R&R Source Total Gage R&R Repeatability Reproducibility Operator Operator*Part Part-To-Part Total Variation VarComp 0.10367 0.90185 P 0.962 0.047263 0.4 F 2.62 27.18120 0.0012917 0.0416019 %Contribution (of VarComp) 10.0371644 0.28358 1.ANOVA Method Two-Way ANOVA Table With Interaction Source Part Operator Part * Operator Repeatability Total DF 9 2 18 30 59 SS 2.

67 0.804 512.0000 36.85 0.0000 0.092 0.85 91.00 0.67 7304.000 220.36 84.000 0.4673 85.4673 0. 221 .00 1354.36 0.2 264.821 558.7 2.00 0.17 %Contribution (of VarComp) 84.804 0.66887 0.3 7304.00 Number of Distinct Categories = 1 Gage R&R for Response Graph window output ANOVA method with GAGEAIAG data 2003 Minitab Inc.Measurement Systems Analysis Part Operator Repeatability Total 2 2 22 26 38990 529 160703 200222 19495.00 39.00 15.64 100.8036 93.03618 0.00 Source Total Gage R&R Repeatability Reproducibility Operator Part-To-Part Total Variation StdDev (SD) 85.328 %Study Var (%SV) 91.50 8659.0547 Study Var (6 * SD) 512.00 0.965 Gage R&R Source Total Gage R&R Repeatability Reproducibility Operator Part-To-Part Total Variation VarComp 7304.55 100.

there is little difference between parts. In the By Part graph (located in upper right corner). you need at least five distinct categories to have an adequate measuring system.GAGEAIAG. little is due to differences between parts. • • Graph Window Output − GAGEAIAG. In this case. The Operator∗Interaction graph is a visualization of the p-value for Oper∗Part − 0. which does not account for this interaction. most of the points in the X-bar and R chart are outside the control limits. there are small differences between operators.MTW • In the Components of Variation graph (located in the upper left corner). In the By Part graph (located in upper right corner). Minitab fits the full model (shown in the ANOVA Table With Operator * Part Interaction (p = 0.25. Look at the %Contribution column in the Gage R&R Table. . See Number of distinct categories statement. • • • • Session Window Output − GAGE2. the ANOVA method will be more accurate than the X-bar and R method. telling you that most of the variation is due to the measurement system − primarily repeatability.67).25. The percent contribution from Part-To-Part (89. as shown by the nearly level line. • • Graph Window Output − GAGE2. When the p-value for Operator by Part is > 0. however. A 1 tells you the measurement system is poor. Minitab fits the model without the interaction and uses the reduced model to define the Gage R&R statistics. more than 10% of the variation is due to the measurement system. the percent contribution from Part-To-Part is larger than that of Total Gage R&R.MTW • 222 2003 Minitab Inc. According to the AIAG.64). In the By Operator graph (located in the middle of the right column). This tells you that much of the variation is due to differences between parts. Thus.33) is larger than that of Total Gage R&R (10. The percent contribution from Total Gage R&R (84. In the Components of Variation graph (located in the upper left corner).MTW • Look at the p-value for the Operator*Part interaction in the ANOVA Table. a Total Gage R&R %Contribution greater than 10% is poor. it cannot distinguish differences between parts. the number of distinct categories is four. most of the variation arises from the measuring system.36) is larger than that of Part-To-Part (15.00)). indicating variation is mainly due to differences between parts. telling you that much of the variation is due to differences between parts. In the Xbar Chart by Operator (located in lower left corner). the percent contribution from Total Gage R&R is larger than that of Part-to-Part.Quality Control ANOVA method with GAGE2 data Interpreting the results Session Window Output . For this data. Look at the %Contribution column in the Gage R&R Table. according to AIAG. as shown by the nonlevel line.MTW • • Look at the p-value for the Operator∗Part interaction in the ANOVA Table. When the p-value for Operator by Part is < 0.00 in this case − indicating a significant interaction between each Part and Operator. However. very little is due to differences between parts. as shown by the nearly level line. there are large differences between parts.

In the By Operator graph (located in the middle of the right column).MTW.00 2003 Minitab Inc. two times per part.030481 0. Under Method of Analysis. as shown by the level line.78 100. In Measurement data.51 2. Three operators measured the three parts. Example of a Gage R&R Study (Crossed) − (X-bar and R method) In this example.00 Source Total Gage R&R Repeatability Reproducibility Part-To-Part Total Variation VarComp 0.08848 %Study Var (%SV) 25. indicating the observed variation is mainly due to the measurement system. Three operators measured the ten parts.0011549 0. For comparison. enter Part. The Operator∗Interaction graph is a visualization of the p-value for Oper∗Part − 0.33 3. enter Operator. In Part numbers. three times per part. ten parts were selected that represent the expected range of the process variation. In Operators.16 18.033983 0.MTW). Session window output X-bar and R method with GAGEAIAG data Gage R&R Study . enter Part. Click OK. enter Response. in a random order.80 96.0308271 0.0009291 0.484 in this case − indicating the differences between each operator/part combination are insignificant compared to the total amount of variation. enter Response.20390 0.27390 0.73 16. 223 .18288 1. In Measurement data. Click OK.045650 0. most of the points in the X-bar and R chart are inside the control limits. choose Xbar and R.Measurement Systems Analysis • • • In the Xbar Chart by Operator (located in lower left corner).0329111 Source Total Gage R&R Repeatability Reproducibility Part-To-Part Total Variation StdDev (SD) 0. choose Xbar and R. Choose Stat > Quality Tools > Gage Study > Gage R&R Study (Crossed).175577 0. we analyze the data using both the X-bar and R method (below) and the ANOVA method. three parts were selected that represent the expected range of the process variation. there are no differences between operators. Choose Stat > Quality Tools > Gage Study > Gage R&R Study (Crossed).05346 1. we do a gage R&R study on two data sets: one in which measurement system variation contributes little to the overall observed variation (GAGEAIAG. Step 2: Use the X-bar and R method with GAGE2 data 1 2 3 4 5 6 7 Open the file GAGE2.67 100. For the GAGEAIAG data set. Under Method of Analysis.0020839 0. You can also look at the same data plotted on a Gage Run Chart. In Part numbers. in a random order. and one in which measurement system variation contributes a lot to the overall observed variation (GAGE2. For the GAGE2 data.82 93.181414 Study Var (6 * SD) 0.MTW. enter Operator.MTW). In Operators. Step 1: Use the X-bar and R method with GAGEAIAG data 1 2 3 4 5 6 7 Open the file GAGEAIAG.XBar/R Method %Contribution (of VarComp) 6.

0116 96.11 78.070 577.00 21.174 0.0000 45.0291 0.0291 85.174 510.94 0. .79 100.Quality Control Number of Distinct Categories = 5 Gage R&R for Response Number of Distinct Categories = 5 X-bar and R method with GAGE2 data Gage R&R Study .2081 Study Var (6 * SD) 510.000 270.99 Source Total Gage R&R Repeatability Reproducibility Part-To-Part Total Variation StdDev (SD) 85.00 Source Total Gage R&R Repeatability Reproducibility Part-To-Part Total Variation VarComp 7229.00 46.38 88.00 2026.XBar/R Method %Contribution (of VarComp) 78.89 100.05 9255.248 %Study Var (%SV) 88.38 0.00 Number of Distinct Categories = 1 Gage R&R for Response Graph window output X-bar and R method with GAGEAIAG data 224 2003 Minitab Inc.94 7229.11 0.

MTW Look at the %Contribution column in the Gage R%R Table. you would be better off using the ANOVA method for this data.89%) is due to differences between parts. In the X-bar Chart by Operator graph. Graph Window Output − GAGEAIAG. which was very large for this data set. although the X-bar and R method does not account for the Operator by Part interaction. That is because the X-bar and R method does not account for the Operator by Part effect. little is due to differences between parts (21.33%) is due to the measurement system (Gage R&R). Most of the points in the X-bar Chart are outside the control limits when the variation is mainly due to part-to-part differences. Graph Window Output − GAGE2. You may want to use the ANOVA method.89%). The measurement system variation (Total Gage R&R) is much smaller than what was found for the same data with the ANOVA method.MTW Look at the %Contribution column in the Gage R&R Table. 2003 Minitab Inc. as explained above.Measurement Systems Analysis X-bar and R method with GAGE2 data Interpreting the results Session Window Output − GAGEAIAG.67%) is due to differences between parts. a high percentage of variation (78. A large percentage (78. Most of the points in the X-bar chart will be within the control limits when the observed variation is mainly due to the measurement system. Here.MTW In the Components of Variation graph. 225 . when the number of distinct categories is 5 it represents an adequate measuring system. it cannot distinguish differences between parts.11%) is due to the measurement system (Gage R&R) − primarily repeatability. Here you get misleading estimates of the percentage of variation due to the measurement system. and a high percentage (93. See Number of distinct categories statement.MTW In the Components of Variation graph. According to AIAG.11%) of the variation in the data is due to the measuring system (Gage R&R). Gage R&R Study (Nested) Gage R& Gage repeatability and reproducibility studies determine how much of your observed process variation is due to measurement system variation. this plot shows you that the interaction is significant. A 1 in Number of distinct categories tells you the measurement system is poor. • Use Gage R&R Study (Crossed) when each part is measured multiple times by each operator. a low percentage of variation (6. Minitab allows you to perform either crossed or nested Gage R&R studies. the X-bar and R method grossly overestimates the capability of the gage. Session Window Output − GAGE2. and the low percentage (21. However. which accounts for the Operator by Part interaction.

Minitab provides two methods for assessing repeatability and reproducibility: X-bar and R. If you are unable to make that assumption then part-to-part variation within a batch will mask the measurement system variation. If you need to use destructive testing. such as in destructive testing. components. In fact. The X-bar and R method breaks down the overall variation into three categories: part-to-part. then choosing between a crossed or nested Gage R&R Study for destructive testing depends on how your measurement process is set up. If each batch is only measured by a single operator. you have a nested design. and operator-by-part. Crash testing is an example of destructive testing. and ANOVA. In fact. Operators: Enter the column containing operator names or numbers (optional). whenever operators measure unique parts. If each batch is only measured by a single operator. Use Gage R&R Study (Nested) when each part is measured by only one operator. you must be able to assume that all parts within a single batch are identical enough to claim that they are the same part. If you are unable to make that assumption then part-to-part variation within a batch will mask the measurement system variation. the measured characteristic is different after the measurement process than it was at the beginning. in part. then use Gage R&R Study (Crossed). <Gage Info> <Options> 226 2003 Minitab Inc.Quality Control • Use Gage R&R Study (Nested) when each part is measured by only one operator. then you must use Gage R&R Study (Nested). The ANOVA method goes one step further and breaks down reproducibility into its operator. you must be able to assume that all parts within a single batch are identical enough to claim that they are the same part. In destructive testing. Gage R&R Study (Nested) Stat > Quality Tools > Gage Study > Gage R&R Study (Nested) Gage repeatability and reproducibility studies determine how much of your observed process variation is due to measurement system variation. 9]. If you need to use destructive testing. because it considers the operator by part interaction [8. Gage R&R Study (Nested) uses the ANOVA method only. the measured characteristic is different after the measurement process than it was at the beginning. you have a nested design. The ANOVA method is more accurate than the X-bar and R method. and reproducibility. then you must use Gage R&R Study (Nested). such as in destructive testing. then use Gage R&R Study (Crossed). you can use Attribute Gage Study (Analytic Method) to analyze the bias and repeatability of your process. then choosing between a crossed or nested Gage R&R Study for destructive testing depends on how your measurement process is set up. repeatability. Crash testing is an example of destructive testing. Dialog box items Part or batch numbers: Enter the column containing the part or batch names or numbers. If all operators measure parts from each batch. If you can make that assumption. and operator-by-part. Gage R&R Study (Crossed) allows you to choose between the X-bar and R method and the ANOVA method. The ANOVA method goes one step further and breaks down reproducibility into its operator. If you can make that assumption. whenever operators measure unique parts. If all operators measure parts from each batch. If you have attribute data. Gage R&R Study (Nested) uses the ANOVA method for assessing repeatability and reproducibility. In destructive testing. . components. Measurement data: Enter the column containing the observed measurements.

83 1.33 . The default multiplier is 6.73% of your process measurements.43 1. you must be able to assume that all parts within a single batch are identical enough to claim that they are the same part. which is the number of standard deviations needed to capture 99. 227 . Part is nested within operator. Gage R&R Study (Crossed). Measure 1. You can estimate any missing observations with the methods described in [7].78 1.. use any of the available dialog box options. Gage Linearity and Bias Study. enter the column of operator names or numbers. because each operator measures unique parts. Measure 1. Parts and operators can be text or numbers. Gage Tolerance: Type the gage tolerance. and the observed measurement...83 1.. PartNum 1 1 2 2 3 3 4 4 5 . 2003 Minitab Inc. If you do not use these fields. Miscellaneous: Type any miscellaneous notes.. Operator Daryl Daryl Daryl Daryl Daryl Daryl Beth Beth Beth . This column shows the percentage of the process tolerance taken up by each variance component. If you like. Gage R&R Study (Nested). A %Tolerance column will be displayed in the bottom table of the Session window output and in the components of variation graph.48 1. Dialog box items Gage name: Type the name of the gage. operator..38 1.38 1. In Measurement data. Note If you use destructive testing. PartNum 1 1 2 2 3 3 1 1 2 . enter the column of measurements. enter the column of part or batch names or numbers. In Part or batch numbers. Dialog box items Study variation: Enter a multiplier to use in the StudyVar column of the Session window output. To do a gage R&R study (nested) 1 2 3 4 5 Choose Stat > Quality Tools > Gage Study > Gage R&R Study (Nested). Gage Info Stat > Quality Tools > Gage Study > choose Gage Run Chart.43 1.Measurement Systems Analysis Data − Gage R&R Study (Nested) Structure your data so that each row contains the part name or number. then click OK. The Gage R&R studies require balanced designs (equal numbers of observations per cell) and replicates. or Attribute Gage Study (Analytic Method) > Gage Info Use these fields to fill in the label on the Session window and graphical output.83 1.. PartNum1 for Daryl is truly a different part from PartNum1 for Beth... Each field corresponds to a line on the label..53 1.. the label will be printed with blank lines. In the example below.33 .83 1. Process tolerance: Enter a known tolerance range. Operator Daryl Daryl Daryl Daryl Daryl Daryl Beth Beth Beth .52 1. Reported by: Type the name of the person reporting the study information.78 1. Gage R&R Study (Nested) − Options Stat > Quality Tools > Gage Study > Gage R&R Study (Nested) > Options Allows you to control the Session and Graph window output. In Operators. Date of study: Type the date.53 1..48 1.52 1.

Session window output Gage R&R Study .46 82.00 Source Total Gage R&R Repeatability Reproducibility Part-To-Part Total Variation StdDev (SD) 1. In Operators.MTW. You decide to conduct a gage R&R study (nested) to determine how much of your observed process variation is due to measurement system variation.13549 1. Each part is unique to operator. enter Part.28933 1.25045 Study Var (6 * SD) 6.996 0.81293 0.0142 22. three operators each measured five different parts twice. Do not display percent contribution: Check to suppress display of percent contribution. Click OK. Draw figures on separate graphs. one figure per graph: Check to display each graph separately.28933 0.52374 1. In Measurement data.255 Gage R&R Source Total Gage R&R Repeatability Reproducibility Part-To-Part Total Variation VarComp 1. Example of gage R&R study (nested) In this example. no two operators measured the same part. In Part or batch numbers.00000 0.00 41.00000 0.54 100.14243 7. Gage R&R Study produces a layout of 6 graphs in one window.13549 0.27430 1. . A %Process column will be displayed in the bottom table of the Session window output and in the components of variation graph.83794 1. for a total of 30 measurements.00000 3.4094 MS 0.46 0.50273 %Study Var (%SV) 90. enter Operator.3400 41.00 17.81 90. Title: Enter a new title to replace the default title on the graphical output. Choose Stat > Quality Tools > Gage Study > Gage R&R Study (Nested).81293 6.88 100.Nested ANOVA Gage R&R (Nested) for Response Source Operator Part (Operator) Repeatability Total DF 2 12 15 29 SS 0.56364 %Contribution (of VarComp) 82.81 0. By default. enter Response. This column shows the percentage of process standard deviation taken up by each variance component.Quality Control Historical standard deviation: Enter a known historical standard deviation.28933 F 0.00385 1.00 Number of Distinct Categories = 1 Gage R&R (Nested) for Response 228 2003 Minitab Inc.0552 19.42549 P 0.00708 1. 1 2 3 4 5 6 Open the worksheet GAGENEST. Do not display percent study variation: Check to suppress display of percent study variation.

The percent contribution for differences between parts (Part-To-Part = 17. This indicates that most of the variation is due to measurement system error. and taste on a 1 to 10 scale In these situations. the possibility exists that the ratings are accurate. The measurements are subjective ratings by people rather than direct physical measurements. Attribute Agreement Analysis Attribute Agreement Analysis Stat > Quality Tools > Attribute Agreement Analysis Use attribute agreement analysis to assess the agreement of nominal or ordinal ratings given by multiple appraisers. Attribute Agreement Analysis.54) is much smaller than the percentage contribution for measurement system variation (Total Gage R&R = 82. To obtain meaningful classifications. quality characteristics are difficult to define and assess. A 1 in number of distinct categories tells you that the measurement system is not able to distinguish between parts. very little is due to differences between part. Examples include: • • • Automobile performance ratings Classification of fabric quality as "good" or "bad" Ratings of wine color. Look at the X-bar Chart − located in the lower left corner. See Number of distinct categories statement. Note Attribute Agreement Analysis was previously called Attribute Gage R&R Study in Minitab Release 13. 229 . rating usefulness is limited. more than one appraiser should classify the response measure. a method to examine the bias and repeatability of an attribute measurement system. Most of the points in the X-bar chart are inside the control limits when the variation is mostly due to measurement system error. If the appraisers agree.46). while a low percentage of variation is due to differences between parts. Dialog box items Data are arranged as Attribute column: Choose to enter the column containing the responses. Samples: Enter the column containing the sample or part number. Most of the variation is due to measurement system error (Gage R&R). is different from Attribute Gage Study (Analytic Method). a technique to assess appraisers' agreement. Look at the Components of Variation Graph − located in upper left corner. If the appraisers disagree. 2003 Minitab Inc.Measurement Systems Analysis Graph window output Gage R&R (Nested) for Response Interpreting the results Look at the %Contribution columns for Total Gage R&R and Part-to-Part. aroma.

the levels in a food tasting study may include crunchy. responses for each appraiser and each trial are in separate columns. For example. Number of trials: Type the number of trials. The assigned ratings can be nominal or ordinal: • • Nominal data are categorical variables that have two or more levels with no natural ordering. and crispy. Categories of the attribute data are ordered: Check to specify that the data have more than two levels and are ordinal. and strongly agree. neutral. Keep the trials for each appraiser in adjoining columns. mushy. For multiple column data. Appraiser names (optional): Type the appraisers' names or enter a column of names. . Data can be stacked in an attribute column or unstacked in multiple columns. Ordinal data are categorical variables that have three or more levels with a natural ordering. such as strongly disagree. The column can contain either numeric or text attributes. For attribute column data. The grouping indicators define each sample. Number of appraisers: Type the number of appraisers. Known standard/attribute (optional): Enter a column containing the attribute or known standard for each sample. Here is the same data set structured both ways: 230 2003 Minitab Inc. Multiple columns (Enter trials for each appraiser together): Choose to enter the columns containing the responses. agree. disagree. Grouping indicators and responses can be text or numeric. stack all of the responses in one column and set up grouping indicator columns for the appraiser and part number. <Information> <Options> <Graphs> <Results> Data − Attribute Agreement Analysis Data can be text or numeric. but the data type needs to match the response type.Quality Control Appraisers: Enter the column containing the appraiser name or number.

In Samples. check Categories of the attribute data are ordered. enter the data columns in Multiple columns. In Number of trials. 2003 Minitab Inc. enter the data column in Attribute column. If you do not use these fields. Each field corresponds to a line on the label. then click OK. you must enter the trials for each operator together. 3 4 Attribute Agreement Analysis − Information Stat > Quality Tools > Attribute Agreement Analysis > Information Use these fields to fill in the label on the Session window and graphical output.Measurement Systems Analysis Attribute column data Sample Appraiser 1 1 1 1 2 2 2 2 3 3 3 3 4 4 4 4 5 5 5 5 A A B B A A B B A A B B A A B B A A B B Response Good Good Bad Good Good Good Good Good Bad Good Bad Bad Good Good Good Good Bad Bad Good Bad Multiple columns data Sample Appraiser A − Trial 1 1 2 3 4 5 Good Good Bad Good Bad Appraiser A − Trial 2 Good Good Good Good Bad Appraiser B − Trial 1 Bad Good Bad Good Good Appraiser B −Trial 2 Good Good Bad Good Bad Note When your data are in multiple columns. The order of the trials must be the same for each operator. use any dialog box options. type the number of appraisers. To do an attribute agreement analysis 1 2 Choose Stat > Quality Tools > Attribute Agreement Analysis. as shown above. If your data have more than two levels and are ordinal. enter the column of appraiser names or numbers. In Number of appraisers. Minitab will display the label with blank lines. 231 . • If your data are unstacked and in multiple columns. Keep the trials for each appraiser together. type the number of trials. In Appraisers. enter the column of sample indicators. If you like. Do one of the following: • If your data are stacked in one column.

Percentages of assessment agreement within and between appraisers: Choose to display the percentages of assessment agreement within and between appraisers.MTW. Kendall's coefficient of concordance. Attribute Agreement Analysis − Results Stat > Quality Tools > Attribute Agreement Analysis > Results Use to control the display of Session window output. In addition. and Kendall's correlation coefficient. . Product name: Enter the name of the product. enter Rating. This graph is displayed only when the attribute is known and provided for each sample. Trials per appraiser and known standard/attribute: Check to display a graph of the matched proportions between the ratings of each appraiser and the attribute. In Known standard/attribute. plus kappa coefficient tables and Kendall's coefficient (ordinal data only) tables. Dialog box items Calculate Cohen's kappa if appropriate: Check to calculate Cohen's kappa when appropriate. Choose Stat > Quality Tools > Attribute Agreement Analysis. Miscellaneous: Enter any miscellaneous notes. kappa and Kendall's (ordinal data) coefficient: Choose to display the output described above. enter Appraiser. 1 2 3 4 5 6 7 Open the worksheet ESSAY. enter Attribute. Minitab will still display graphs. Confidence level: Enter the confidence level of the interval estimation of the percentages of the assessment agreement within appraisers and between each appraiser and standard. In Appraisers. Example of Attribute Agreement Analysis An educational testing company is training five new appraisers for the written portion of the twelfth-grade standardized essay test. Dialog box items Plot the percentage of agreed responses among Trials per appraiser (no graph for single trial): Check to display a graph of the matched proportions for each appraiser. Cohen's kappa. Check Categories of the attribute data are ordered and click OK. -1. Reported by: Enter the name of the person reporting the study information. 232 2003 Minitab Inc. Attribute Agreement Analysis − Graphs Stat > Quality Tools > Attribute Agreement Analysis > Graphs Use to choose which graphs to display. 2).Quality Control Dialog box items Date of study: Enter the date the study was conducted. The appraisers' ability to rate essays consistent with standards needs to be assessed. In Attribute column. The default is set at 95. Attribute Agreement Analysis − Options Stat > Quality Tools > Attribute Agreement Analysis > Options Use to specify a value for the confidence level. This graph is displayed when the number of trials for each appraiser is more than one. Display figures on one graph: Check to display both graphs on one page. Dialog box items Control the Display of Results Display nothing: Choose to suppress all Session window output. 1. enter Sample. Each appraiser rated fifteen essays on a five-point scale (-2. Minitab will calculate Cohen's kappa when two appraisers rate a single trial or when each appraiser rates two trials. In Samples. including Fleiss' kappa. 0.

0001 0.00) (81.64550 1.258199 0.33 86.00000 1.0000 0.87298 7.44099 0.0008 0.14667 5.87298 6.00000 1.81366 1.94347 3.130924 Z 2.87298 3.258199 0.93955 5.87298 2. 98.70796 1.258199 0.34) (81.258199 0.87298 3.87298 3.2593 0.14508 2. 233 .16667 0.0001 0.192450 0.258199 0.0000 0.00 100.0000 0.42308 0.61554 4.00000 1.81366 0.Measurement Systems Analysis Session window output Attribute Agreement Analysis for Rating Each Appraiser vs Standard Assessment Agreement Appraiser Duncan Hayes Holmes Montgomery Simpson # Inspected 15 15 15 15 15 # Matched 8 13 15 15 14 Percent 53.14667 4.258199 0.00000 1.00000 1.63857 3.0074 0.0001 0.58333 0.41176 0.59.89779 0.258199 0.258199 0.00000 1.258199 0.43855 3.00000 1.87298 3.05.61584 3.258199 0.00000 1.0001 0.82955 1.00 93.91597 SE Kappa 0.258199 0.0000 0.90.258199 0.44099 0.25924 0.0001 0.0008 0.00000 1.258199 0.192450 Z 4.192450 0.0001 0.00000 0.81366 1.130924 0.258199 0.131305 0. 100.0000 0.258199 0.87298 7.0119 0.0001 0.0016 0.0001 0.87298 3.83) # Matched: Appraiser's assessment across trials agrees with the known standard.192450 0.87298 3.00000 0.258199 0.0000 Hayes Holmes Montgomery Simpson Kendall's Correlation Coefficient Appraiser Duncan Hayes Holmes Montgomery Simpson Coef 0. 100.0000 0.258199 0.258199 0.15131 3.96014 1.258199 0.33 95 % CI (26.81366 0.62963 0.0001 0.18307 3.0001 0.258199 0.0008 0.93258 SE Coef 0.00000 1.00000 1.15131 3.0000 0.0438 0.0008 0.99619 P(vs > 0) 0.258199 0.134164 0.15131 6. 99.0001 0.00000 0.87298 3.192450 0.76000 0.73) (59.61584 3.67 100.131305 0. Fleiss' Kappa Statistics Appraiser Duncan Response -2 -1 0 1 2 Overall -2 -1 0 1 2 Overall -2 -1 0 1 2 Overall -2 -1 0 1 2 Overall -2 -1 0 1 2 Overall Kappa 0.00) (68.54.00000 1.0008 0.258199 0.79636 P 0.00000 1.258199 0.258199 0. 78.87298 3.00000 0.00000 1.87298 3.0438 0.87298 3.70796 1.90.0001 0.0001 0.258199 0.0000 Between Appraisers Assessment Agreement 2003 Minitab Inc.0001 0.15131 3.0507 0.

642479 0. 67.0207 16.831455 SE Kappa 0.3383 14.1100 P 0.3822 8.958102 SE Coef 0. Fleiss' Kappa Statistics Response -2 -1 0 1 2 Overall Kappa 0.3331 7.0000 0.6663 7. Fleiss' Kappa Statistics Response -2 -1 0 1 2 Overall Kappa 0.0816497 0.0000 0.1136 P(vs > 0) 0.115470 0.0000 All Appraisers vs Standard Assessment Agreement # Inspected 15 # Matched 6 Percent 40.680398 0.0000 Kendall's Coefficient of Concordance Coef 0.0000 0.71) # Matched: All appraisers' assessments agree with the known standard.0000 0.Quality Control # Inspected 15 # Matched 6 Percent 40.796066 0.34.Sq 67.3693 6.115470 0.847348 0.058911 Z 7. 67.707602 0.602754 0.0860663 Z 11.0000 0.0000 0.672965 SE Kappa 0.0816497 0.0816497 0.0816497 0.842593 0.0000 0.115470 0.850932 0.736534 0.0000 234 2003 Minitab Inc.8941 7.71) # Matched: All appraisers' assessments agree with each other.115470 0.115470 0.0000 0.3210 P(vs > 0) 0.0000 Kendall's Correlation Coefficient Coef 0.0816497 0.6422 DF 14 P 0.0000 0.0412331 Z 8.0000 0.802932 0.9536 7.8687 9.2971 6.34. .00 95 % CI (16.00 95 % CI (16.966317 Chi .

however. You can easily compare the assessment agreements for the five appraisers. Holmes and Montgomery. Between Appraisers. Because no actual measurements from the gage are available to estimate bias and repeatability. However. Hayes. you conclude that Duncan. Kappa and Kendall's statistics are included in each of these tables. Note that each percentage matched is associated with a confidence interval. Attribute Gage Study (Analytic Method) Attribute Gage Study (Analytic Method) Οverview Attribute gage studies assess the amount of bias and repeatability of a gage when the response is a binary attribute variable. you must have exactly 20 trials per part. 235 . Minitab calculated bias and repeatability by fitting the normal distribution curve using the calculated probabilities of acceptance and the known reference values for all parts. The number of required runs performed on each part depends on the method you use to test the bias of the gage. You should select the parts at nearly equidistant intervals and know the reference value for each selected part. and All Appraisers vs Standard. The minimum number of trials necessary to use the regression method is 15. The Appraiser vs Standard graph provides a graphical display of the Each Appraiser vs Standard assessment agreement table.966317 (p=0. and Simpson need additional training. Based on this data. However. Overall. Minitab offers two methods to test whether the bias is zero: AIAG and regression. Minitab calculates the bias and repeatability of the gage using the method described in [1]. 2003 Minitab Inc. Kendall's coefficient for all appraisers vs the standard is 0.0000). The overall Kendall coefficient between appraisers is 0. for a percentage match of 100%. matched 15 of 15 assessments.0000). an attribute gage does not give information on the quality of a part but only indicates whether a part is accepted or rejected based on a set of specified standards. Note Use Attribute Agreement Analysis when you assess the agreement of ratings given by multiple appraisers. See Data − Attribute Gage Study (Analytic Method) for more information on the data requirements.958102 (p=0. Unlike a variable gage.Measurement Systems Analysis Graph window output Interpreting the results Minitab displays three assessment agreement tables: Each Appraiser vs Standard. Attribute gage study results depend heavily on the way you select the parts and the number of runs performed on each part. If you use AIAG (default). a closer look at the Each Appraiser vs Standard table indicates that Duncan's and Hayes' assessments match poorly with the standard. Note The Within Appraiser table is not displayed because each appraiser conducted a single trial. 20 or more trials are recommended. these statistics suggest good agreement.

Attribute for calculation: Enter the attribute for estimating probabilities. For the lower limit. You can enter the number of trials as a constant or a column of data. The number of trials for each part must be greater than or equal to 15. Minitab accepts either summarized or raw data for attribute gage studies. When the number of trials is not identical for all parts. a method to examine the precision of an attribute measurement system. the part with the lowest reference value must have 0 acceptances and the part with the highest reference must have the maximum number of possible acceptances. . Note Attribute Gage Study (Analytic Method). See Data − Attribute Gage Study (Analytic Method) for more information. you must enter a column containing the number of trials for each part. Dialog box items Part numbers: Enter the column containing the part names or numbers. For the upper limit. <Gage Info> <Options> Data − Attribute Gage Study (Analytic Method) You must have at least eight parts to conduct an attribute gage study. you should strictly follow the rules in [1] to select parts with known reference values. You must specify either a lower or upper tolerance limit Display the other limit: Enter another limit to display on the Gage Performance Curve (optional). the number of acceptances increases. Tolerance limit for calculation Lower limit: Enter the lower tolerance limit. you must have exactly 20 trials per part. For the AIAG method for testing bias. 236 2003 Minitab Inc. You must specify either a lower or upper tolerance limit Upper limit: Enter the upper tolerance limit. Attribute Agreement Analysis was previously called Attribute Gage R&R Study in Minitab Release 13. If you specify the lower tolerance limit. Raw data: Choose to enter a column containing the raw data. If you specify the upper tolerance limit. is different from Attribute Agreement Analysis. The smallest part must have zero acceptances. Number of trials: Enter the number of trials. To obtain good estimates of bias and repeatability. You must set up your reference value column differently for summarized and raw data. If you use the AIAG method (default) to test the bias of the gage.Quality Control Attribute Gage Study (Analytic Method) Stat > Quality Tools > Gage Study > Attribute Gage Study (Analytic Method) Attribute gage studies calculate the amount of bias and repeatability of a measurement system when the response is a binary attribute variable. exactly six parts must be greater than zero number of acceptances and less than 20--the maximum number of acceptances allowed in the AIAG situation. a technique to assess appraisers' agreement. Attribute label: Enter the attribute label (default is Acceptance). Reference values: Enter the column containing the numeric reference values. the part with the lowest reference value must have the maximum number of possible acceptances and the part with the highest reference value must have 0 acceptances. and the largest part must have the maximum number of possible acceptances. For the regression method. you may have more than six parts between the two extreme reference values. the number of acceptances decrease. Binary attributes in Summarized counts: Choose to enter a column containing the summarized counts. as the reference values increase. which must match one of the two raw data values. as the reference values increase.

To do an attribute gage study (Analytic method) 1 2 3 4 Choose Stat > Quality Tools > Gage Study > Attribute Gage Study (Analytic Method). which must match one of the two distinct responses. In Attribute label. reference value. In Part numbers. the label will be printed with blank lines. In Lower limit.. If you like. In Attribute for calculation. In Reference values. • When the binary responses are raw data.65 1.45 1. enter the number of trials.40 1. reference value. Gage R&R Study (Nested).70 Response Reject Reject Reject Reject .. Date of study: Type the date.. Reference values and counts must be numeric Structure raw data so that each row contains the part name or number. use any dialog box options.70 1. type the attribute label (the default is Acceptance). Gage R&R Study (Crossed). 8 8 8 8 Reference 1.50 1. 1. Miscellaneous: Type any miscellaneous notes.. Each field corresponds to a line on the label. Accept Accept Accept Accept Structure summarized data so that each row contains the part name or number. enter the column of part numbers.35 1. and summarized count.35 1. and binary response. or Attribute Gage Study (Analytic Method) > Gage Info Use these fields to fill in the label on the Session window and graphical output. In Upper limit. If you do not use these fields. then click OK. Dialog box items Gage name: Type the name of the gage. Parts and responses can be text or numbers.Measurement Systems Analysis Summarized Data Part Number Reference 1 2 3 4 5 6 7 8 1. Do one of the following: • When the binary responses are summarized counts. 2003 Minitab Inc. enter the column of counts in Summarized Counts. enter the column of reference values.35 .70 1. type the upper tolerance limit.35 1. enter the column of raw data in Raw data. Reported by: Type the name of the person reporting the study information. Parts can be text or numbers. 5 6 Gage Info Stat > Quality Tools > Gage Study > choose Gage Run Chart.35 1. 237 ..70 Raw Data Part Number Acceptances 0 3 8 13 15 18 19 20 1 1 1 1 . Attribute Gage Study (Analytic Method) − Options Stat > Quality Tools > Gage Study > Attribute Gage Study (Analytic Method) > Options Allows you to determine a method to test the bias of the gage and enter a new title to replace the default title.60 1. Gage Tolerance: Type the gage tolerance. type the attribute for calculation. Reference values must be numeric.55 1.. Gage Linearity and Bias Study.70 1. In Number of trials. type the lower tolerance limit.

020 and an upper tolerance of 0. 238 2003 Minitab Inc. df=19.Quality Control Dialog box items Test bias = 0 using AIAG Method: Choose to use the AIAG Method. Graph window output Interpreting the results The bias in the attribute gage system is 0. enter Reference.005 from −0. In Number of trials.0097955. through the gage 20 times. suggesting that bias is present in the attribute measurement system.MTW. Click Summarized counts and enter Acceptances. The manufacturer ran 10 parts. Click OK.020. type 20. p = 0. enter Part number. and the adjusted repeatability is 0.020.005. which had reference values at intervals of 0. . In Part numbers. 1 2 3 4 5 6 Open the worksheet AUTOGAGE.050 to −0.00). Title: Enter a new title to replace the default title on the graphical output. Choose Stat > Quality Tools > Gage Study > Attribute Gage Study (Analytic Method). Regression Method: Choose to use the Regression Method. Example of Attribute Gage Study (Analytic Method) An automobile manufacturer wants to measure the bias and repeatability of an automated attribute measurement system. Click Lower limit and type -0.70123.0458060. The test for bias indicates that bias is significantly different from 0 (t = 6. The system has a lower tolerance of −0. In Reference values.

so you do not have to record the "size" of the invoices sampled. you may record the number of errors in billing statements. Minitab provides capability analyses based on both normal and nonnormal probability models. Each item can result in one of two possible outcomes ("success/failure". The billing statements may be considered to be the "same size. Or. If you are an assembler. and the number of people scheduled to work that day. You determine capability by comparing the width of the process variation with the width of the specification limits. if you manufacture appliances. The rate of defects per unit of space or time is the same for each item. Choosing a capability command Minitab provides you with the options of identifying the distribution that fits your data or to transform your data to follow normal distribution.). or statistics. For example. You can use individual distribution identification to select the distribution that best fits your data prior to conducting a capability analysis. you can use capability statistics to compare the capability of one process to another. The commands that use a normal probability model provide a more complete set of statistics. • A Poisson distribution could model these examples if the following conditions are met: • • Process Capability Overview Once a process is in statistical control. you might have a go/no-go gauge that determines whether an item is defective or not. prior to using capability analysis. so you may want to record the size of each surface sampled. Binomial data Binomial data is usually associated with recording the number of defective items out of the total number of items sampled. The process needs to be in control before you assess its capability. Because the process information is reduced to a single number. you can perform capability analyses for: • • • normal or nonnormal probability models (for measurement data) normal data that might have a strong source of between-subgroup variation binomial or Poisson probability models (for attributes or count data) It is essential to choose the correct distribution when conducting a capability analysis. Capability indices. but your data must approximate the normal distribution 2003 Minitab Inc. You can assess process capability graphically by drawing capability histograms and capability plots. you may want to record the number of scratches on one of the surfaces of the appliance." because they have the same number of opportunities for errors. 239 . you could record the number of parts sent back due to poor fit in the assembly process and the total number of parts purchased. you probably then want to determine if it is capable. "go/no-go. if it is not. Because the sizes of the items may vary. you may want to record the number of breaks in a piece of wire. Poisson data Poisson data is usually associated with the number of defects observed in an item. The probability of a success (or failure) is constant for each item. where the item occupied a specified amount of time or a specified space. you may want to keep track of each item's size. The number of defects observed in the items are independent of each other. which are ratios of the specification tolerance to the natural process variation. Depending on the nature of data and its distribution. You can also calculate capability indices. For example. that is producing consistently. Or. These graphics help you assess the distribution of your data and verify that the process is in control. you could record the number of people who call in sick on a particular day. that is meeting specification limits and producing "good" parts. are a simple way of assessing process capability. if you are a manufacturer. The lengths of the pieces may vary. then you will get incorrect estimates of process capability. Here are some examples: • • If you manufacture electrical wiring. The size of the surfaces may be different for different appliances. You could then record the number of items that were failed by the gauge and the total number of items inspected. so you may also want to record the length for each piece of wire.Process Capability Overview Process Capability for Attributes Data You can produce process capability studies for either binomial or Poisson data. The outcomes of the items are independent of each other." etc. Or. These examples could be modeled by a binomial distribution if the following conditions are met: • • • • Each item is the result of identical conditions.

The report also includes a table of between/within and overall process capability statistics. Similarly. Based on probability plots and goodness-of-fit tests. The report also includes a table of overall process capability statistics. which helps you to verify that the process is in a state of control. which helps you make a visual assessment of the assumption of normality. For example. With Minitab. Capability Analysis (Nonnormal) calculates parts per million out-of-spec using a nonnormal distribution that best fits your data. Capability Analysis (Normal) produces a process capability report when your data do not follow a normal distribution. the validity of the statistics depends on the validity of the assumed distribution. Understanding both sources of subgroup variation may provide you with a more realistic estimate of the potential capability of a process. Capability Analysis (Normal) estimates expected parts per million out-of-spec using the normal probability model. based on the binomial and Poisson probability models. If you suspect that there may be a strong between-subgroup source of variation in your process. Minitab also allows you to store the transformed data for future analysis. use Capability Analysis (Between/Within) or Capability Sixpack (Between/Within). you can transform the data using Johnson distribution system or Box-Cox power transformation or use nonnormal probability model. it is better to either transform the data to make the normal distribution a more appropriate model. Capability Analysis (Between/Within) and Capability Sixpack (Between/Within) computes both within and between standard deviations and then pools them to calculate the total standard deviation. If the data are badly skewed. You can also use this feature when a grouping variable is associated with a continuous variable. For example. or choose a nonnormal probability model for the data. which can be used to verify that the process is in a state of control − a capability histogram and normal probability plot. In that case. You can also classify products based on the number of defects (use Capability Analysis (Poisson)). The report draws a U chart. Capability Sixpack (Between/Within) combines the following charts into a single display. Capability Analysis (Binomial) is appropriate when your data consists of the number of defectives out of the total number of parts sampled. histogram of %defectives. Minitab also provides capability analyses for attributes (count) data. Subgroup data may have. Capability Sixpack (Normal) combines the following charts into a single display. and that they follow an approximately normal distribution. along with a subset of the capability statistics: • • • • • • • • • • 240 2003 Minitab Inc. Use this analysis for subgroup data in which there is a strong between-subgroup source of variation. Capability Analysis (Nonnormal) draws a capability histogram of the individual measurements overlaid with a distribution curve based on the process parameters. Here is a summary of Minitab's capability commands: • Individual Distribution Identification allows you to fit data with 15 parametertric distribution families. Capability Analysis Multiple Variables (Nonnormal) allows you to compare process capability for multiple variables that do not follow or transformed to follow normal distribution. The report also includes a chart of the cumulative mean DPU (defects per unit). Johnson Transformation transforms your data to follow normal distribution using Johnson distribution system. you want to compare the width of parts from the same process before and after a process improvement. Nonnormal data compares these two methods. the estimated proportion of defective items may be extremely over or under estimated. . you can choose a distribution that best fits your data prior to conducting a capability analysis. along with a subset of the capability statistics: − an Xbar (or Individuals). which displays the process variability compared to the specifications Capability Sixpack (Between/Within) is appropriate for subgroup data in which there is a strong between-subgroup source of variation. you want to compare the width of the parts from the same process before and after a process improvement.Quality Control for the statistics to be appropriate for the data. The report also includes a table of process capability statistics. in addition to random error within subgroups. in addition to the within-subgroup variation. The report also includes probability plot to assess whether a distribution fits your data well. which helps you verify that the process is in a state of control. R or S (or Moving Range). This helps you make a visual assessment of the assumption that your data follow a specified distribution. including both within and overall statistics. The report includes histogram with overlaid normal curve and capability statistics for each variable or group. and defective rate plot. The report includes probability plot and Anderson-Darling statistics for original and transformed data for comparison. The report also includes a chart of cumulative %defectives. Capability Analysis (Poisson) is appropriate when your data take the form of the number of defects per item. You can also use this feature to perform capability analysis for a process with one output but having a grouping variable associated with it. The report includes a capability histogram overlaid with a distribution curve based on the distribution parameters to assess whether your data follows the selected distribution. and a defect rate plot. Interpretation of these statistics rests on two assumptions: that the data are from a stable process. The report draws a P chart. and run chart. Capability Analysis (Between/Within) draws a capability histogram of the individual measurements overlaid with normal curves. In both cases. histogram of DPU. Capability Analysis Multiple Variables (Normal) allows you to compare process capability for multiple variables following or transformed to follow normal distribution. For example. which can be used to verify that the data are normally distributed − a capability plot. products may be compared against a standard and classified as defective or not (use Capability Analysis (Binomial)). For example. random variation between subgroups. The report includes histogram with overlaid distribution cure based on process parameters and capability statistics for each variable or group.

which can be used to verify that the process is in a state of control a capability histogram and normal probability plot. Continuing Process Control and Process Capability Improvement. pp 133−141.135th percentile) Based on the default (ISO) method. which displays the process variability compared to the specifications Although the Capability Sixpack commands give you fewer statistics. 30.A. Owen. April. "Bootstrap Lower Confidence Limits for Capability Indices. July. [5].L." Journal of Quality Technology.B.A. Ford Motor Company (1983). it is not good practice to rely on a single capability statistic to characterize a process. L. "Lower Confidence Limits on Process Capability Indices. Cheng. and R.(or Individuals)." Journal of Quality Technology. [10]. Mason (1998). L.196−210.Process Capability − − − • an Individuals Chart. Borrego (1990). 241 . but they have distributional properties that are not fully understood.965th percentile − 0. For a stable process. and R Chart or S Chart. D. pp. D'Agostino and M. Use the table below as a guide: Assumed distribution Normal (within analysis) Normal (between/within analysis) Capability measures Potential capability Parameters estimates used Process mean (Xbar) Within-subgroup standard deviation Process mean (Xbar) Between/Within standard deviation Process mean (Xbar) Overall standard deviation Process median (50th percentile) Process spread (99.S. [6]. S. (1983). and run chart. The parameters used in estimating the indices differ based on the distribution used in the analysis. McGraw Hill. D. [9]. Michigan.R. "Measuring Process Capability".A. Moving Range Chart. S. Stephens (1986). pp.223−229. which can be used to verify that the data come from a specified distribution − a capability plot. Y. Capability statistics are simple to use. "Transforming nonnormal Data to Normality in Statistical Process Control. 20. Polansky.K. In general. "A New Measure of Process Capability: Cpm. which displays the process variability compared to specifications Capability Sixpack (Nonnormal) combines the following charts into a single display. 22.4. reliable estimates can be made of the various process parameters required for capability studies. and Shenton. Chan. October. R. Ford Motor Company. which can be used to verify that the process is in a state of control − a capability histogram and probability plot.M. Y. L. and F. Franklin and G.R. Bothe (1997). which can be used to verify that the data are normally distributed a capability plot. Chou. Goodness-of-Fit Techniques. pp. [4]. and [11] for a discussion. See [2]. 2003 Minitab Inc. July. Between/Within capability Normal (within and within/between analysis) Nonnormal Overall capability Overall capability References − Process Capability [1] [2] [3] [4] [5] [6] [7] [8] Bowman. Wasserman (1992). Note Process parameters Process capability statistics are numerical measures of process capability − that is. Chou.O." Journal of Quality Technology. Spiring (1988). A. 303 − 314. pp. Marcel Dekker. K. Dearborn. Encyclopedia of Statistical Sciences Vol.(or Moving Range).162−175. 24." Journal of Quality Technology. "Johnson's System of Distributions". along with a subset of the capability statistics: − an Xbar. it provides an array of charts that can be used to verify that the process is in control and that the data follow the chosen distribution. they measure how capable a process is of meeting specifications. R.W.

[4].L. Yen and A." Quality Progress.E. pp.15− 21. March. B. • • To check which distribution best fits your data. Classical and Modern Regression with Applications. "Modified goodness-of-fit test for the Laplace distribution." Journal of Quality Technology. Capability Analysis (Nonnormal). Academic Press. Sullivan (1984). [20] C.176−187.P. D." Journal of Quality Technology.F. Indiana. Second Edition. pp.S. 216−231. October. "Recent Developments in Process Capability Analysis. Jaehn (1989). "Development and Evaluation of Control Charts Using Exponentially Weighted Moving Averages. Statistical Quality Control Handbook. Gunter (1989). Ng and K. Ryan (1989).H. "Distributional and Inferential Properties of Process Capability Indices. Chambers.E. use Individual distribution identification. Ed. Johnson and S. Capability Sixpack (Normal). pp. Hurley (1992)." Journal of the Royal Statistical Society. A. May. (2000).72. J. John Wiley & Sons. 109. [21] [22] [23] [24] [25] [26] [27] [28] [29] F. S. "Goodness of fit for the inverse Gaussian distribution. R. Capability Analysis Multiple Variables (Normal).P. Rueda (1992). October. "Estimation and Tests of Fit for the Three-parameter Weibull Distribution. and Capability Sixpack (Between/Within) − Johnson transformation.A. Kushler and P.A. 242−250. W. Godfrey (1986). N.H. pp. Capability Analysis Multiple Variables (Nonnormal). R. H." The Canadian Journal of Statistics. [9]. (1972). Brooks/Cole Publishing Company. W.C. Kotz. use Individual distribution identification to select the distribution that best fits your data. Statistical Models and Methods for Lifetime Data. R. 24." Quality Progress. 80. 22. Capability Analysis (Normal). Kane (1986).188−195. John Wiley & Sons. yet many units will still be "out of spec. . Moore (1998). Johnson (1992). Discrete Distributions. [19] R. 22. July. 41−52.Quality Control [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] B. L. October. "How to Estimate Percentage of Product Failing Specifications. Myers.L. K. Kotz (1969). "The Use and Abuse of Cpk. the capability indices based on normal distribution such as Ppk and Cpk may look "good". "Reducing Variability: A New Approach to Quality. One advantage of Capability Sixpack is that it includes the appropriate charts as part of its output. 24.J. Escobar (1998). O'Reilly and R. [5]." Journal of Quality Technology." When the distribution of your data is not known. Lawless (1982). Meeker and L. you can either transform the data in such a way that the normal distribution is a more appropriate model. 1984. "Confidence Bounds for Capability Indices. Statistical Methods for Reliability Data. S. V. John Wiley & Sons. and N. In order to make proper use of the capability statistics. J. W.H. Understanding Statistical Process Control. Vol 20. or choose a nonnormal probability model for the data. 491-500. A Vol 17.L.Q. Part 2. Lockhart and M. "The Use and Abuse of Cpk. Stephens. and [11] for a discussion. Numerical Methods for Unconstrained Optimization." Journal of Quality Technology. 242 2003 Minitab Inc. 18. use Johnson transformation. 24.B. Indianapolis. Modern Methods for Quality Control and Improvement. Minitab also allows you to transform your data to follow normal distribution using Box-Cox or Johnson transformation methods. V.H.N. pp. pp. B Vol 56. "Process Capability Indices. Inc. H. Inc. pp. Inc. Gunter (1989). use Individual distribution identification.108. Murray. See [2]. [10]." Journal of Quality Technology. You should verify that the process is stable by checking a control chart for the location (such as an Xbar chart or an I chart) and one for the variation (such as an R chart. Pp 275−281. an S chart or an MR chart). Part 3. pp.227−228. pp.79. Assumptions of Process Capability There is no capability unless the process is in control. John Wiley & Sons." Quality Progress. Pearn." Communication Statistics. 21. Western Electric (1956). [6]. and Capability Sixpack (Nonnormal). Case (1989). Capability Analysis (Between/Within). Wheeler and D. pp 387-397. John Wiley & Sons. Rodriguez (1992). Nonnormal data When you have nonnormal data. Stephens (1994).A. Wadsworth. (1992). When observations come from a skewed distribution. T. and A. 3rd edition. To transform the data to follow a normal distribution − Box-Cox power transformation. Validity of statistics is based on validity of assumed distribution.M." Tappi. Western Electric Corporation. SPC Press. Statistical Methods for Quality Improvement.

and Pp are the same. Cpk. you can use capability statistics to compare the capability of one process to another. 243 . Use Cpm to assess if the process is centered at target. Pp. The higher the Cpm index. The capability indices based on a nonnormal distribution can be interpreted in the same manner as the indices based on a normal distribution. For example. then the overall standard deviation approximately equals the within-subgroup standard deviation. Examples of Capability Analysis Capability analysis (normal) Capability analysis (normal) using a Box-Cox transformation Capability analysis (between/within) Capability analysis (nonnormal) Capability analysis multiple variables (normal) Capability analysis multiple variables (nonnormal) Capability analysis (Binomial) 2003 Minitab Inc.33 to be a minimum acceptable value for the process capability statistics. Normal distribution: • Cp and Cpk − Cp does not consider the location of the process mean in relation to the specification limits. 50th.13th percentiles is equivalent to the 6σ spread in the normal case.13th. If Cpm. use Capability Analysis (Nonnormal). If Cpk and Ppk are the same. Compare Pp and Ppk to assess if the process median is close to the specification midpoint. and overall capability Nonnormal model Uses actual data units for the histogram Calculates only overall capability Draws a normal curve over the histogram to help you Draws the chosen nonnormal distribution curve over the determine whether the transformation made the data "more histogram to help you determine whether the data fits the normal" specified distribution Which method is better? The only way to answer that question is to see which model fits the data better. and Ppk statistics are measures of potential and overall capability. The capability statistics differ in interpretation depending on the distribution used for the analysis.Process Capability • To use a nonnormal probability model. it is probably better to choose the normal model. and Capability Sixpack (Nonnormal). the process median is off the specification midpoint and closer to one of the specification limits. • • • Nonnormal distribution: Minitab calculates the capability statistics using 0. Cpk is calculated using the within-subgroup variation. Ppk is for the whole process. Many practitioners consider 1. will result in low Cpk and high Cp. Because the process information is reduced to a single number. then the process mean coincides with the target. use the information below for guidance. The distance between the 99. Cpm − Minitab calculates Cpm when you specify a target value. If Pp is greater than Ppk. between/within (when both within and between subgroup variation exists). Cpk vs Ppk − Minitab calculates Ppk using the overall variation Both the between-subgroup and within-subgroup contributes to the overall variation. the better the process. Capability statistics allows you to monitor and report the improvement of the system over time. The 50th percentile represents the process median. Ppk. since it provides estimates of both overall and within process capability. To understand and interpret the capability statistics. a process may be performing within minimum variation but away from the target (or process mean) and closer to one of the specification limits. Cpk is higher only when you are meeting the target (or process mean when no target value is specified) with minimum variation. Capability Analysis Multiple Variables (Nonnormal). This table summarizes the differences between the models. Interpreting capability statistics Cp. and 99. Pp and Ppk − Pp and Ppk are interpreted in the same way as Cp and Cpk respectively. Normal model Uses actual or transformed data for the histogram Calculates within. Cpk is an index which measures how close a process is running to its specification limits in relation to the process spread. For. but not the shift and drift between subgroups.87th percentiles for the distribution used in the analysis.87th and 0. If both models fit the data about the same. A value less than 1 indicates that your process variation is wider than the specification spread.

To change the length of the moving range from 2. • the average of the subgroup standard deviations − choose Sbar. check Use moving range of length and enter a number in the box. • the pooled standard deviation (the default) − choose Pooled standard deviation. Capability Analysis Multiple Variables (Normal) − Within subgroups. . check Use moving range of length and enter a number in the box. or Capability Sixpack (Normal) main dialog box. To not use an unbiasing constant in the estimation. see [1]. base the estimate on: • the average of the moving range (the default) − choose Average moving range. base the estimate on: • the average of the subgroup ranges − choose Rbar. 3 Click OK. Do one of the following: For subgroup sizes greater than one. For individual observations (subgroup size is one).Quality Control Capability analysis (Poisson) Examples of Capability Sixpack Capability sixpack (normal) Capability sixpack (normal) using Box-Cox transformation Capability sixpack (between/within) Capability sixpack (nonnormal) Examples of Quality Tools The following examples illustrate how to use the various quality tools. For a discussion of the relative merits of these methods. Choose an example below: Run Chart Pareto Chart Cause-and-Effect Individual Distribution Identification Johnson Transformation Capability Analysis Capability Sixpack Gage Run Chart Gage Linearity and Bias Study Gage R&R Study (Crossed) (ANOVA method) Gage R&R Study (Crossed) (X-bar and R method) Gage R&R Study (Nested) Attribute Gage Study (Analytic Method) Attribute Agreement Analysis Multi-Vari Chart Symmetry Plot To specify a method for estimating within-subgroup standard deviation 1 2 In the Capability Analysis (Normal). uncheck Use unbiasing constants. • the square root of MSSD (mean of the squared successive differences) − choose Square root of MSSD. Note 244 2003 Minitab Inc. To change the length of the moving range from 2. uncheck Use unbiasing constants. uncheck Use unbiasing constants. click Estimate. • the median of the moving range − choose Median moving range. To not use an unbiasing constant in the estimation. To not use an unbiasing constant in the estimation.

Specify: Choose to specify one to four distributions to fit your data. uncheck Use unbiasing constants. To change the length of the moving range from 2. click Estimate. choose one of the following: • the average of the subgroup ranges − choose Rbar. then enter a column. Choose from 14 distributions. Distribution 2: Check to fit a distribution (default is Exponential). uncheck Use unbiasing constants. the square root of MSSD (mean of the squared successive differences) − choose Square root of MSSD. then select a distribution of your choice. To change the method for estimating σwithin. Note Individual Distribution Identification Individual Distribution Identification Stat > Quality Tools > Individual Distribution Identification Use to evaluate the optimal distribution for your data based on the probability plots and goodness-of-fit tests prior to conducting a capability analysis study. <Box-Cox> <Options> <Results> Individual Distribution Identification − Box-Cox Transformation Stat > Quality Tools > Individual Distribution Identification > Box-Cox Use a Box-Cox transformation to transform your data to follow a normal distribution and store the transformed data for further analysis. choose one of the following: • • • 4 the average of the moving range (the default) − choose Average moving range. Distribution 1: Check to fit a distribution (default is Normal). Distribution 4: Check to fit a distribution (default is Gamma). • the average of the subgroup standard deviations − choose Sbar. To not use an unbiasing constant in the estimation. You can also use distribution identification to transform your data to follow a normal distribution using a Box-Cox transformation. To not use an unbiasing constant in the estimation. check Use moving range of length and enter a number in the box. check Use moving range of length and enter a number in the box. To not use an unbiasing constant in the estimation.Process Capability To specify methods for estimating within-subgroup and between-subgroup standard deviation 1 2 In the Capability Analysis (Between/Within). 245 . For a discussion of the relative merits of these methods. uncheck Use unbiasing constants. Distribution 3: Check to fit a distribution (default is Weibull). the median of the moving range − choose Median moving range. then select a distribution of your choice. Capability Analysis Multiple Variables (Normal) − Between/Within subgroups. then select a distribution of your choice. To change the method for estimating σbetween. Dialog box items Box-Cox power transformation (W=Y**Lambda) [for normal distribution]: Check to transform your data to follow a normal distribution. To change the length of the moving range from 2. see [1]. 3 Click OK. Use all distributions: Choose to fit your data with each of the 14 parametric distributions. Subgroups across rows of: Choose if subgroups are arranged in rows across several columns. Note You can only use this procedure with nonnegative data. 2003 Minitab Inc. Dialog box items Data are arranged as Single column: Choose if data are in one column. or Capability Sixpack (Between/Within) main dialog box. then select a distribution of your choice. then enter the columns. • the pooled standard deviation (the default) − choose Pooled standard deviation.

Lambda = 0. Click OK. 1 2 3 4 Open the worksheet TILES. Dialog box items Display one probability plot per page: Check to display one probability plot per graph window. Lambda = 0 (natural log): Choose to use zero as the lambda value. Subgroups across rows of: Choose to store data in subgroups arranged in rows across several columns. Choose Stat > Quality Tools > Individual Distribution Identification. Confidence levels: Enter a confidence level between 0 and 100. Dialog box items Descriptive statistics: Check to display descriptive statistics for the data. then enter a column.5 (square root): Choose to use the square root of 0. 1. you measured warping in 10 tiles each working day for 10 days. and are concerned about warping in the tiles. and a method for calculating the plot points using Tools > Options > Individual Graphs > Probability Plots. 0. then enter Warping. Data must be numeric.135.5 as the lambda value. use '∗' to represent missing data. . Store transformed data in Single columns: Choose to store data in one column. choose Single column. 2.Quality Control Use optimal lambda: Choose to allow Minitab to estimate the optimal lambda value. Other (enter a value between -5 and 5): Choose to enter your own lambda value. then enter the percents (0 < P < 100) individually or as a column. Percentiles for these percents: Check to estimate percentiles for selected percents. Individual Distribution Identification allows you to fit these data with 14 parametric distributions.MTW. Example of individual distribution identification Suppose you work for a company that manufactures floor tiles. Under Data are arranged as. and 5 are the default percent. Distribution parameter estimates: Check to display distribution parameters estimated from the data. You can only use a Box-Cox transformation with nonnegative data. The default is to display four probability plots per graph window. You can set preference for Y-scale type. Data − Individual Distribution Identification You can arrange the data in one column or in subgroups across several columns. The distribution of the data is unknown. Choose Use all distributions. If your subgroups are not all the same size. 0. All subgroups should be the same size. 246 2003 Minitab Inc. Do not display confidence intervals on probability plots: Check to display probability plots without confidence intervals. graph orientation. then enter a value. To ensure production quality. then enter the columns Individual Distribution Identification − Options Stat > Quality Tools > Individual Distribution Identification > Options Use to control the display of the graphical output. Goodness of fit tests: Check to display goodness-of-fit tests results for each distribution chosen in the main dialog box. Individual Distribution Identification − Results Stat > Quality Tools > Individual Distribution Identification > Results Use to control the display of the Session window output.5.

238 * 0.09064 Skewness 0.50491 3.79590 0.982 3.410 0.028 1.90969 1.225 0.684 0.60726 Minimum 0.64402 3.359 3.879 1.248 0.010 >0.26997 Threshold -1.003 <0.005 * <0.007 0.010 <0.20988 1.135236 Goodness of Fit Test Distribution Normal Lognormal 3-Parameter Lognormal Exponential 2-Parameter Exponential Weibull 3-Parameter Weibull Smallest Extreme Value Largest Extreme Value Gamma 3-Parameter Gamma Logistic Loglogistic 3-Parameter Loglogistic AD 1.78597 0.467 <0.09575 2. 247 .42168 0.99241 1.477 0.40015 0.99693 1.27904 0.12768 2.09399 * Scale: Adjusted ML estimate 2003 Minitab Inc.92307 2.27812 2.000 0.28186 Maximum 8.92307 0.692 P 0.01616 0.250 0.504 0.74444 0.523 5.24768 1.005 * LRT P 0.69368 1.013 <0.37877 Shape Scale 1.213 0.33208 1.Process Capability Session window output Distribution Identification for Warping Descriptive Statistics N 100 N* 0 Mean 2.010 0.84429 1.92307 StDev 1.547 0.41965 1.085 ML Estimates of Distribution Parameters Distribution Normal* Lognormal* 3-Parameter Lognormal Exponential 2-Parameter Exponential Weibull 3-Parameter Weibull Smallest Extreme Value Largest Extreme Value Gamma 3-Parameter Gamma Logistic Loglogistic 3-Parameter Loglogistic Location 2.763 0.78597 Median 2.30433 0.41843 2.08883 -1.34280 2.86413 2.239 0.707725 Kurtosis 0.489 0.

Quality Control Graph window output 248 2003 Minitab Inc. .

Process Capability Interpreting the results Minitab displays descriptive statistics.78597. Descriptive statistics − The table of descriptive statistics provides you with summary information for the whole column of data. goodness-of-fit test results.92307 and σ = 1. For these data. µ = 2. All the statistics are based on the non-missing (N = 100) values. and probability plots. 2003 Minitab Inc. 249 .

For a critical value α. and Distribution 4. The LRT P value of 0. Minitab displays normal probability plots for original and transformed data and their p-values for comparison. 250 2003 Minitab Inc. The middle line is the expected percentile from the distribution based on maximum likelihood parameter estimates. These distribution are labeled as SB. enter the data column in Single column.Quality Control Goodness-of-fit test − The table includes Anderson-Darling (AD) statistics and the corresponding p-value for a distribution. You can also store the transformed data for further analysis. If more than one distribution fits your data. since it provides estimates of both overall and within process capability. then click OK. see selecting a distribution for guidance. Probability plot − The probability plot includes percentile points for corresponding probabilities of an ordered data set. 0. then enter a column. where B. . The probability plots shows that the data points fall approximately on a straight line and within the confidence intervals for the 2-parameter Weibull. Johnson transformation optimally selects a function from three families of distributions of a variable. Selecting a distribution If more than one distribution fits your data: • • Select a distribution with the largest p-value. If you like. and U refer to the variable being bounded. If both normal and nonnormal models fit the data about the same. 3 4 Johnson Transformation Johnson Transformation Stat > Quality Tools > Johnson Transformation You can try to transform your data so it follows a normal distribution. and SU. Subgroups across rows of: Choose if subgroups are arranged in rows across several columns. • When subgroups are in rows. . Distribution 2. choose a distribution under Distribution 1. enter a series of columns in Subgroups across rows of. Distribution 3. which tests whether a 2-parameter distribution would fit the data equally well compared to its 3-parameter counterpart. Note Johnson transformation does not always find an optimal function to transform your data. lognormal. L.4671. choose Use all distribution. 0. − A distribution based on capability statistics. and gamma distributions fit the data well. The left and right line represent the lower and upper bounds for the confidence intervals of each percentile. SL. To fit data with four distributions of your choice. select − A distribution that you have used previously for a similar data set.25.2383 indicate that the 2-parameter Weibull. and gamma distribution. use any dialog box items. choose Specify. see Example of Capability Analysis for Nonnormal Data and Example of Capability Analysis using a Box-Cox Transformation.2246 suggests that the 3-parameter Weibull distribution does not significantly improve the fit compared to the 2-parameter Weibull distribution. and unbounded. Among extremely close p-values. Choose the one that is most conservative. and 0. Do one of the following: • When subgroups are in one column. Do one of the following: • To fit your data with all 14 distributions. 3-parameter Weibull. Dialog box items Data are arranged as Single column: Choose if data are in one column. 3-parameter Weibull. For other example using these data.2128. a p-value greater than α suggests that the data follow that distribution. To identify a distribution for a capability study 1 2 Choose Stat > Quality Tools > Individual Distribution Identification. The p-value of 0. it is probably better to choose the normal model. largest extreme value. largest extreme value. • To fit your data with the distribution of your choice. then enter the columns. which are easily transformed into a standard normal distribution. Minitab also includes a p-value for Likelihood ratio test (LRT P).

To ensure production quality. Data must be numeric. Data − Johnson Transformation You can arrange the data in one column or in subgroups across several columns. Click Options. choose Single column. then enter Warping.05. Example of a Johnson transformation Suppose you work for a company that manufactures floor tiles. You cannot transform data that is in one arrangement and store it in the other arrangement. In P-Value to select best fit. choose Single column. and are concerned about warping in the tiles. you measured warping in 10 tiles each working day for 10 days. graph orientation. Use Johnson transformation to select the best transformation. Because your data is highly right-skewed. The default is 0.Process Capability Store transformed data in Single column: Enter a storage column for the transformed data. 251 . Choose Stat > Quality Tools > Johnson Transformation. Under Store transformed data in. and a method for calculating the plot points using Tools > Options > Individual Graphs > Probability Plots. Subgroups across rows of: Enter storage columns for the transformed data. type 0. 2003 Minitab Inc. Dialog box items P-Value to select best fit: Type a p-value for selecting the best transformation. where the subgroups are arranged in rows across several columns. Click OK in each dialog box. Under Data are arranged as. <Options> Johnson Transformation − Options Stat > Quality Tools > Johnson Transformation > Options You can enter a p-value for selecting the best transformation. then enter C2. Note 1 2 3 4 5 6 You can set preference for Y-scale type. but one desires to apply a methodology that requires the underlying distribution to be normal.MTW. Open the worksheet TILES. Johnson transformation is helpful when the collected data are nonnormal.10. you want to transform it so the data follow a normal distribution.

799). You can compare the normality of the transformed data stored in C2 to the original data. enter a series of columns in Subgroups across rows of. • For subgroup in rows. P = 0. For example of Box-Cox transformation using this data. The scatterplot of P versus Z over the range of 0.25 shows that the best transformation function is selected at Z = 0. 3 4 252 2003 Minitab Inc. P = 0. The normal probability plot for the: original data indicates that the data do not follow a normal distribution (AD = 1. see Example of Capability Analysis using a Box-Cox transformation.6.25 to 1.2310. enter a column in Single column. Click OK. Do one of the following: • To store the transformed data from a single column.01).Quality Control Graph window output Interpreting the results • The normal probability plot allows you to compare the transformed data to the original data. enter the data column in Single column. enter a series of columns in Subgroups across rows of. • To store the transformed data from several columns. • • More To do a Johnson transformation 1 2 Choose Stat > Quality Tools > Johnson Transformation. using Stat > Basic Statistics > Normality Test. The table shows parameter estimates for the selected transformation function and their corresponding p-values and Zvalues.028. . Do one of the following: • For individual observations in one column. transformed data indicates that the data follow a normal distribution (AD = 0.

If you do not specify a value for the mean. Historical standard deviation (optional): Enter a value for standard deviation of the population distribution if you have a known process parameter or an estimate obtained from past data. Check Boundary to define the number as a "hard" limit. If your data are very skewed. Use optimal lambda: Choose if you want Minitab to estimate an optimal value of lambda. such as the process mean. or one of three ways if you have a subgroup size of 1. and the expected within and overall performance. 253 . Lower spec: Enter the lower specification limit. Subgroups across rows of: Choose if subgroups are arranged in rows across several columns. See <Estimate>. 2003 Minitab Inc. Check Boundary to define the number as a "hard" limit.Process Capability Capability Analysis . See Upper spec below for an explanation. The output includes histogram for the original and transformed data with their specification limits for comparison. the target (if you enter one). Other (enter a value between -5 and 5): Choose and enter your own value of lambda to bypass the lambda estimation step. Minitab can estimate standard deviation in one of three ways for sample sizes larger than 1. The report also includes statistics of the process data. Historical mean (optional): Enter a value for the mean of the population distribution if you have a known process parameter or an estimate obtained from past data. So the report can be used to visually assess whether the data are normally distributed. Subgroup size (use a constant or an ID column): Enter the subgroup size (for equal-size subgroups) or a column of subscripts (for unequal-size subgroups). it is estimated from the data. meaning that it is impossible for a measurement to fall outside the limit. the observed performance. Dialog box items Data are arranged as Single column: Choose if data is in one column.5 (square root): Choose if you want Minitab to use lambda equal to square root of 0. Lambda = 0 (natural log): Choose if you want Minitab to use lambda equal to zero. and a complete table of overall and within capability statistics. Note You can only use this procedure with nonnegative data.Box-Cox Transformation Stat > Control Charts > Capability Analysis > Normal > Box-Cox Use this option to transform nonnormal data to follow normal distribution and select a method for determining lambda. Enter a column. When you define upper and lower specification limits as boundaries. and whether it is capable of consistently meeting the process specifications. Upper spec: Enter the upper specification limit. The two normal curves are generated using the process mean and within standard deviation and the process mean and overall standard deviation.5. it is estimated from the data. Note You must enter a lower specification limit and/or an upper specification limit. Minitab sets the expected percentage of values that are out of specifications to ∗ for a boundary. If you do not specify a value for standard deviation. A model that assumes the data are from a normal distribution suits most process data. Enter the columns. and the process specifications. Dialog box items Box-Cox power transformation (W=Y**Lambda): Check to perform Box-Cox transformation. <Box-Cox> <Estimate> <Options> <Storage> Capability Analysis (Normal Distribution) . whether the process is centered on the target. The report includes a capability histogram overlaid with two normal curves. the within and overall standard deviation. see the discussion under nonnormal data.Normal Capability Analysis (Normal Distribution) Stat > Quality Tools > Capability Analysis > Normal Use Capability Analysis (Normal) to produce a process capability report when your data are from a normal distribution or to transform the data to follow normal distribution using Box-Cox transformation. Lambda = 0.

Use moving range of length: Choose to enter the number of observations used to calculate the moving range. Median moving range: Choose to estimate standard deviation using the median of the moving range. You can set preferences for the estimation of σ using Tools > Options > Control Charts and Quality Tools > Estimating Sigma. a span of 2 is used as consecutive values have the greatest chance of being alike. . Percents: Choose to display the expected and observed out-of-spec in percents. Sbar: Choose to estimate standard deviation using the average of the subgroup standard deviations. Pp): Choose to display capability statistics. displaying the target and Cpm values in a table below the graph. But when the subgroup size is 1 you cannot calculate sample standard deviations or ranges. You can also choose how the capability analysis is displayed and what statistics are calculated. Minitab estimates standard deviation using moving ranges instead.0). Calculate statistics using a___sigma tolerance: Enter a number to calculate the capability statistics using an interval other than 6 standard deviations wide (3 on either side of the process mean). Minitab then calculates Ppm in addition to the standard capability statistics. Perform Analysis Within subgroup analysis: Check to perform the within-subgroup analysis only.. Display Parts per million: Choose to display the expected and observed out-of-spec in parts per million. Dialog box items Target (adds Cpm to table): Enter a number that Minitab uses as a target. (for subgroup size > 1): Rbar: Choose to estimate standard deviation using the average of the subgroup ranges. Average moving range: Choose to estimate standard deviation using the average of the moving range (the default). (for subgroup size = 1): Minitab generally estimates standard deviation using the sample standard deviations or ranges of each subgroup. Square root of MSSD: Choose to estimate standard deviation using the mean squared successive differences. 254 2003 Minitab Inc. The default is to perform both the within-subgroup and overall analyses. Use unbiasing constants: Check to estimate standard deviation either using or not using unbiasing constants. You can set preferences for the sigma tolerance for capability statistics using Tools > Options > Control Charts and Quality Tools > Capability Analysis. and Ppk. The default is to use the unbiasing constants. Minitab then calculates Cpm. Include confidence intervals: Check to get the confidence interval for Cp. Overall analysis: Check to perform the overall analysis only. Confidence level: Enter a confidence level between 0 and 100 (default is 95. Capability Analysis (Normal Distribution) − Options Stat > Quality Tools > Capability Analysis > Normal > Options Allows you to transform the data using a Box-Cox power transformation. Confidence intervals: Choose to get two-sided confidence intervals (the default) or just an upper or lower confidence bound. or nominal specification.Quality Control Capability Analysis (Normal Distribution) − Estimation of Standard deviation Stat > Quality Tools > Capability Analysis > Normal > Estimate Controls number of observations used in estimating standard deviation. The span must be < 100. Dialog box items Methods of estimating standard deviation: See [2] for a discussion of the relative merits of the methods listed below. You can specify a process target. Pp. Capability stats (Cp. The default is to perform both the within-subgroup and overall analyses. in addition to the standard capability statistics. Benchmark Z's: Choose to display benchmark Z statistics. Title: Type a new title to replace the default title. By default. Pooled standard deviation: Choose to estimate standard deviation using a pooled standard deviation (the default). Cpk.

PPM or % Total: Check to store the PPM Total or % Total.USL: Check to store the CPU or Z. PPL or Z. Expected "Within" Performance PPM or % < LSL: Check to store the PPM < LSL or % < LSL. 255 .USL: Check to store the PPU or Z. CCpk: Check to store the CCpk statistic.Bench statistic. StDev (Within): Check to store the within standard deviation. Lower bound for Cpk: Check to store lower bound for Cpk statistic.Bench statistic.LSL: Check to store the CPL or Z. USL: Check to store the upper specification limit.LSL statistic.Bench statistic. StDev (Overall)*: Check to store the overall standard deviation of the transformed data. USL*: Check to store the transformed upper specification limit. Cpm: Check to store the Cpm statistic. Upper bound for Ppk: Check to store the upper bound for Ppk statistic. Upper bound for Cpk: Check to store upper bound for Cpk statistic. PPU or Z. Lower bound for Cpm: Check to store the lower bound for Cpm statistic. Mean*: Check to store the sample mean of the transformed data. Lower bound for Cp or Z.Bench: Check to store the Cp or Z.Bench: Check to store lower bound for Pp or Z. Ppk: Check to store the Ppk statistic.Process Capability Capability Analysis (Normal Distribution) − Storage Stat > Quality Tools > Capability Analysis > Normal > Storage Allows you to store your choice of statistics in the worksheet.USL statistic. Cpk:Check to store Cpk statistic. Target*: Check to store the transformed Target value.Bench: Check to store lower bound for Cp or Z. Pp or Z. Target: Check to store the Target value. Lower bound for Ppk: Check to store the lower bound for Ppk statistic.LSL statistic. Lower bound for Pp or Z.Bench: Check to store upper bound for Pp or Z. Cp or Z. StDev (Within)*: Check to store the within standard deviation of the transformed data. Dialog box items Variable name: Check to store the column names used in the analysis.Bench statistic. Note Transformed statistics is indicated by ∗. StDev (Overall): Check to store the overall standard deviation. LSL*: Check to store the transformed lower specification limit. Sample N: Check to store the number of sample observations (N).LSL: Check to store the PPL or Z. Observed Performance PPM or % < LSL: Check to store the PPM < LSL or % < LSL. The statistics available for storage depend on the options you have chosen in the Capability Analysis (Normal) dialog box and subdialog boxes.Bench: Check to store upper bound for Cp or Z. PPM or % Total: Check to store the PPM Total or % Total.USL statistic. CPU or Z. Upper bound for Pp or Z. PPM or % > USL: Check to store the PPM > USL or % > USL.Bench statistic. LSL: Check to store the lower specification limit. Expected "Overall" Performance 2003 Minitab Inc. CPL or Z. Upper bound for Cp or Z. PPM or % > USL: Check to store the PPM > USL or % > USL. Mean: Check to store the sample mean.Bench: Check to store the Pp or Z.Bench statistic.

must be 600 mm + 2 mm long to meet engineering specifications. There has been a chronic problem with camshaft lengths being out of specification − a problem that has caused poor-fitting assemblies down the production line and high scrap and rework rates. Data − Capability Analysis Normal You can use individual observations or data in subgroups. enter 602. Box-Cox Lambda: Check to store the lambda value used in the Box-Cox transformation. Subgroup data can be structured in one column. you must have two or more observations in at least one subgroup in order to estimate the process standard deviation. Upon examination of the inventory records.Quality Control PPM or % < LSL: Check to store the PPM < LSL or % < LSL. data must be positive. you discovered that there were two suppliers for the camshafts. enter 598. In Subgroup size. so you decided to stop accepting production runs from them until they get their production under control. 1 2 3 4 5 Open the worksheet CAMSHAFT. Example of Capability Analysis (normal probability model) Suppose you work at an automobile manufacturer in a department that assembles engines. After dropping Supplier 2.MTW. see Data. In Lower spec. the number of poor quality assemblies has dropped significantly. Choose Stat > Quality Tools > Capability Analysis > Normal. For examples. In Target (adds Ppm to table). enter Supp1. enter 600. Click OK in each dialog box. a camshaft. enter 5. Minitab omits it from the calculations. In Single column. If an observation is missing. PPM or % Total: Check to store the PPM Total or % Total. You decide to run a capability study to see whether Supplier 1 alone is capable of meeting your engineering specifications. Individual observations should be structured in one column. Click Options. An Xbar and R chart showed you that Supplier 2's camshaft production was out of control. then set up a second column of subgroup indicators. . 256 2003 Minitab Inc. One of the parts. In Upper spec. To use the Box-Cox transformation. enter the data in a single column. If you have data in subgroups. but the problems have not completely disappeared. When you have subgroups of unequal size. PPM or % > USL: Check to store the PPM > USL or % > USL. or in rows across several columns.

A histogram shows that your data do not follow a normal distribution. is 3621.Process Capability Graph window output Interpreting the results If you want to interpret the process capability statistics. and therefore. But you can see that the process mean (599.548) falls short of the target (600). the PPM < LSL − the number of parts per million whose characteristic of interest is less than the lower spec. And the left tail of the distribution falls outside the lower specification limits. The Cpk index for Supplier 1 is only 0. so you decide to use the Box-Cox power transformation to try to make the data "more normal. 257 . you will work with them to improve their process. This means that approximately 3621 out of a million camshafts do not meet the lower specification of 598 mm. To ensure production quality. as shown by the histogram overlaid with a normal curve. you measure warping in ten tiles each working day for ten days. your data should approximately follow a normal distribution.06." 2003 Minitab Inc. indicating that they need to improve their process by reducing variability and centering the process on the target. The Cpk index indicates whether the process will produce units within the tolerance limits. Example of Capability Analysis (Box-Cox transformation) Suppose you work for a company that manufactures floor tiles and are concerned about warping in the tiles. This requirement appears to have been fulfilled. your own. This means you will sometimes see camshafts that do not meet the lower specification of 598 mm. Since Supplier 1 is currently your best supplier.90. Likewise.

426207. In Single column. So you will run the Capability Analysis with a Box-Cox transformation. 258 2003 Minitab Inc.5). Graph window output The best estimate of lambda is 0. Then you will do the capability analysis. using λ = 0.5 is a reasonable choice because it falls within the 95% confidence interval. In Subgroup size. such as the square root (a lambda of 0. but practically speaking. enter Warping. type 10.5. as marked by vertical lines on the graph. . 1 2 3 Open the worksheet TILES. 1 Choose Stat > Quality Tools > Capability Analysis (Normal). Choose Stat > Control Charts > Box-Cox Transformation. performing the Box-Cox transformation with that value.Quality Control First you need to find the optimal lambda ( λ ) value for the transformation. you may want a lambda value that corresponds to an intuitive transformation. In our example. 0. Click OK.MTW.

see Example of Capability Analysis for Nonnormal Data and Example of Capability Sixpack using a Box-Cox Transformation. type 10. Check Box-Cox power transformation (W = Y**Lambda). and enter the columns containing the rows in the box. the capability statistics printed are CPU and Cpk. You must enter at least one 3 2003 Minitab Inc. Click Box-Cox. type 8. Do one of the following: • When subgroups or individual observations are in one column. For other examples using this data. Now the process capability statistics are appropriate for this data. To perform a capability analysis (normal probability model) 1 2 Choose Stat > Quality Tools > Capability Analysis > Normal. enter a lower and/or upper specification limit. enter Warping. Graph window output Interpreting the results As you can see from the normal curve overlaying the histogram.Process Capability 2 3 4 5 In Single column. Choose Lambda = 0. You can also see on the histogram that some of the process data fall beyond the upper spec limit. enter a subgroup size of 1. Because you only entered an upper specification limit. For individual observations. so your process does not appear to be capable. Both statistics are 0. Click OK in each dialog box. respectively. choose Subgroups across rows of. • When subgroups are in rows. see example of individual distribution identification. Note To choose the best fitting distribution using the same data.76. In Lower spec or Upper spec. below the guideline of 1. the Box-Cox transformation "normalized" the data. In Subgroup size. In Upper spec. In Subgroup size.33. enter a subgroup size or column of subgroup indicators.5 (square root). 259 . enter the data column in Single column.

When you collect data in subgroups. Note You can only use this procedure with nonnegative data. See Upper spec below for an explanation. meaning that it is impossible for a measurement to fall outside the limit. Subgroups across rows of: Choose if subgroups are arranged in rows across several columns. and a complete table of overall and between/within capability statistics. . If you do not specify a value for the mean. or you may specify historical standard deviations. Enter a column. These will be combined (pooled) to compute the between/within standard deviation. If you do not specify a value for sigma. Choose Boundary to define the number as a "hard" limit. use any of the dialog box options. Enter the columns Lower spec: Enter the lower specification limit. When you define upper and lower specification limits as boundaries. There may also be random error between subgroups. Historical standard deviations: Within subgroup: (optional): Enter a value for the within-subgroup standard deviation of the population distribution if you have a known process parameter or an estimate obtained from past data. Choose Boundary to define the number as a "hard" limit. such as the process mean. Capability Analysis . 260 2003 Minitab Inc. Subgroup size (use a constant or an ID column): Enter the subgroup size (for equal-size subgroups) or a column of subscripts (for unequal-size subgroups). Between subgroups: (optional): Enter a value for the between-subgroup standard deviation of the population distribution if you have a known process parameter or an estimate obtained from past data. Minitab can estimate within-subgroup sigma in one of three ways. then click OK. If you do not specify a value for sigma. The normal curves are generated using the process mean and overall standard deviation and the process mean and between/within standard deviation.Quality Control of the specification limits. the between/within process variation is due to both the between-subgroup variation and the within-subgroup variation. The report also includes statistics of the process data. and observed and expected performance. The output includes histogram for the original and transformed data with their specification limits for comparison. The report includes a capability histogram overlaid with two normal curves. such as Cp and Cpk. it is estimated from the data. if you enter one. Minitab sets the expected percentage of values that are out of specifications to ∗ for a boundary. Under these conditions. Upper spec: Enter the upper specification limit. it is estimated from the data. random error within subgroups may not be the only source of variation to consider. The between/within standard deviation will be used to calculate the capability statistics. Historical mean (optional): Enter a value for the mean of the population distribution if you have a known process parameter or an estimate obtained from past data. Dialog box items Data are arranged as Single column: Choose if data is in one column. 4 If you like. between/within and overall standard deviation. Minitab can estimate between-subgroup sigma in one of three ways <Box-Cox> <Estimate> <Options> <Storage> Capability Analysis (Between/Within) . Note You must enter at least an upper or lower spec. it is estimated from the data.Between/Within Capability Analysis (Between/Within) Stat > Quality Tools > Capability Analysis > Between/Within Use Capability Analysis (Between/Within) to produce a process capability report using both between-subgroup and withinsubgroup variation. Capability Analysis (Between/Within) computes standard deviations within-subgroups and between-subgroups.Box-Cox Stat > Control Charts > Capability Analysis > Between/Within > Box-Cox Use this option to transform nonnormal data to follow a normal distribution and select a method for determining lambda. target. Dialog box items Box-Cox power transformation (W=Y**Lambda): Check to perform Box-Cox transformation.

Between subgroups Average moving range: Choose to estimate between-subgroups standard deviation using the average of the moving range (the default). Benchmark Z's: Choose to display the benchmark Z statistics. Dialog box items Methods of estimating standard deviation: See [1] for a discussion of the relative merits of the methods listed below. Use moving range of length: Choose to enter the number of observations used to calculate the moving range. You can set preferences for the sigma tolerance for capability statistics using Tools > Options > Control Charts and Quality Tools > Capability analysis. Square root of MSSD: Choose to estimate between-subgroups standard deviation using the mean squared successive differences. The default is to perform both the between/withinsubgroup and overall analyses. Dialog box items Target (adds Cpm to table): Enter a number that Minitab uses as a target.5.Process Capability Use optimal Lambda: Choose if you want Minitab to estimate an optimal value of lambda. Capability stats (Cp. Use unbiasing constants: Choose to estimate σ either using or not using unbiasing constants. a span of 2 is used as consecutive values have the greatest chance of being alike. Capability Analysis (Between/Within) − Estimation of Standard Deviation Stat > Quality Tools > Capability Analysis > Between/Within > Estimate Use to control the method for estimating within and between standard deviation. By default. Median moving range: Choose to estimate between-subgroups standard deviation using the median of the moving range. Other (enter a value between -5 and 5): Choose to enter your own value of lambda to bypass the lambda estimation step. Pp): Choose to display the capability statistics. Lambda = 0. or nominal specification. Minitab then calculates Cpm. Overall analysis: Check to perform the overall analysis only.5 (square root): Choose if you want Minitab to use lambda equal to square root of 0. You can also choose how the capability analysis is displayed and what statistics are calculated. Percents: Choose to display the expected and observed out-of-spec in percents. The default is to perform both the between/within-subgroup and overall analyses. Pooled standard deviation: Choose to estimate within-subgroup standard deviation using a pooled standard deviation (the default). in addition to the standard capability statistics. Capability Analysis (Between/Within) − Options Stat > Quality Tools > Capability Analysis > Between/Within > Options Allows you to transform the data using a Box-Cox power transformation. Within subgroup Rbar: Choose to estimate within-subgroup standard deviation using the average of the subgroup ranges. Minitab then calculates Cpm in addition to the standard capability statistics. 2003 Minitab Inc. Enter a value. 261 . The span must be < 100. Sbar: Choose to estimate within-subgroup standard deviation using the average of the subgroup standard deviations. Lambda = 0 (natural log): Choose if you want Minitab to use lambda equal to zero. The default is to use the unbiasing constants. Calculate statistics using: (sigma tolerance): Enter a number to calculate the capability statistics using an interval other than 6 standard deviations wide (3 on either side of the process mean). Display Parts per million: Choose to display the expected and observed out-of-spec in parts per million. You can specify a process target. displaying the target and Cpm values in a table below the graph. Perform Analysis Between/within analysis: Check to perform the between/within-subgroup analysis only. You can set preferences for the estimation of σ using Tools > Options > Control Charts and Quality Tools > Estimating Standard Deviation.

Quality Control

Include confidence intervals: Check to get the confidence interval for Cp, Cpk, Pp, and Ppk. Confidence level: Enter a confidence level between 0 and 100 (default is 95.0). Confidence intervals: Choose to get two-sided confidence intervals (the default) or just an upper or lower confidence bound. Title: Type a new title to replace the default title.

**Capability Analysis (Between/Within) − Storage
**

Stat > Quality Tools > Capability Analysis > Between/Within > Storage Allows you to store you choice of statistics in the worksheet. The statistics available for storage depend on the options you have chosen in the Capability Analysis (Between/Within) dialog box and subdialog boxes. Note Transformed statistics is indicated by ∗.

**Dialog box items
**

Variable name: Check to store the variable name used in the analysis. LSL: Check to store the lower specification limit. LSL*: Check to store the transformed lower specification limit. Target: Check to store the target value. Target*: Check to store the transformed target value. USL: Check to store the upper specification limit. USL*: Check to store the transformed upper specification limit. Mean: Check to store the sample mean. Mean*: Check to store the sample mean of the transformed data. Sample N: Check to store the number of sample observations (N). StDev (Between): Check to store the between standard deviation. StDev (Within): Check to store the within standard deviation. StDev (B/W): Check to store the between/within standard deviation. StDev (Between)*: Check to store the between standard deviation of the transformed data. StDev (Within)*: Check to store the within standard deviation of the transformed data. StDev (B/W)*: Check to store the between/within standard deviation of the transformed data. StDev (Overall): Check to store the overall standard deviation. StDev (Overall)*: Check to store the overall standard deviation of the transformed data. Cp or Z.Bench: Check to store the Cp or the Z.Bench statistic. Lower bound for Cp or Z.Bench: Check to store the lower bound for Cp or Z.Bench statistic. Upper bound for Cp or Z.Bench: Check to store the upper bound for Cp or Z.Bench statistic. CPU or Z.USL: Check to store the CPU or Z.USL statistic. CPL or Z.LSL: Check to store the CPL or Z.LSL statistic. Cpk: Check to store the Cpk statistic. Lower bound for Cpk: Check to store the lower bound for Cpk statistic. Upper bound for Cpk: Check to store the upper bound for Cpk statistic. CCpk: Check to store the CCpk statistic. Cpm: Check to store the Cpm statistic. Lower bound for Cpm: Check to store the lower bound for Cpm statistic. Pp or Z.Bench: Check to store the Pp or Z.Bench statistic. Lower bound for Pp or Z.Bench: Check to store the lower bound for Pp or Z.Bench statistic. Upper bound for Pp or Z.Bench: Check to store the upper bound for Pp or Z.Bench statistic. PPU or Z.USL: Check to store the PPU or Z.USL statistic. PPL or Z.LSL: Check to store the PPL or Z.LSL statistic. Ppk: Check to store the Ppk statistic. Lower bound for Ppk: Check to store the lower bound for Ppk statistic. Upper bound for Ppk: Check to store the upper bound for Ppk statistic.

262

2003 Minitab Inc.

Process Capability

Observed Performance PPM or % < LSL: Check to store PPM < LSL or % < LSL. PPM or % > USL: Check to store PPM > USL or % > USL. PPM or % Total: Check to store PPM Total or % Total. Expected "Between/Within' Performance PPM or % < LSL: Check to store PPM < LSL or % < LSL. PPM or % > USL: Check to store PPM > USL or % > USL. PPM or % Total: Check to store PPM Total or % Total. Expected"Overall" Performance PPM or % < LSL: Check to store PPM < LSL or % < LSL. PPM or % > USL: Check to store PPM > USL or % > USL. PPM or % Total: Check to store PPM Total or % Total. Box-Cox Lambda: Check to store the lambda value used in the Box-Cox transformation.

**Data for Capability Analysis (Between/Within)
**

You can use data in subgroups, with two or more observations. Subgroup data can be structured in one column, or in rows across several columns. To use the Box-Cox transformation, data must be positive. Ideally, all subgroups should be the same size. If your subgroups are not all the same size, due to missing data or unequal subgroup sizes, only subgroups of the majority size are used for estimating the between-subgroup variation.

**Example Capability Analysis (Between/Within)
**

Suppose you are interested in the capability of a process that coats rolls of paper with a thin film. You are concerned that the paper is being coated with the correct thickness of film and that the coating is applied evenly throughout the roll. You take three samples from 25 consecutive rolls and measure coating thickness. The thickness must be 50 +3 mm to meet engineering specifications. 1 2 3 4 Open the worksheet BWCAPA.MTW. Choose Stat > Quality Tools > Capability Analysis > Between/Within. In Single column, enter Coating. In Subgroup size, enter Roll. In Lower spec, enter 47. In Upper spec, enter 53. Click OK

2003 Minitab Inc.

263

Quality Control Graph window output

**Interpreting the results
**

You can see that the process mean (49.8829) falls close to the target of 50. The Cpk index indicates whether the process will produce units within the tolerance limits. The Cpk index is only 1.21, indicating that the process is fairly capable, but could be improved. The PPM Total for Expected "Between/Within" Performance is 193.94. This means that approximately 194 out of a million will not meet the specification limits. This analysis tells you that your process is fairly capable.

**To perform a capability analysis (between/within)
**

1 2 Choose Stat > Quality Tools > Capability Analysis > Between/Within. Do one of the following: • When subgroups are in one column, enter the data column in Single column. In Subgroup size, enter a subgroup size or column of subgroup indicators. • When subgroups are in rows, choose Subgroups across rows of, and enter the columns containing the rows in the box. In Lower spec or Upper spec, enter a lower and/or upper specification limit, respectively. You must enter at least one of them. If you like, use any of the available dialog box options, then click OK.

3 4

**Capability Analysis - Nonnormal
**

Capability Analysis (Nonnormal Distribution)

Stat > Quality Tools > Capability Analysis > Nonnormal Use to produce a process capability report when your data do not follow a normal distribution. The report includes a capability histogram overlaid with a distribution curve, and a table of overall capability statistics. The report also includes statistics of the process data, such as the mean, distribution parameter estimates, target (if you enter one), and process specifications; the actual overall capability (Pp, Ppk, PPU, and PPL), and the observed and expected overall performance. Minitab bases the calculations on maximum likelihood estimates of the distribution parameters, rather than mean and

264

2003 Minitab Inc.

Process Capability

variance estimates as in the normal case. You can use the report to visually assess the distribution of the process relative to the target, whether the data follow a specified distribution, and whether the process is capable of meeting the specifications consistently.

**Dialog box items
**

Data are arranged as Single column: Choose if data are in one column. Enter a column. Subgroups across rows of: Choose if subgroups are arranged in rows across several columns. Enter the columns. Fit data with Distribution: Choose a distribution. Johnson transformation: Choose to transform your data to follow normal distribution using a Johnson transformation method. Lower spec: Enter a lower specification limit. Check Boundary to define the number as a "hard" limit, meaning that it is impossible for a measurement to fall outside the limit. Upper spec: Enter an upper specification limit. Check Boundary to define the number as a "hard" limit, meaning that it is impossible for a measurement to fall outside the limit. Note You must enter a lower specification limit and/or an upper specification limit. When you define upper and lower specification limits as boundaries, Minitab sets the expected percentage of values that are out of specifications to ∗ for a boundary. <Estimate> <Options> <Storage>

**Capability Analysis (Nonnormal Distribution) − Estimate
**

Stat > Quality Tools > Capability Analysis (Nonnormal) > choose Distribution > Estimate Distributions Weibull, Gamma, Exponential, Lognormal, Loglogistic, Smallest extreme value, Largest extreme value, 2parameter exponential, 3-parameter Weibull, 3-parameter gamma, 3-parameter lognormal, and 3parameter loglogistic Use to control the method for determining distribution parameters.

**Dialog box items
**

Estimate parameters of distributions: Choose to estimate the distribution parameters from the data. Minitab estimates any parameters not specified below. Set shape (Weibull/gamma) or scale (other dists): Enter the shape or scale parameter depending on your choice of distribution. Set threshold (3-parameter): Enter the threshold parameter. Use historical estimates (Use the order: Location/Shape, Scale, Threshold): Choose to specify historical estimates of parameters. Enter constants or a column. The number of constants and rows in the column must equal the number of parameters in the distribution. Tip For 3-parameter loglogistic distribution, enter values of the parameters in the following order: location, scale, and threshold.

**Capability Analysis (Nonnormal Distribution) − Estimate
**

Stat > Quality Tools > Capability Analysis > Nonnormal > choose Johnson transformation > Estimate You can enter a p-value for selecting the best transformation.

**Dialog box items
**

P-Value to select best fit: Type a p-value.

**Capability Analysis (Nonnormal Distribution) − Options
**

Stat > Quality Tools > Capability Analysis > Nonnormal > Options Use to specify process target or confidence level for the confidence intervals of Cp, Cpk, Pp, Ppk, and Cpm. Enter a new title to replace the default title.

2003 Minitab Inc.

265

Quality Control

You can set preferences for the sigma tolerance for capability statistics using Tools > Options > Control Charts and Quality Tools > Capability Analysis.

**Dialog box items
**

Target: Enter the process target value. Calculate statistics using ____ sigma tolerance: Enter a number (default is 6.0) to calculate the capability statistics using an interval other than 6 standard deviations wide (3 on each side of the process mean). Display Capability stats (Pp): Choose to display capability statistics (the default). Benchmark Z's (sigma level): Choose to display Benchmark Z statistics. Include confidence intervals: Check to display confidence intervals for Cp, Cpk, Pp, Ppk, and Cpm. Confidence level: Enter a confidence level between 0 and 100 (default is 95.0). Confidence intervals: Choose to get two-sided confidence intervals (the default) or just an upper or lower confidence bound. Title: Type a new title to replace the default title.

**Capability Analysis (Nonnormal Distribution) − Storage
**

Stat > Quality Tools > Capability Analysis (Nonnormal) > choose Distribution > Storage Distributions Weibull, Gamma, Exponential, Lognormal, Loglogistic, Smallest extreme value, Largest extreme value, 2-parameter exponential, 3-parameter Weibull, 3-parameter gamma, 3-parameter lognormal, and 3parameter loglogistic

Store your choice of statistics in the worksheet. The statistics available for storage depend on the options you have chosen in the Capability Analysis (Nonnormal) dialog box and its subdialog boxes.

**Dialog box items
**

Variable name: Check to store the variable name in the worksheet. LSL: Check to store the lower specification limit Target: Check to store target in the worksheet. USL: Check to store the upper specification limit. Mean: Check to store the sample mean. Sample N: Check to store the number of sample observations. Location: Check to store the location parameter of a distribution. Shape: Check to store the shape parameter of a distribution. Scale: Check to store the scale parameter of a distribution. Threshold: Check to store the threshold parameter of a distribution. Pp: Check to store the Pp or Z.Bench statistic. PPU: Check to store the PPU or Z.USL statistic. PPL: Check to store the PPL or Z.LSL statistic. Ppk: Check to store the Ppk. Observed Performance PPM < LSL: Check to store PPM < LSL. PPM > USL: Check to store PPM > USL. PPM Total: Check to store PPM Total. Expected Performance PPM < LSL: Check to store PPM < LSL. PPM > USL: Check to store PPM > USL. PPM Total: Check to store PPM Total.

**Capability Analysis (Nonnormal Distribution) − Storage
**

Stat > Quality Tools > Capability Analysis (Nonnormal) > choose Distribution > Storage You can store your choice of statistics in the worksheet. The statistics available for storage depend on the options you chose in the Capability Analysis (Nonnormal) dialog and subdialog boxes.

266

2003 Minitab Inc.

Process Capability

Note

Transformed statistics is indicated by ∗.

**Dialog box items
**

Variable name: Check to store the variable name in the worksheet. LSL: Check to store the lower specification limit. LSL*: Check to store the transformed lower specification limit. Target: Check to store the target value. Target*: Check to store the transformed target value. USL: Check to store the upper specification limit. USL*: Check to store the transformed upper specification limit. Mean: Check to store the sample mean. Mean*: Check to store the sample mean of the transformed data. Sample N: Check to store the number of observations (N). StDev (Overall): Check to store the overall standard deviation. StDev (Overall)*: Check to store the transformed overall standard deviation. Pp or Z.Bench: Check to store the Pp or Z.Bench statistic. Lower bound for Pp or Z.Bench: Check to store the lower boundary for the Pp or Z.Bench statistic. Upper bound for Pp or Z.Bench: Check to store the upper boundary for the Pp or Z.Bench statistic. PPU or Z.USL: Check to store the PPU or Z.USL statistic. PPL or Z.LSL: Check to store the PPL or Z.LSL statistic. Ppk: Check to store the Ppk statistic. Lower bound for Ppk: Check to store the lower boundary for the Ppk statistic. Upper bound for Ppk: Check to store the upper boundary for the Ppk statistic. Johnson Transformation Distribution Type: Check to store the distribution type. First shape: Check to store the first shape parameter. Second shape: Check to store the second shape parameter. Location: Check to store the location parameter. Scale: Check to store the scale parameter. Observed Performance PPM < LSL: Check to store the PPM < LSL. PPM > USL: Check to store the PPM > USL. PPM Total: Check to store the PPM Total. Expected Performance PPM < LSL: Check to store the PPM < LSL. PPM > USL: Check to store the PPM > USL. PPM Total: Check to store the PPM Total. Note The definitions for the expected performance statistics when using Johnson transformation is the same as the expected overall performance statistics for capability analysis based on a normal distribution.

**Data − Capability Analysis (Nonnormal Distributions)
**

You can enter your data in a single column or in multiple columns if you have arranged subgroups across rows. Because the nonnormal capability analysis does not calculate within capability statistics, Minitab does not use subgroups in calculations. For examples, see Data. If an observation is missing, Minitab omits it from the calculations.

**Example of capability analysis with nonnormal data
**

Suppose you work for a company that manufactures floor tiles, and are concerned about warping in the tiles. To ensure production quality, you measured warping in 10 tiles each working day for 10 days.

2003 Minitab Inc.

267

000 tiles warp more than the upper specification limit of 8mm. below the guideline of 1. Thus. Under Fit data with.73.00. so you decide to perform a capability analysis based on a Weibull probability model. . see Example of a capability analysis with a Box-Cox transformation. choose Single column. Likewise. Graph window output Interpreting the results The histogram does not show evidence of any serious discrepancies between the assumed model and the data. To analyze the same data with Capability Analysis (Normal). Both indices are 0.000 out of 1. choose one of the following: • If data are in one column. Click OK. Note 1 2 3 4 5 You can set preference for a method to calculate the capability statistics using Tools > Options > Control Charts and Quality Tools > Capability Analysis. This value means that 20.MTW. the observed PPM > USL (number of parts per million above the upper specification limit) is 20000. Under Data are arranged as. To perform a capability analysis for nonnormal data 1 2 Choose Stat > Quality Tools > Capability Analysis > Nonnormal. Open the worksheet TILES. But. Under Data are arranged as. your process does not appear to be capable. you can see that the right tail of the distribution falls over the upper specification limit. The Ppk and PPU indices tell you whether the process produces tiles within the tolerance limits. This means you will sometimes see warping higher than the upper specification of 8mm. then enter Warping.Quality Control A histogram showed that the data do not come from a normal distribution.33. Choose Stat > Quality Tools > Capability Analysis > Nonnormal. choose Single column and enter the column containing the data. choose Distribution. type 8. then select Weibull. In Upper spec. also see example of individual distribution identification.000. 268 2003 Minitab Inc. To choose the best fitting distribution for the same data.

or a column. otherwise. Minitab sets the expected percentage of values that are out of specifications to ∗ for a boundary. which means that measurements do not fall outside the limit. Boundary: Check to define the lower specification limit as a hard limit. If you check By variable. For example. enter ∗. multiple values. Subgroup sizes: Enter one value.Process Capability • 3 If subgroups are in rows. then perform capability analysis on the transformed data. otherwise. One value applies to all variables. and enter the columns. You can enter one or both specification limits. multiple values. One value applies to all variables. or a column to specify the lower specification limit for multiple variables or groups. Variation Within subgroups: Choose to perform analysis for within subgroups variation (default). Upper spec: Enter one value. you want to compare the width of parts from the same process before and after a process improvement. Under Fit data with. you can enter only one column in Variables. 269 . In Lower spec and Upper spec. One value applies to all variables. By variables: Check to perform a separate capability analysis for each unique value in the column or combination of values in columns. and enter the columns containing the data. then click OK. <Historical> <Box-Cox> <Estimate> <Graphs> <Options> <Storage> 2003 Minitab Inc. To define a lower specification limit as a hard limit enter 1. Enter one value. Multiple values or columns correspond to columns listed in Variables. If you like. multiple values. enter lower and upper specification limits. choose Distribution and select a distribution. or both. If you have no lower specification limit. Lower spec: Enter one value. multiple values. or columns to specify the subgroup size for multiple variables or groups. When you define the upper and lower specification limits as boundaries. enter 0. You can also use this feature to correct nonnormality in your data using a Box-Cox transformation.Normal Capability Analysis for Multiple Variables (Normal Distribution) Stat > Quality Tools > Capability Analysis > Multiple Variables (Normal) You can assess the capability of an in-control process when each of the continuous variables follow a normal distribution. Dialog box items Variables: Enter the columns containing the data for multiple variables. 4 5 Capability Analysis for Multiple Variables . enter ∗. To define a lower specification limit as a hard limit enter 1. Multiple values correspond to columns listed in Variables or groups in By variables. do one of the following: • To perform capability analysis using a nonnormal distribution. Multiple values correspond to columns listed in Variables or groups listed in By variables. or a column. • To transform your data to follow a normal distribution. Multiple values correspond to columns listed in Variables or groups in By variables. Between/Within subgroups: Choose to perform analysis for between and within subgroups variation. choose Johnson transformation. Use this procedure when: • • Comparing the results for multiple variables A grouping variable is associated with a continuous variable. which means that measurements do not fall outside the limit. or a column to specify the upper specification limit for multiple variables or groups. Boundary: Check to define the lower specification limit as a hard limit. choose Subgroups across rows of. One value or a column applies to all variables or groups. One value applies to all variables. If you have no lower specification limit. use any dialog box options. enter 0. Multiple values correspond to columns listed in Variables or groups in By variables. Enter one value. multiple values. Note You must enter a lower or upper specification limit.

(for subgroup size = 1) Average moving range: Choose to estimate standard deviation using the average of the moving range. One value applies to all variables. One value applies to all variables. Capability Analysis for Multiple Variables (Normal Distribution) − Estimation of Standard Deviation Stat > Quality Tools > Capability Analysis > Multiple Variables (Normal) > Choose Within subgroups > Estimate Use to control the method to estimate standard deviation. . The maximum subgroup size allowed is 100. Note Minitab estimates the mean and standard deviation from the data for missing values. Historical standard deviations Within subgroups: Enter one value. Dialog box items Historical means: Enter one value. Square root of MSSD: Choose to estimate standard deviation using the mean of the squared successive differences. or a column to specify the within-subgroups standard deviation of the population distribution. multiple values. A default value of 2 is used as consecutive values have the greatest chance of being alike. Multiple values correspond to columns listed in Variables. 270 2003 Minitab Inc. multiple values. One value applies to all variables. Dialog box items Box-Cox power transformation (W=Y**Lambda): Check to perform a Box-Cox transformation. You can set preferences for the estimation of σ using Tools > Options > Control Charts and Quality Tools > Estimating Sigma. Note For subgroup of size one.Quality Control Capability Analysis for Multiple Variables (Normal Distribution) − Historical Stat > Quality Tools > Capability Analysis > Multiple Variables (Normal) > Historical Use to the specify historical mean and standard deviation of the population for each variable. or a column. or a column to specify the mean of the population distribution. See <Estimate> for details on estimation alternatives. One value applies to all variables. Multiple values correspond to columns listed in Variables. Multiple values correspond to columns listed in Variables. you cannot calculate sample standard deviation and range. Capability Analysis for Multiple Variables (Normal Distribution) − Box-Cox Stat > Quality Tools > Capability Analysis > Multiple Variable (Normal) > Box-Cox Use to transform nonnormal data to follow a normal distribution and select a method for determining lambda. Minitab estimates standard deviation using moving ranges. Enter one value. Use optimal lambdas: Choose to estimate an optimal value of lambda. Median moving range: Choose to estimate standard deviation using the median of the moving range. Maximum allowed length is 100. Multiple values correspond to columns in Variables. multiple values. multiple values. The output includes capability analysis and histograms for the original and transformed data for comparison. or a column to specify the between-subgroup standard deviation of the population distribution. Sbar: Choose to estimate standard deviation using the average of the subgroup standard deviations. Pooled standard deviation: Choose to estimate standard deviation using a pooled standard deviation. Enter lambdas between -5 and 5 [Enter * to estimate lambda from data]: Choose to enter your own value of lambda and bypass the lambda estimation step. See <Estimate> for details on estimation alternatives. Between subgroups: Enter one value. Note You can only use this procedure with nonnegative data. Use moving range of Length: Check to change the number of observations used to calculate the moving range. (for subgroup size > 1) Rbar: Choose to estimate standard deviation using the average of the subgroup ranges. Dialog box items Methods of estimating standard deviation Use unbiasing constants: Check to estimate standard deviation using unbiasing constants.

Minitab uses a default value of 2 because consecutive values have the greatest chance of being alike. Dialog box items Methods of estimating standard deviation Use unbiasing constants: Check to estimate standard deviation using unbiasing constants. If you do not enter a minimum value. You can also specify the range to be plotted on the histogram. You can enter multiple values individually or in a column. Minitab then calculates Cpm in addition to the standard capability statistics. You can set preferences for the estimation of σ using Tools > Options > Control Charts and Quality Tools > Estimating Sigma. Between subgroups Average moving range: Choose to estimate standard deviation using the average of the moving range.Options Stat > Quality Tools > Capability Analysis > Multiple Variables (Normal) > Options Use to specify a process target. or overall analysis. Maximum subgroup size allowed is 100. Minitab determines the value from the data. Minitab estimates standard deviation using moving ranges. Minimum: Enter a minimum value to display on the x-axis of the histogram of all variables. Sbar: Choose to estimate standard deviation using the average of the subgroup standard deviations. 271 . Maximum allowed length is 100. Note For subgroup size of 1. you cannot calculate sample standard deviation and range. Square root of MSSD: Choose to estimate standard deviation using the mean squared successive differences. Capability Analysis for Multiple Variables (Normal Distrbution) . Same x-axis scalings: Choose to display a histogram of all variables with same x-axis scale.Process Capability Capability Analysis for Multiple Variables (Normal Distribution) − Estimation of Standard Deviation Stat > Quality Tools > Capability Analysis > Multiple Variables (Normal) > choose Between/Within subgroups > Estimate Use to control the method for estimating standard deviation. Pooled standard deviation: Choose to estimate standard deviation using a pooled standard deviation. You can also choose to include confidence intervals for capability indices. Minimum: Enter a minimum value for each variable. 2003 Minitab Inc. between/within subgroup. Use moving range of length: Check to change the number of observations used to calculate the moving range. Within subgroup Rbar: Choose to estimate standard deviation using the average of the subgroup ranges. Different x-axis scalings: Choose to display a histogram of each variable with different x-axis scale. Minitab determines the value from the data. Capability Analysis for Multiple Variables (Normal Distribution) − Graphs Stat > Quality Tools > Capability Analysis > Multiple Variables (Normal) > Graphs Use to display probability plots and histograms with overlaid density functions for each variable. You can set preferences for the sigma tolerance for capability statistics using Tools > Options > Control Charts and Quality Tools > Capability Analysis. You can choose to perform just the within subgroup. Normal probability plot: Check to display normal probability plots for each variable. Dialog box items Histograms with density function overlaid: Check to display histograms with overlaid density function for each variable. You can enter multiple values individually or in a column. If you do not enter a maximum value. Median moving range: Choose to estimate standard deviation using the median of the moving range. Maximum: Enter a maximum value for each variable. Maximum: Enter a maximum value to display on the x-axis of the histogram of all variables.

The statistics available for storage depend on the options you have chosen in the Capability Analysis for Multiple Variables (Normal) dialog and subdialog boxes.Bench statistic. Sample N: Check to store the number of sample observations (N). StDev (Within)*: Check to store the within standard deviation of the transformed data.Bench statistic.Bench: Check to store the upper bound for Cp or Z. Upper bound for Cpk: Check to store the upper bound for Cpk statistic CCpk: Check to store the CCpk statistic. Calculate Capability stats (Cp.Quality Control Dialog box items Targets (Display Cpm): Enter one value. Percents: Choose to display the expected and observed out-of-spec in percents. Calculate statistics using ___ sigma tolerance: Enter one value. Cpk. Lower bound for Cp or Z. 272 2003 Minitab Inc. Capability Analysis for Multiple Variables (Normal Distribution) − Storage Stat > Quality Tools > Capability Analysis > Multiple Variables (Normal) > Choose Within subgroups > Storage Use to store your choice of statistics in the worksheet. Pp. LSL*: Check to store the transformed lower specification limit.LSL statistic. StDev (Overall): Check to store the overall standard deviation. CPU or Z.LSL: Check to store the CPL or Z. One value applies to all variables. Mean: Check to store the sample mean.0). Parts per million: Choose to display the expected and observed out-of-spec in parts per million.USL statistic. USL: Check to store the upper specification limit. LSL: Check to store the lower specification limit. Confidence level: Enter a confidence level between 0 and 100 (default is 95. Mean*: Check to store the sample mean of the transformed data. Multiple values correspond to columns listed in Variables. or a column that Minitab uses as a target. USL*: Check to store the transformed upper specification limit. Pp): Choose to display capability statistics. Upper bound for Cp or Z. Target: Check to store Target. in addition to the standard capability statistics. Include confidence intervals: Check to include confidence intervals of Cp. Cp or Z. StDev (Within): Check to store the within standard deviation. Target*: Check to store transformed Target. Perform Within or between/within subgroup analysis: Check to perform within or between/within analysis. Overall analysis: Check to perform overall analysis. multiple values or a column that Minitab uses to calculate capability statistics using an interval other than 6 standard deviations wide (3 on each side of the process mean). . Cpm: Check to store the Cpm statistic.Bench: Check to store the Cp or Z. Lower bound for Cpk: Check to store the lower bound for Cpk statistic. Note Transformed statistics is indicated by ∗. StDev (Overall)*: Check to store the overall standard deviation transformed data. Benchmark Z's (sigma level): Choose to display benchmark Z statistics. multiple values. Dialog box items Variable name: Check to store the column names or BY levels in a column. Cpk: Check to store the Cpk statistic. Minitab then calculates Cpm.USL: Check to store the CPU or Z. and Ppk. Confidence intervals: Choose to get two-sided confidence intervals (the default) or just an upper or lower confidence bound.Bench: Check to store the lower bound for Cp or Z.Bench statistic. CPL or Z.

273 . Capability Analysis for Multiple Variables (Normal Distribution) − Storage Stat > Quality Tools > Capability Analysis > Multiple Variables (Normal) > Choose Between/Within subgroups > Storage Use to store your choice of statistics in the worksheet.Process Capability Lower bound for Cpm: Check to store the lower bound for Cpm statistic Pp or Z. Cp or Z. Expected "Within" Performance PPM or % < LSL: Check to store PPM < LSL or % < LSL. Target*: Check to store the transformed target. StDev (Within): Check to store the within standard deviation. PPM or % Total: Check to store PPM Total or % Total. Note Transformed statistics is indicated by ∗. Lower bound for Pp or Z. StDev (B/W): Check to store the between/within standard deviation.Bench: Check to store the Cp or Z. Expected "Overall" Performance PPM or % < LSL: Check to store PPM < LSL or % < LSL. PPM or % Total: Check to store PPM Total or % Total. Box-Cox Lambda: Check to store the lambda value used in the Box-Cox transformation.USL: Check to store the PPU or Z.LSL statistic.USL statistic.Bench: Check to store the upper bound for Pp or Z. StDev (Overall)*: Check to store the overall standard deviation of the transformed data. StDev (Within)*: Check to store the within standard deviation of the transformed data. Lower bound for Ppk: Check to store the lower bound for Ppk statistic. PPM or % Total: Check to store PPM Total or % Total. Upper bound for Ppk: Check to store the upper bound for Ppk statistics. Mean*: Check to store the sample mean of the transformed data. StDev (Between): Check to store the between standard deviation. PPU or Z. Ppk: Check to store the Ppk statistic.Bench: Check to store the lower bound for Pp or Z.Bench statistic.Bench: Check to store the Pp or Z. LSL*: Check to store the transformed lower specification limit. StDev (Between)*: Check to store the between standard deviation of the transformed data.Bench statistic. StDev (B/W)*: Check to store the between/within standard deviation of the transformed data. 2003 Minitab Inc. The statistics available for storage depend on the options you have chosen in the Capability Analysis for Multiple Variable (Normal) dialog and subdialog boxes. PPM or % > USL: Check to store PPM > USL or % > USL. PPL or Z.Bench statistic. LSL: Check to store the lower specification limit. Dialog box items Variable name: Check to store the column names and BY levels in a column. PPM or % > USL: Check to store PPM > USL or % > USL. Observed Performance PPM or % < LSL: Check to store PPM < LSL or % < LSL. USL: Check to store the upper specification limit. Target: Check to store target.LSL: Check to store the PPL or Z.Bench statistic. Sample N: Check to store the number of observations (N). USL*: Check to store the transformed upper specification limit. PPM or % > USL: Check to store PPM > USL or % > USL. StDev (Overall): Check to store the overall standard deviation. Mean: Check to store the sample mean. Upper bound for Pp or Z.

CPU or Z. Box-Cox Lambda: Check to store the lambda value used in the Box-Cox transformation. Observed Performance PPM or % < LSL: Check to store PPM < LSL or % < LSL . If the subgroup sizes of the variables differ.Bench: Check to store the lower bound for Pp or Z.99 27.01 Subgroup 1 1 1 1 2 2 3 3 3 Variable 2 0.Bench: Check to store the upper bound for Pp or Z. In the example below.59 27.LSL: Check to store the CPL or Z. then create a column of subgroup indicators for each variable.USL: Check to store the PPU or Z. Upper bound for Cpk: Check to store the upper bound for Cpk statistic CCpk: Check to store the CCpk statistic. PPM or % Total: Check to store PPM Total or % Total.USL: Check to store the CPU or Z. Cpm: Check to store the Cpm statistic.Quality Control Lower bound for Cp or Z. Lower bound for Pp or Z. Cpk: Check to store the Cpk statistic.275 0.Bench statistic.45 27. Variable 1 27. Upper bound for Pp or Z.36 27.354 0. PPM or % > USL: Check to store PPM > USL or % > USL.286 0.236 0.USL statistic.421 0. Lower bound for Cpk: Check to store the lower bound for Cpk statistic. Lower bound for Ppk: Check to store the lower bound for Ppk statistic. Upper bound for Ppk: Check to store the upper bound for Ppk statistic. CPL or Z.Bench statistic. .Bench: Check to store the upper bound for Cp or Z. Expected Performance "Overall" PPM or % < LSL: Check to store PPM < LSL or % < LSL.21 28. Data − Capability Analysis for Multiple Variables (Normal) Enter the data for each variable in a separate column. then you do not need to create a column of subgroup indicators.Bench statistic. Ppk: Check to store the Ppk statistic. PPM or % > USL: Check to store PPM > USL or % > USL. Expected Performance "Between/Within" PPM or % < LSL: Check to store PPM < LSL or % < LSL. Lower bound for Cpm: Check to store the lower bound for Cpm statistic.Bench: Check to store the lower bound for Cp or Z. PPM or % Total: Check to store PPM Total or % Total.78 26.LSL statistic. PPM or % > USL: Check to store PPM > USL or % > USL. PPU or Z.Bench: Check to store the Pp or Z.Bench statistic. PPL or Z. Upper bound for Cp or Z. If the subgroup size for each variable is the same.325 0. Variable 1 and Variable 2 have different subgroup sizes.321 Subgroup 2 1 1 2 2 2 3 3 3 274 2003 Minitab Inc.LSL: Check to store the PPL or Z. Pp or Z.LSL statistic.Bench statistic.284 0.61 28. PPM or % Total: Check to store PPM Total or % Total.USL statistic.

44 Target * * USL 10.8948 10. In Upper spec.284 0. In Variables.929 1.59 27. the analysis is based on nonmissing observations. You are concerned that the thickness of the beam might be affected by the morning and evening shift.703 3. The thickness must be between 10. You measure the thickness of 5 samples out of 10 boxes produced in each shift. For between-subgroup case.0234052 StDev(Overall) 0.325 0.45 27.96 Sample Mean 10. Check By variables.96 mm to meet the requirement. 1 2 3 4 5 6 Open the worksheet MCAPA.008 Cpk 0. For Between/Within subgroup analysis of each variable: • • • All subgroups should be the same size.703 3. type 10.0234052 0.44 10.321 Subgroup 1 1 2 2 2 3 3 3 To use Box-Cox transformation.286 0.21 28.96 10.236 0. Session window output Process Capability Analysis of Thickness by Shift Process Data Shift 1 2 Shift 1 2 LSL 10. type 5.0219872 0. then only one column is needed.0226887 Potential Capability Shift 1 2 Cp 3. Minitab displays error message when most of subgroup sizes are 1.008 CCpk 3.99 27.397 CPU 0. enter Shift.354 0. data must be positive. only subgroups of the majority size are used for estimating between-subgroup variation. type 10. Click OK. due to missing data or unequal subgroup sizes. Choose Stat > Quality Tools > Capability Analysis > Multiple Variable (Normal).61 28.01 Variable 2 0. In Subgroup sizes. If subgroups are not all the same size.275 0.96. Variable 1 and Variable 2 have the same subgroup sizes.703 CPL 6.44. enter Thickness.8892 Sample N 50 50 StDev(Within) 0. Example of Capability Analysis for Multiple Variables (Normal) Suppose you are interested in the capability of a process that produces support beams.703 2003 Minitab Inc. 275 . Variable 1 27.44 mm and 10. In Lower spec. In the example below.421 0.36 27.929 1.Process Capability If all variables have the same subgroup sizes.78 26.477 6.MTW. Missing values are allowed in within-subgroup analysis.

.30 Exp.58 1243.66 902.73 276 2003 Minitab Inc. Overall Performance Shift 1 2 PPM < LSL 0.0000000 PPM > USL 2670. Within Performance Shift 1 2 PPM < LSL 0.820 PPL 6.0000000 0.988 1.30 PPM Total 2670.Quality Control Overall Capability Shift 1 2 Pp 3.988 1.66 902.0000000 PPM > USL 1511.040 Cpm * * Observed Performance Shift 1 2 PPM < LSL 0 0 PPM > USL 0 0 PPM Total 0 0 Exp.58 1243.895 6.040 Ppk 0.0000000 0.599 PPU 0.942 3.73 PPM Total 1511.

Process Capability Graph window output 2003 Minitab Inc. 277 .

if you ignore the centering problem of the process. In Variables. 278 2003 Minitab Inc. Pp = 3. PPM > USL (2670. To conduct capability analysis for multiple variables 1 2 3 Choose Stat > Quality Tools > Capability Analysis > Multiple Variables (Normal). enter a lower and/or upper specification limit. This requirement appears to have been fulfilled.Nonnormal Capability Analysis for Multiple Variables (Nonnormal Distribution) Stat > Quality Tools > Capability Analysis > Multiple Variables (Nonnormal) You can assess the capability of an in-control process when each of the continuous variables follow a nonnormal distribution. For both the shifts. If you check By variables you can only enter one column in Variables. A generally accepted minimum value for the indices is 1. then perform a capability analysis on the transformed data. the process seems to have high capability. and Ppk are similarly defined as their counterparts in the Within table. You can enter up to 100 columns. CPL and CPU relates the process variation (σwithin) and mean (µ) to the specification limits. see data. PPL. enter one value. multiple values or a column where. or column(s) of subgroup indicators. For example. Dialog box items Variables: Enter columns containing the data for multiple variables. then you should investigate the issue and improve the process.820 times greater than the process range. PPU. If you like.58) indicates that 2670 out of 1 million beams exceed the upper specification limit of 10.942 indicate that the specification interval is 3.Quality Control Interpreting the results If you want to interpret the process capability statistics. PPU and PPL relates the process spread (the 3-σoverall variation) to a single-sided specification spread (µ-LSL or USLµ).942 times greater than the process range. your data should approximately follow a normal distribution. enter a grouping variable in By variable. Cp = 3. one value applies to all variables and multiple values correspond to columns listed in Variables. You can enter one value. you want to compare the width of parts from the same process before and after a process improvement. This is also evident in the histograms. The CPL and CPU for both the shifts are not equal. use any of the available dialog box options. By variables: Check to perform a separate capability analysis for each unique value in the column.703 times greater than the process range. The PPL and PPU for both the shifts indicate that the process is not centered The potential capability indices are close to the actual indices for both the shifts. You must specify at least one specification limit. You can also try to correct nonnormality in your data using a Johnson transformation. as shown by the probability plot. CPL is the ratio of µ-LSL to 3-σwithin variation. indicating that the process is not centered at the specification midpoint. which suggests that the variation across subgroup means are small for both shifts. In Lower spec or Upper spec. 4 5 4 Capability Analysis for Multiple Variables . However. Pp. For shift 1. Industry guidelines determine whether the process is capable. Fit data with Distribution: Choose a distribution. . high value of Cp and low value of Cpk indicate that the process has a centering problem. Use this procedure when: • • Comparing the results for multiple variables A grouping variable is associated with a continuous variable. Cp is defined as the ratio of the specification range to the potential process range ( the 6-σwithin variation). if centering of the process is a concern. enter the columns of quality characteristics. In Subgroup sizes. Cpk is the minimum of CPU and CPL.703 for both the shifts indicating that the specification spread for both the shifts is 3. To analyze a continuous variable associated with a grouping variable. For more details. Multiple values or columns corresponds to columns listed in Variables. respectively. multiple values. One value or a column applies to all variables or groups. For both the shifts.96 mm. For shift 2. The capability indices in the Overall table measures the actual capability of a process using σoverall variation as the process variation . and enter the column. CPU is the ratio of USL-µ to σwithin variation.820 indicate that the specification interval is 3. Pp = 3.33. then click OK.

3-parameter gamma. enter ∗. Set shape (Weibull or gamma) or scale (other dists) at: Enter the shape or scale parameter depending on your choice of distribution. If you have no upper specification limit. If each variable has different parameter values. Set threshold at: Enter the threshold parameter. or a column to specify the lower specification limit for multiple variables or groups. gamma. <Estimate> <Graphs> <Options> <Storage> Capability Analysis for Multiple Variables (Nonnormal) − Estimate Stat > Quality Tools > Capability Analysis > Multiple Variables (Nonnormal) > choose Distribution > Estimate Distributions Weibull. 2parameter exponential. To define a lower specification limit as a hard limit. and threshold. which means that measurements do not fall outside the limit. Minitab sets the expected percentage of values that are out of specifications to ∗ for a boundary. One value applies to all variables or groups. If you have no lower specification limit. enter constants or a column. Enter one value. largest extreme value. or a column. 279 . multiple values. 3-parameter lognormal. The default is 0. multiple values. enter 1. Multiple values correspond to columns listed in Variables or groups in By variables. Note You must enter a lower or upper specification limit. or a column to specify the upper specification limit for multiple variables or groups. enter columns equal to the number of variables in your study. Dialog box items P-Value to select best fit: Enter one value. lognormal. multiple values. Boundary: Check to define the lower specification limit as a hard limit. scale. or a column of p-values for selecting the best transformation. Multiple values correspond to columns listed in Variables or groups in By variables. scale. which means that measurements do not fall outside the limit. and 3parameter loglogistic Use to control the method for determining distribution parameters. One value applies to all variables or groups. One value applies to all variables. 3-parameter Weibull. The number of constants or values in a column must equal the number of parameter in a distribution. Capability Analysis for Multiple Variables (Nonnormal Distributions) − Estimate Stat > Quality Tools > Capability Analysis > Multiple variables (nonnormal) > choose Johnson transformation > Estimate You can enter a p-value for selecting the best transformation. 2003 Minitab Inc. Boundary: Check to define the lower specification limit as a hard limit. loglogistic. or both. multiple values. To define a lower specification limit as a hard limit. The number of values in each column must equal the number of parameter in a distribution Tip For 3-parameter loglogistic distribution. Dialog box items Estimate parameters of distributions: Choose to estimate the distribution parameters from the data. threshold): Check to specify historical estimates of parameters. enter 1. Enter one value. One value applies to all variables or groups. If all variables share same parameter values. One value applies to all variables or groups. Minitab estimates any parameters not specified below. enter 0. enter values of the parameters in the following order: location. Upper spec: Enter one value. otherwise. Use historical estimates (Use the order: location/shape. or a column. smallest extreme value. enter ∗. multiple values.1. enter 0.Process Capability Johnson transformation: Choose to transform your data to follow a normal distribution using a Johnson transformation. When you define upper and lower specification limits as boundaries. otherwise. Lower spec: Enter one value. Multiple values correspond to the columns listed in Variables. Multiple values correspond to columns listed in Variables or groups in By variables. Multiple values correspond to columns listed in Variables or groups in By variables. exponential.

Confidence level: Enter a confidence level between 0 and 100 (default is 95. Pp. Same x-axis scaling: Choose to display a histogram of all variables with the same x-axis scale. Cpk. Capability Analysis for Multiple Variables (Nonnormal Distributions) − Storage Stat > Quality Tools > Capability Analysis > Multiple Variables (Nonnormal) > choose Distribution > Storage Distributions Weibull. Calculate statistics using ___ sigma tolerance: Enter one value. Minimum: Enter a minimum value to display on the x-axis of the histogram of all variables. You can enter multiple values individually or in a column. Display Capability stats (Pp): Choose to display capability statistics. multiple values. USL: Check to store the upper specification limit. smallest extreme value. gamma. You can also specify the range to plot on the histogram. 280 2003 Minitab Inc. Capability Analysis for Multiple Variables (Nonnormal Distribution) − Options Stat > Quality Tools > Capability Analysis > Multiple Variables (Nonnormal) > Options Use to specify a target or confidence level for the confidence intervals for capability indices. Dialog box items Target (adds Cpm to table): Enter one value. loglogistic. Different x-axis scaling Choose to display a histogram of each variable with a different x-axis scale. Cpk. Minitab uses this level to calculate the confidence intervals of Cp. . Minitab then calculates Cpm in addition to the standard capability statistics. or a column that Minitab uses as a target. and Ppk. 3-parameter lognormal. Probability plot: Check to display a probability plot of the chosen distribution for each variable. If you do not enter a minimum value. and Ppk. Mean: Check to store the mean of the distribution. Multiple values correspond to columns listed in Variables. Minitab determines the value from the data. exponential. Maximum (Optional): Enter a maximum value for each variable. Maximum: Enter a maximum value to display on the x-axis of the histogram of all variables. Minitab determines the value from the data. Dialog box items Variable name: Check to store the variable name in the worksheet. One value applies to all variables. One value applies to all variables. and 3parameter loglogistic You can store your choice of statistics in the worksheet. Minimum (Optional): Enter a minimum value for each variable. Include confidence intervals: Check to include confidence intervals or bounds for Cp.0). Multiple values correspond to columns listed in Variables. The statistics available for storage depend on the options you chose in the Capability Analysis Multiple Variables (Nonnormal) dialog and subdialog boxes. Benchmark Z's: Choose to display benchmark Z statistics. 3-parameter Weibull. LSL: Check to store the lower specification limit. Target: Check to store target value in the worksheet. Dialog box items Histograms with density functions overlaid: Check to display a histogram with an overlaid density function for each variable. If you do not enter a maximum value. multiple values. Sample N: Check to store the number of observations (N). Confidence interval: Choose to get two-sided confidence intervals (the default) or just an upper or lower confidence bound. or a column that Minitab uses to calculate capability statistics using an interval other than six standard deviations wide (three on each side of the process mean). You can set preferences for the sigma tolerance for capability statistics using Tools > Options > Control Charts and Quality Tools > Capability Analysis.Quality Control Capability Analysis for Multiple Variables (Nonnormal Distribution) − Graphs Stat > Quality Tools > Capability Analysis > Multiple Variables (Nonnormal) > Graphs Use to display probability plots and histograms with an overlaid density function for each variable. Pp. largest extreme value. 3-parameter gamma. You can enter multiple values individually or in a column. 2parameter exponential. lognormal.

PPM > USL: Check to store the PPM > USL. Shape: Check to store the shape parameter of the distribution. Note Transformed statistics is indicated by ∗. LSL*: Check to store the transformed lower specification limit. Lower bound for Pp: Check to store the lower boundary for the Pp statistic. USL: Check to store the upper specification limit. PPM > USL: Check to store the PPM > USL. PPU: Check to store the PPU statistic. Johnson Transformation Distribution Type: Check to store the distribution type. Location: Check to store the location parameter. Pp: Check to store the Pp statistic. First shape: Check to store the first shape parameter. PPL: Check to store the PPL statistic. Threshold: Check to store the threshold parameter of the distribution. Scale: Check to store the scale parameter of the distribution. Dialog box items Variable name: Check to store the variable name in the worksheet. Upper bound for Ppk: Check to store the upper boundary for the Ppk statistic. USL*: Check to store the transformed upper specification limit. Ppk: Check to store the Ppk statistic. Capability Analysis for Multiple Variables (Nonnormal Distributions) − Storage Stat > Quality Tools > Capability Analysis > Multiple Variables (Nonnormal) > choose Johnson transformation > Storage You can store your choice of statistics in the worksheet. Expected Performance PPM < LSL: Check to store the PPM < LSL. Ppk: Check to store the Ppk statistic. 281 . Target: Check to store the target value in the worksheet. Upper bound for Pp: Check to store the upper boundary for the Pp statistic. StDev (Overall)*: Check to store the transformed overall standard deviation. PPU: Check to store the PPU statistic.Process Capability Location: Check to store the location parameter of the distribution. Sample N: Check to store the number of observations (N). StDev (Overall): Check to store the overall standard deviation. Lower bound for Ppk: Check to store the lower boundary for the Ppk statistic. LSL: Check to store the lower specification limit. PPM Total: Check to store the PPM Total. Mean*: Check to store the transformed mean of the distribution. 2003 Minitab Inc. The statistics available for storage depend on the options you chose in the Capability Analysis Multiple Variables (Nonnormal) dialog and subdialog boxes. Observed Performance PPM < LSL: Check to store the PPM < LSL. PPM Total: Check to store the PPM Total. Target*: Check to store the transformed target value in the worksheet. PPL: Check to store the PPL statistic. Second shape: Check to store the second shape parameter. Mean: Check to store the mean of the distribution. Pp: Check to store the Pp statistic.

99 27.36 27.45 27. Capability Analysis for Multiple Variables (Nonnormal Distribution) − Results Stat > Quality Tools > Capability Analysis > Mulitple Variables (Nonnormal) > Results Use to control the display of the output in the Session window. Dialog box items Control the display of results Display nothing: Choose to suppress the display of Session window results.61 28.236 0. Variable 1 27.321 Subgroup 1 1 2 2 2 3 3 3 282 2003 Minitab Inc.325 0.286 0. PPM Total: Check to store the PPM Total. Expected Performance PPM < LSL: Check to store the PPM < LSL. PPM > USL: Check to store the PPM > USL. PPM > USL: Check to store the PPM > USL. PPM Total: Check to store the PPM Total. Data − Capability Analysis for Multiples Variables (Nonnormal) Enter the data for each variable in a separate column.01 Subgroup 1 1 1 1 2 2 3 3 3 Variable 2 0.78 26.59 27.21 28. Display all: Choose to display Session window results. If the subgroup sizes of the variables differ.421 0.284 0.286 0.275 0.325 0. Variable 1 and Variable 2 have different subgroup sizes. In the example below.284 0.78 26. Observed Performance PPM < LSL: Check to store the PPM < LSL.01 Variable 2 0. Variable 1 and Variable 2 have equal subgroup sizes. In the example below. you do not need to create a column of subgroup indicators.36 27.354 0.Quality Control Scale: Check to store the scale parameter.275 0. create a column of subgroup indicators for each variable.236 0.45 27.21 28.421 0. . If the subgroup size for each variable is the same.321 Subgroup 2 1 1 2 2 2 3 3 3 If all variables have equal subgroup sizes.59 27.354 0. Variable 1 27. only one column is needed.99 27.61 28.

In Lower spec. Open the worksheet MNCAPA. Choose Stat > Quality Tools > Capability Analysis > Multiple Variable (Nonnormal). In Fit data with.2 2003 Minitab Inc. In Upper spec.90 PPL 1. 283 . The process is in control and the manager is wants to assess whether the amount packaged is affected by the usage of the two machines. In Variables. choose Distribution and then select Largest extreme value.48752 PPM > USL 10904.04393 Overall Capability Machine 1 2 Pp 0. Check BY variables and enter Machine. Overall Performance Machine 1 2 PPM < LSL 1.66 0.MTW. Random samples of size 50 for each machine are selected from the output and weighed.0 6803.84 0. enter Weight.03209 0.Process Capability Example of Capability Analysis for Multiple Variables (Nonnormal) A manufacturer uses two machines to package frozen food. Click OK.74 Ppk 0.1 6804. type 35.74 Observed Performance Machine 1 2 PPM < LSL 0 0 PPM > USL 0 0 PPM Total 0 0 Exp.5766 30.33 1.12097 1.7941 Scale 1.9409 29. Note 1 2 3 4 4 5 6 You can set preference for a method to calculate the capability statistics using Tools > Options > Control Charts and Quality Tools > Capability Analysis. type 27. The specifications for the weight of the frozen food packages are 31+4 oz.3686 Sample N 50 50 Location 29.35 PPU 0.7 PPM Total 10905. Session window output Process Capability Analysis of Weight by Machine Distribution: Largest Extreme Value Process Data Machine 1 2 LSL 27 27 Target * * USL 35 35 Sample Mean 30.66 0.

Quality Control Graph window output 284 2003 Minitab Inc. .

1 out of 1 million is expected to fall below the lower specification limit of 27 oz. You must specify at least one specification limit. A generally accepted minimum value for the indices is 1.5 − LSL to X0.5.33. You can enter up to 100 columns. 10904 out of 1 million are expected to exceed the upper specification limit of 35 oz. The process tends to put more food in a package than the upper limit. This also indicates that the process has median close to the lower specification limit. X0. you might have a pass/fail gage that determines whether an item is defective or not. Or.0013. Binomial distributions are usually associated with recording the number of defective items out of the total number sampled. then click OK. The percentiles are calculated using the parameter estimates for the largest extreme value distribution. This also indicates that more than 0. Under Fit data with.9974. • To transform your data to follow a normal distribution. For both machines the capability indices are lower than 1.66.13 percentiles denoted as X0.5.33 and PPU = 0. For machine 2.25. high value of Pp and low value of Ppk indicate that the process median is off the specification midpoint.84 and 0. choose Johnson transformation. use any dialog box options.5. For machine 1. The capability statistics are based on the 0. The PPM > USL (10904) indicates that for machine 1.Binomial Capability Analysis (Binomial) Stat > Quality Control > Capability Analysis > Binomial Use Capability Analysis (Binomial) to produce a process capability report when your data are from a binomial distribution. Machine 2 show similar results. This is also evident in the histogram. or a column where one value applies to all variables and multiple values correspond to the columns listed in Variables. PPL = 1. In Variables. indicating that the probability that the process produces conforming frozen food packets is slightly less than 0. 5 6 Capability Analysis . 285 .90 respectively. Pp is defined as the ratio of the specification range (USL − LSL) to the potential process range (X0.33. do one of the following: • To perform capability analysis using a nonnormal distribution. enter a grouping variable in By variable. For machine 1. Industry guidelines determine whether the process is capable. or both.0013 ). For example. For both machines. The PPM < LSL (1. enter the columns of quality characteristics.03209) indicates that for machine 1. The manufacturer needs to take immediate steps to improve the process.9987 − X0. you could record the number of people who call in sick on a particular day and the number of people scheduled to work each day. Ppk is the minimum of PPU and PPL. PPL is the ratio of X0.341 and P > 0. In Lower spec or Upper spec. enter a lower or upper specification limit. AD = 0.0013. AD = 0. choose Distribution and select a distribution. You could then record the total number of parts inspected and the number failed by the gage. Pp for machine 1 and machine two are 0. 99. Machine two show similar results. indicating that more than 0.9987.335 and P > 0.25.9987 − X0.5 − X0.5 to X0.Process Capability Interpreting the results The probability plot confirms that the data follows largest extreme value distribution. go/no-go) the probability of a success (or failure) is constant for each item the outcomes of the items are independent of each other P chart − verifies that the process is in a state of control Chart of cumulative %defective − verifies that you have collected data from enough samples to have a stable estimate of %defective Histogram of %defective − displays the overall distribution of the %defectives from the samples collected Capability Analysis (Binomial) produces a process capability report that includes the following: 2003 Minitab Inc. multiple values. and X0. You can enter one value. PPU is the ratio of USL − X0. Use Capability Analysis (Binomial) if your data meet the following conditions: • • • • • • • each item is the result of identical conditions each item can result in one or two possible outcomes (success/failure.13 percent pf the process output is more than the upper specification limit. If you like.87 and 0.13 percent of the process output is outside at least one of the specification limits. To conduct a capability analysis for multiple variables when data are nonnormal 1 2 3 4 Choose Stat > Quality Tools > Capability Analysis > Multiple Variables (Nonnormal). To analyze a continuous variable associated with a grouping variable.

all increasing or decreasing Test 4 Fourteen points in a row. you can perform the four tests for special causes. then check any combination of the tests listed below. as shown in the table below. enter number for the % defective target. Enter a target % Defective for this process (Optional) Target: If you like. Each test. Choose specific tests to perform: Choose to specify a subset of the four tests. The occurrence of a pattern suggests a special cause for the variation that should be further investigated. it is marked with the test number on the plot. all increasing or all decreasing Fourteen points in a row. Use Sizes In: Enter a column containing the number of items in each sample or subgroup. If a point fails more than one test. detects a specific pattern in the plotted data. When a point fails a test. This value must be between 0 and 1. alternating up and down Dialog box items Tests For Special Causes Perform all four tests: Choose to perform all four tests shown below. Test 1 line One point more than 3 sigmas from center Test 2 Nine points in a row on same side of center line Test 3 Six points in a row. alternating up and down 286 2003 Minitab Inc. .Quality Control • Defective rate plot − verifies that the %defective is not influenced by the number of items sampled Dialog box items Defectives: Enter a column containing the number of defectives from each sample or subgroup. <Tests> <Options> <Storage> Capability Analysis (Binomial) − Tests Stat > Quality Control > Capability Analysis > Binomial > Tests With Capability Analysis (Binomial). the number of the first test you request is the number printed on the plot. Minitab prints a summary table in the Session window with the complete information. You can set preferences for the tests to perform using Tools > Options > Control Charts and Quality Tools > Test to perform and the sensitivity of these tests using Tools > Options > Control Charts and Quality Tools > Define Tests. Sample Size Constant size: If all the samples are the same size. One point more than 3 sigmas from center line Nine points in a row on same side of center line Six points in a row. you may enter a constant as the sample size. Historical p (Optional): Enter a historical value for the proportion of defectives.

Enter the number that failed inspection in one column. Upper bound for %Defective: Check to store the upper bound for the percent of defectives. Lower bound for PPM Defective: Check to store the lower bound for parts per million defectives. If the total number inspected varies. %Defective: Check to store percent of defectives. there is a gap in the P chart where the subgroup would have been plotted. The statistics available for storage depend on the options you have chosen in the Capability Analysis (Binomial) dialog box. both numbers may vary. When subgroup sizes are unequal.Process Capability Capability Analysis (Binomial) − Options Stat > Quality Control > Capability Analysis > Binomial > Options Allows you to change the default title. The other plots and charts simply exclude the missing observation(s). Upper bound for Process Z: Check to store the upper bound for process Z. Upper bound for Average P: Check to store the upper bound for average P value. Process Z: Check to store the process Z. Each entry in the worksheet column should contain the number of defectives for a subgroup. 287 . you must also enter a corresponding column of subgroup sizes. Lower bound for Average P: Check to store the lower bound for average P value. Capability Analysis (Binomial) − Storage Stat > Quality Tools > Capability Analysis > Binomial > Storage Allows you to store you choice of statistics in the worksheet. Upper bound for PPM Defective: Check to store the upper bound for parts per million defectives. Dialog box items Variable name: Check to store the variable name. Dialog box items Title: Type a new title to replace the default title. Lower bound for Process Z: Check to store the lower bound for process Z. PPM Defective: Check to store the parts per million defectives. 2003 Minitab Inc. Data − Capability Analysis (Binomial) Use data from a binomial distribution. Suppose you have collected data on the number of parts inspected and the number of parts that failed inspection. On any given data. Target: Check to store the target value. enter subgroup size in another column: Failed 8 13 13 16 14 15 13 10 24 12 Inspect 968 1216 1004 1101 1076 995 1202 1028 1184 992 Missing Data If an observation is missing. Lower bound for %Defective: Check to store the lower bound for the percent of defectives. Average P: Check to store the average P value.

For example.Quality Control Unequal Subgroup Sizes In the P chart. how capable it is of answering incoming calls. the control limits are further from the center line for smaller subgroups than they are for larger ones. that is. the plot of %defective versus sample size will permit you to verify that there is no relationship between the two. Click OK. You record the number of calls that were not answered (a defective) by sales representatives due to unavailability each day for 20 days. In Use sizes in. When you do have unequal subgroup sizes.00 standard deviations from center line. You also record the total number of incoming calls. enter Calls. One point more than 3. Test Failed at points: 3 288 2003 Minitab Inc. 1 2 3 4 Open the worksheet BPCAPA. which is a common problem. Example of Capability Analysis (Binomial probability model) Suppose you are responsible for evaluating the responsiveness of your telephone sales department. In Defectives. In general. . Choose Stat > Quality Tools > Capability Analysis > Binomial. Graph window output Session window output Binomial Process Capability Analysis of Unavailable Test Results for P Chart of Unavailable TEST 1. the control limits are a function of the subgroup size. if you tend to have a smaller %defective when more items are sampled.MTW. this could be caused by fatigued inspectors. enter Unavailable. The subgroup size has no bearing on the other charts because they only display the %defective.

if you manufacture electrical wiring. where the item occupies a specified amount of time or specified space. then click OK. you may want to record the number of breaks in a piece of wire.Process Capability Interpreting the results The P chart indicates that there is one point out of control. For example. Enter the column. which is very poor. Use Sizes In: Choose to use a column containing the size of each item sampled. Enter a target DPU for this process (Optional). enter the sample size value in Constant size. 4 Capability Analysis . detects a specific pattern in the plotted data. • When your sample sizes vary. <Tests> <Options> <Storage> Capability Analysis (Poisson) − Tests Stat > Quality Control > Capability Analysis > Poisson > Tests With Capability Analysis (Poisson). 289 . you can perform the four tests for special causes. If you like. as shown in the table below. 2003 Minitab Inc. Use Capability Analysis (Poisson) when your data meet the following conditions: • • the rate of defects per unit of space or time is the same for each item the number of defects observed in the items are independent of each other Capability Analysis (Poisson) produces a process capability report for data from a Poisson distribution. The size of the item may vary. you may also record the size of each surface sampled. In Defectives. Target: If you like. if you manufacture appliances. This process could use a lot of improvement. The chart of cumulative %defect shows that the estimate of the overall defective rate appears to be settling down around 22%. Poisson data is usually associated with the number of defects observed in an item.Poisson Capability Analysis (Poisson) Stat > Quality Control >Capability Analysis > Poisson Use Capability Analysis (Poisson) to produce a process capability report when your data are from a Poisson distribution. The report includes the following: • • • • U chart − verifies that the process was in a state of control at the time the report was generated Chart of cumulative mean DPU (defects per unit) − verifies that you have collected data from enough samples to have a stable estimate of the mean Histogram of DPU − displays the overall distribution of the defects per unit from the samples collected Defect plot rate − verifies that DPU is not influenced by the size of the items sampled Dialog box items Defects: Enter a column containing the number of defects from each item sampled. The occurrence of a pattern suggests a special cause for the variation that should be further investigated. say in square inches. you may want to record the number of scratches on the surface of the appliance. enter the target DPU. enter the column containing the number of defectives. use any of the available dialog box options. you will have to record the size of each piece sampled. so you may also keep track of the size.75. Each test. Sample Size Constant size: Choose to use a constant sample size for the analysis. Enter the value. Do one of the following: • When your sample size is constant. Historical mu (Optional): Enter a column containing the size of each item sampled. The process Z is around 0. but more data may need to be collected to verify this. enter the column containing sample sizes in Use sizes in. Since the sizes of the surface may be different. The rate of defectives does not appear to be affected by sample size. If the lengths of the wire vary. To perform a capability analysis (binomial probability model) 1 2 3 Choose Stat > Quality Tools > Capability Analysis > Binomial. Or.

Dialog box items Variable name: Check to store the variable names used in the analysis. Lower bound for Mean Defective: Check to store the lower bound for average P statistics. Dialog box items. all increasing or decreasing Test 4 Fourteen points in a row. When a point fails a test. alternating up and down Capability Analysis (Poisson) − Options Stat > Quality Control > Capability Analysis > Poisson > Options Allows you to change the default title. 290 2003 Minitab Inc. then check any combination of the tests listed below. Title: Enter a title for the report. The statistics available for storage depend on the options you have chosen in the Capability Analysis (Poisson) dialog box. Capability Analysis (Poisson) − Storage Stat > Quality Tools > Capability Analysis > Poisson > Storage Allows you to store you choice of statistics in the worksheet. all increasing or all decreasing Fourteen points in a row. Minitab prints a summary table in the Session window with the complete information. alternating up and down Dialog box items Tests For Special Causes: Enter the number of tests you want performed in the P chart portion of the report.Quality Control You can set preferences for the tests to perform using Tools > Options > Control Charts and Quality Tools > Test to perform and the sensitivity of these tests using Tools > Options > Control Charts and Quality Tools > Define Tests. Upper bound for Mean Defective: Check to store the upper bound for average P statistics. Mean Defective: Check to store the average defectives. If a point fails more than one test. Test 1 One point more than 3 sigmas from center Test 2 Nine points in a row on same side of center line line Test 3 Six points in a row. Choose specific tests to perform: Choose to specify a subset of the four tests. If you do not enter a title. One point more than 3 sigmas from center line Nine points in a row on same side of center line Six points in a row. Perform all four tests: Choose to perform all four tests shown below. . the number of the first test you request is the number printed on the plot.. it is marked with the test number on the plot. a default title will be used.

You take random lengths of electrical wiring and test them for weak spots in their insulation by subjecting them to a test voltage. In Defects. Choose Stat > Quality Tools > Capability Analysis > Poisson. because they only display the DPU. The subgroup size has no bearing on the other charts. the control limits are further from the centerline for smaller subgroups than they are for larger ones. fatigued inspectors could cause this. 1 2 3 4 Open the worksheet BPCAPA. In general. Max DPU: Check to store the maximum DPU statistics. the control limits are a function of the subgroup size. The other plots and charts simply exclude the missing observation(s). there is a gap in the U chart where the subgroup would have been plotted. so you enter the subgroup size in another column. the plot of defects per unit (DPU) versus sample size will permit you to verify that there is no relationship between the two. You enter the number that failed to pass inspection in one column. enter Weak Spots. enter Length. Target DPU: Check to store the target DPU. When subgroup sizes are unequal. Lower bound for Mean DPU: Check to store the lower bound for average P statistics. if you tend to have a smaller DPE when more items are sampled. In Uses sizes in. Unequal Subgroup Sizes In the U chart. In this case.Process Capability Mean DPU: Check to store the mean defects per unit. Suppose you have collected daily data on the number of parts that have been inspected and the number of parts that failed to pass inspection. When you do have unequal subgroup sizes. For example. you must also enter a corresponding column of subgroup sizes. You record the number of weak spots and the length of each piece of wire (in feet). Example of Capability Analysis (Poisson probability model) Suppose you work for a wire manufacturer and are concerned about the effectiveness of the wire insulation process. 2003 Minitab Inc. Failed 8 13 13 16 14 15 13 10 24 12 Inspect 968 1216 1004 1101 1076 995 1202 1028 1184 992 Missing Data If an observation is missing. Min DPU: Check to store the minimum DPU statistics.MTW. the total number inspected varies from day to day. On any given day both numbers may vary. Upper bound for Mean DPU: Check to store the upper bound for average P statistics. Click OK. which is a common problem. 291 . Data − Capability Analysis (Poisson) Each entry in the worksheet column should contain the number of defectives or defects for a subgroup.

. To perform a capability analysis (Poisson distribution model) 1 2 3 Choose Stat > Quality Tools > Capability Analysis > Poisson. Test Failed at points: 36 Interpreting the results The U Chart indicates that there are three points out of control.00 standard deviations from center line. enter the sample size value in Constant size. • When your sample sizes vary. 4 292 2003 Minitab Inc. One point more than 3. In Defects. If you like. then click OK. signifying that enough samples were collected to have a good estimate of the mean DPU. The chart of cumulative DPU (defects per unit) has settled down around the value 0. enter the column containing sample sizes in Use sizes in.0265.Quality Control Graph window output Session window output Poisson Capability Analysis of Weak Spots Test Results for U Chart of Weak Spots TEST 1. use any of the available dialog box options. enter the column containing the number of defectives. Do one of the following: • When your sample size is constant. The rate of DPU does not appear to be affected by the lengths of the wire.

Historical standard deviation (optional): Enter a value for standard deviation of the population distribution if you have a known process parameter or an estimate obtained from past data. the report includes: To assess capability. the control limits will not be straight lines but will vary with the subgroup size. Note You must enter at least a lower spec and/or upper spec.5 (square root): Choose if you want Minitab to use lambda equal to square root of 0. Cpk. Minitab uses the transformed data for capability analysis. and Ppk To confirm normality. If you do not specify a value for the mean. you may want to force the control limits to be constant by specifying a fixed subgroup size. Cpm (if you specify a target). Pp. the report includes: • • • • • • • An Xbar chart (or Individuals chart for individual observations) An R chart or S chart (for subgroups size greater than 8) A run chart of the last 25 subgroups (or last 25 observations) A histogram of the process data A normal probability plot (also displays. Historical mean (optional): Enter a value for the mean of the population distribution if you have a known process parameter or an estimate obtained from past data. <Box-Cox> <Tests> <Estimate> <Options> Capability Sixpack (Normal Distribution) − Box-Cox Stat > Control Charts > Capability Sixpack > Normal > Box-Cox Use this option to transform nonnormal data to follow normal distribution and select a method for determining lambda. then perform capability on the transformed data. Subgroup size (use a constant or an ID column): Enter the subgroup size (for equal-size subgroups) or a column of subscripts (for unequal-size subgroups). Subgroups across rows of: Choose if subgroups are arranged in rows across several columns. Lower spec: Enter the lower specification limit. Lambda = 0 (natural log): Choose if you want Minitab to use lambda equal to zero. To confirm process stability. Enter the columns. it is estimated from the data. A model that assumes the data are from a normal distribution suits most process data. when this variation is proportional to the mean). See <Estimate> for estimation options.Process Capability Capability Sixpack . Anderson-Darling and P values) A process capability plot within and overall capability statistics. Use optimal lambda: Choose if you want Minitab to estimate an optimal value of lambda. Dialog box items Box-Cox power transformation (W=Y**Lambda): Check to perform Box-Cox transformation. Upper spec: Enter the upper specification limit. If your data are either very skewed or the within-subgroup variation is not constant (for example.5. If the subgroups are not equal. Cp. the report includes: You can also use this feature to correct normality in your data using a Box-Cox transformation. 2003 Minitab Inc. it is estimated from the data. If the subgroup sizes do not vary much. Dialog box items Data are arranged as Single column: Choose if data is in one column. 293 .Normal Capability Sixpack (Normal Distribution) Stat > Quality Tools > Capability Sixpack > Normal Use to produce process capability report when your data follow a normal distribution. see the discussion under nonnormal data. Enter a column. If you do not specify a value for standard deviation. Lambda = 0.

You can set preferences for the estimation of σ using Tools > Options > Control Charts and Quality Tools > Estimating Sigma. all increasing or all decreasing Fourteen points in a row. By default.Quality Control Other (enter a value between -5 and 5): Choose and enter your own value of lambda to bypass the lambda estimation step. The default is to use the unbiasing constants. alternating up and down Two out of three points more than 2 standard deviation from center line (same side) Four out of five points more than 1 standard deviation from center line (same side) Fifteen points in a row within 1 standard deviation of center line (either side) Eight points in a row more than 1 standard deviation from center line (either side) Capability Sixpack (Normal Distribution) − Estimation of Standard Deviation Stat > Quality Tools > Capability Sixpack > Normal > Estimate Allows you to specify the method for estimating the population standard deviation. When a point fails a test. Minitab estimates standard deviation using moving ranges instead. Subgroup sizes must be equal to perform the tests. Average moving range: Choose to estimate standard deviation using the average of the moving range. Minitab marks it by the lowest numbered test. . You can set preferences for the tests to perform using Tools > Options > Control Charts and Quality Tools > Test to perform and the sensitivity of these tests using Tools > Options > Control Charts and Quality Tools > Define Tests. Square root of MSSD: Choose to estimate standard deviation using the mean squared successive differences. 294 2003 Minitab Inc. Tests Stat > Quality Tools > Capability Sixpack > Normal > Tests Lets you select a subset of the eight tests for special causes. Use moving range of length: Choose to enter the number of observations used to calculate the moving range. Each test detects a specific pattern in the data plotted on the chart. Sbar: Choose to estimate standard deviation based on the average of the subgroup standard deviations. (for subgroup size > 1): Rbar: Choose to estimate standard deviation based on the average of the subgroup ranges. If a point fails more than one test. But when the subgroup size is 1 you cannot calculate sample standard deviations or ranges. Median moving range: Choose to estimate standard deviation using the median of the moving range. Dialog box items Tests For Special Causes (default definitions) Perform all eight tests: Choose to perform all eight tests for special causes. Dialog box items Methods of estimating standard deviation: See [2] for a discussion of the relative merits of the methods listed below. Use unbiasing constants: Check to estimate standard deviation either using or not using unbiasing constants. a span of 2 is used as consecutive values have the greatest chance of being alike. one that should be investigated. The occurrence of a pattern suggests a special cause for the variation. (for subgroup size = 1): Minitab generally estimates standard deviation using the sample standard deviations or ranges of each subgroup. The span must be < 100. Lets you control the sigma tolerance level and break out data for the final subgroups. Minitab marks it with the test number on the plot. Pooled standard deviation: Choose to estimate standard deviation using a pooled standard deviation. Capability Sixpack (Normal Distribution) − Options Stat > Quality Tools > Capability Sixpack > Normal > Options Allows you to transform the data. then check any combination of the tests listed below. One point more than 3 standard deviation from center line Nine points in a row on same side of center line Six points in a row. and prints a summary table with complete information in the Session window. Choose specific tests to perform: Choose to specify a subset of the eight tests.

In Single column. enter Supp1. Such an omission may cause the control chart limits and the center line to have different values for that subgroup. An chart showed you that Supplier 2's camshaft production was out of control. type 598. To use the Box-Cox transformation. Minitab then calculates Cpm. 1 2 3 4 5 Open the worksheet CAMSHAFT. In Upper spec. One of the parts. Capability Sixpack calculates Cpk using an interval that is 6 standard deviation wide. In Subgroup size. There has been a chronic problem with camshaft lengths being out of specification − a problem that has caused poor-fitting assemblies down the production line and high scrap and rework rates. Data − Capability Sixpack Normal You can enter individual observations or data in subgroups. Choose Stat > Quality Tools > Capability Sixpack > Normal. type the desired text in this box. Dialog box items Break out data for last___subgroups: Specifies the number of subgroups to use for the Run chart at the lower left of the graph. type 602. Title: To replace the default title with your own custom title. This is referred to as "6-sigma quality. or in rows across several columns. 6 on either side of the process mean. After dropping Supplier 2. you discovered that there were two suppliers for the camshafts. but the problems have not completely disappeared. the number of poor quality assemblies has dropped significantly. see Data. If an entire subgroup is missing. then set up a second column of subgroup indicators. For examples. you must have two or more observations in at least one subgroup in order to estimate the process standard deviation. must be 600 mm +2 mm long to meet engineering specifications. Click OK in each dialog box. When you have subgroups of unequal size. if you enter 12. Individual observations should be structured in one column. 3 on either side of the process mean." By default. In Lower spec. The default is 25. type 5.MTW. 2003 Minitab Inc. Subgroups need not be the same size. Example of Capability Sixpack (normal probability model) Suppose you work at an automobile manufacturer in a department that assembles engines. Target (adds Cpm to table): Enter a number that Minitab uses as a target. displaying the target and Cpm values in a table below the graph. In Target (adds Cpm to table). data must be positive. a camshaft. so you decided to stop accepting production runs from them until they get their production under control. You decide to run a capability sixpack to see whether Supplier 1 alone is capable of meeting your engineering specifications. Subgroup data can be structured in one column. Click Options. If you have data in subgroups. 295 . Minitab omits it from the calculations of the statistics for that subgroup. enter the subgroups in a single column. and R Upon examination of the inventory records. enter 600. in addition to the standard capability statistics. For example. Calculate statistics using___sigma tolerance: Enter the value you would like to use as the sigma tolerance level for calculating Cpk. there is a gap in the chart where the statistic for that subgroup would have been plotted. If a single observation in the subgroup is missing.Process Capability You can set preferences for the sigma tolerance for capability statistics using Tools > Options > Control Charts and Quality Tools > Capability Analysis. that says to use an interval 12 standard deviations wide.

.5 makes the data "more normal." For details. implying a stable chart to see if the points follow each process. again. So you will run the capability sixpack using a Box-Cox transformation on the data. you can see that the process tolerance falls below the lower specification limit. you measure warping in ten tiles each working day for ten days.5 (square root). If you want to interpret the process capability statistics. with no trends or shifts − also indicating process stability. Click OK in each dialog box. This means you will sometimes see camshafts that do not meet the lower specification of 598 mm. These patterns indicate that the data are normally distributed. type 10. the points are randomly distributed between the control limits. It is also important to compare points on the R chart with those on the other.MTW.16) and Cpk (0. Also. Choose Lambda = 0. the values of Cp (1. In Single column. you found that the tile data do not come from a normal distribution. the data approximately follow the normal curve. Choose Stat > Quality Tools > Capability Sixpack > Normal. implying a stable process.Quality Control Graph window output Interpreting the results On both the chart and the R chart. From previous analyses. To ensure production quality.33. On the capability histogram. On the normal probability plot. Example of Capability Sixpack (Box-Cox transformation) Suppose you work for a company that manufactures floor tiles. your data should approximately follow a normal distribution. In Subgroup size. In Upper spec. and that a Box-Cox transformation using a lambda value of 0. see Example of a capability analysis with a Box-Cox transformation. Yours do not. 1 2 3 4 5 6 Open the worksheet TILES. But from the capability plot. the points approximately follow a straight line. Click Options. and are concerned about warping in the tiles. 296 2003 Minitab Inc. indicating that Supplier 1 needs to improve their process. Check Box-Cox power transformation (W = Y**Lambda).90) are below the guideline of 1. The points on the run chart make a random horizontal scatter. enter Warping. type 8.

and enter the columns containing the rows in the box.Process Capability Graph window output Interpreting the results On both the chart and the R chart.33. You must enter at least one of them. enter a subgroup size or column of subgroup indicators. For individual observations. the points are randomly distributed between the control limits. In Lower spec or Upper spec. respectively. To make a Capability Sixpack (normal probability) 1 2 Choose Stat > Quality Tools > Capability Sixpack > Normal Do one of the following: • When subgroups or individual observations are in one column. enter a subgroup size of 1.74) fall below the guideline of 1. with no trends or shifts − also indicating process stability. Now the process capability statistics are appropriate for this data. the data follow the normal curve. shows that the process is not meeting specifications. As you can see from the capability histogram. the points approximately follow a straight line. These patterns indicate that the Box-Cox transformation "normalized" the data. use any of the available dialog box options. 3 4 2003 Minitab Inc. enter a lower and/or upper specification limit. And the values of Cpk (0. implying a stable process. The capability plot. In Subgroup size. • When subgroups are in rows. enter the data column in Single column. Yours do not − again.76) and Ppk (0. It is also important to compare points on the R chart with those on the chart for the same data to see if the points follow each other. on the normal probability plot. 297 . If you like. choose Subgroups across rows of. The points on the run chart make a random horizontal scatter. implying a stable process. so your process does not appear to be capable. however. Also. then click OK.

Between/Within Capability Sixpack (Between/Within) Stat > Quality Tools > Capability Sixpack > Between/Within Use to produce process capability report when your data follow a normal distribution. the report includes: • • • • • • • An Xbar chart (or Individuals chart for individual observations) An Range (R) chart or S chart (for subgroups size greater than 8) A run chart of the last 25 subgroups (or last 25 observations) A histogram of the process data A normal probability plot (also displays. Enter a column. see the discussion under Nonnormal data. If your data are very skewed. Cpk. Between subgroups: (optional): Enter a value for the between-subgroup sigma (standard deviation of the population distribution) if you have a known process parameter or an estimate obtained from past data. Historical sigmas Within subgroup: (optional): Enter a value for the within-subgroup sigma (standard deviation of the population distribution) if you have a known process parameter or an estimate obtained from past data. it is estimated from the data. Minitab can estimate between-subgroup sigma in one of three ways. it is estimated from the data. Enter the columns. To confirm process stability. the report includes: You can also use this feature to correct normality in your data using a Box-Cox transformation. 298 2003 Minitab Inc. Upper spec: Enter the upper specification limit. Note You must enter a lower or upper specification limit. Use this feature when you suspect that you may have both between-subgroup and within-subgroup variation. or both. If you do not specify a value for sigma. Minitab can estimate within-subgroup sigma in one of three ways. Historical mean (optional): Enter a value for the mean of the population distribution if you have a known process parameter or an estimate obtained from past data. A model that assumes that the data are from a normal distribution suits most process data. If the subgroups are not equal. Subgroups across rows of: Choose if subgroups are arranged in rows across several columns. the report includes: To assess capability. Cp. Minitab uses transformed data for capability analysis. Cpm (if you specify a target). and Ppk To confirm normality. Note You can only use this procedure with nonnegative data. Dialog box items Data are arranged as Single column: Choose if data is in one column. Anderson-Darling and P values) A process capability plot Between/within and overall capability statistics. If you do not specify a value for sigma.Quality Control Capability Sixpack . Subgroup size (use a constant or an ID column): Enter the subgroup size (for equal-size subgroups) or a column of subscripts (for unequal-size subgroups). <Box-Cox> <Tests> <Estimate> <Options> Capability Sixpack (Between/Within) − Box-Cox Stat > Control Charts > Capability Sixpack > Between/Within > Box-Cox Use this option to transform nonnormal data to follow normal distribution and select a method for determining lambda. you may want to force the control limits to be constant by specifying a fixed subgroup size. Pp. If you do not specify a value for the mean. Lower spec: Enter the lower specification limit. then perform capability on the transformed data. the control limits will not be straight lines but will vary with the subgroup size. . If the subgroup sizes do not vary much. it is estimated from the data.

Process Capability Dialog box items Box-Cox power transformation (W=Y**Lambda): Check to perform Box-Cox transformation. all increasing or all decreasing Fourteen points in a row. 299 . One point more than 3 standard deviations from center line Nine points in a row on same side of center line Six points in a row. Use optimal lambda: Choose if you want Minitab to estimate an optimal value of lambda. 2003 Minitab Inc. alternating up and down Two out of three points more than 2 standard deviations from center line (same side) Four out of five points more than 1 standard deviation from center line (same side) Fifteen points in a row within 1 standard deviation of center line (either side) Eight points in a row more than 1 standard deviation from center line (either side) Capability Sixpack (Between/Within) − Estimation of Standard Deviation Stat > Quality Tools > Capability Sixpack > Between/Within > Estimate Allows you to specify the method for estimating the population standard deviation. Dialog box items Tests For Special Causes (default definitions) Perform all eight tests: Choose to perform all eight tests. When a point fails a test. Minitab marks it with the test number on the plot. Each test detects a specific pattern in the data plotted on the chart. Within subgroup Rbar: Choose to estimate within-subgroup standard deviation using the average of the subgroup ranges. Pooled standard deviation: Choose to estimate within-subgroup standard deviation using a pooled standard deviation. and prints a summary table with complete information in the Session window. Capability Sixpack (Between/Within) − Tests Stat > Quality Tools > Capability Sixpack > Between/Within > Tests Lets you select a subset of the eight tests for special causes. The default is to use the unbiasing constants. one that should be investigated. You can set preferences for the tests to perform using Tools > Options > Control Charts and Quality Tools > Test to perform and the sensitivity of these tests using Tools > Options > Control Charts and Quality Tools > Define Tests. Between subgroups Average moving range: Choose to estimate between-subgroups standard deviation using the average of the moving range. Dialog box items Methods of estimating standard deviation: See [1] for a discussion of the relative merits of the methods listed below.5 (square root): Choose if you want Minitab to use lambda equal to square root of 0. The occurrence of a pattern suggests a special cause for the variation. Choose specific tests to perform: Choose to specify a subset of the four tests. Sbar: Choose to estimate within-subgroup standard deviation using the average of the subgroup standard deviations. You can set preferences for the estimation of σ using Tools > Options > Control Charts and Quality Tools > Estimating Sigma. If a point fails more than one test.5. then check any combination of the tests listed below. Minitab marks it by the lowest numbered test. when more than half the subgroups are the same size. Lambda = 0 (natural log): Choose if you want Minitab to use lambda equal to zero. Note Minitab performs all eight tests. Lambda = 0. Use unbiasing constants: Check to estimate standard deviation either using or not using unbiasing constants. Other (enter a value between -5 and 5) : Choose and enter your own value of lambda to bypass the lambda estimation step. Median moving range: Choose to estimate between-subgroups standard deviation using the median of the moving range.

Use moving range of length: Choose to enter the number of observations used to calculate the moving range. enter 47. To use the Box-Cox transformation. Minitab then calculates Cpm in addition to the standard capability statistics.MTW. Click Tests. enter 53. Control limits for the Individuals and Moving Range charts are based on the majority subgroup size. Capability Sixpack (Between/Within) − Options Stat > Quality Tools > Capability Sixpack > Between/Within > Options Allows you to transform the data using a Box-Cox power transformation. a span of 2 is used as consecutive values have the greatest chance of being alike. Select Stat > Quality Tools > Capability Sixpack > Between/Within. The span must be < 100. You can specify a process target. only subgroups of the majority size are used for estimating the between-subgroup variation.Quality Control Square root of MSSD: Choose to estimate between-subgroups standard deviation using the mean squared successive differences. You are concerned that the paper is being coated with the correct thickness of film and that the coating is applied evenly throughout the roll. data must be non-negative. Ideally. displaying the target and Cpm values in a table below the graph. Example of capability sixpack (between/within) Suppose you are interested in the capability of a process that coats rolls of paper with a thin film. Click OK in each dialog box. 1 2 3 4 5 Open the worksheet BWCAPA. If your subgroups are not all the same size. In Subgroup size. you use Minitab to conduct a Capability Sixpack (Between/Within). You can set preferences for the sigma tolerance for capability statistics using Tools > Options > Control Charts and Quality Tools > Capability Analysis. Minitab then calculates Cpm. in addition to the standard capability statistics. Choose Perform all eight tests. enter Coating. Subgroup data can be structured in one column or in rows across several columns. Data − Capability Sixpack (Between/Within) You can enter data in subgroups. . Dialog box items Target (add Cpm to table): Enter a number that Minitab uses as a target. In Upper spec. enter Roll. In Lower spec. all subgroups should be the same size. The thickness must be 50 + 3 mm to meet engineering specifications. 300 2003 Minitab Inc. By default. In Single column. You take three samples from 25 consecutive rolls and measure coating thickness. Because you are interested in determining whether or not the coating is even throughout a roll. with two or more observations per subgroup. type the desired text in this box. or nominal specification. due to missing data or unequal sample sizes. Title: To replace the default title with your own custom title. Calculate statistics using: (sigma tolerance): Enter a number to calculate the capability statistics using an interval other than 6 standard deviations wide (3 on either side of the process mean).

on the normal probability plot. so your process could use some improvement.796740 Interpreting the results If you want to interpret the process capability statistics.14) fall just below the guideline of 1. • When subgroups are in rows. No points failed the eight tests for special causes. In Lower spec or Upper spec.685488 0.Process Capability Graph window output Session window output I-MR-R/S Standard Deviations of Coating Standard Deviations Between Within Between/Within 0. choose Subgroups across rows of. Also. This criteria appears to have been met. and enter the columns containing the rows in the box. The points on the Individuals and Moving Range chart do not appear to follow each other.33. enter a lower and/or upper specification limit.21) and Ppk (1. the data approximately follow the normal curve. the points approximately follow a straight line. enter a subgroup size or column of subgroup indicators. respectively.406080 0. again indicating a stable process. thereby implying that your process is in control. The values of Cpk (1. The capability plot shows that the process is meeting specifications. enter the data column in Single column. You must enter at least one 3 2003 Minitab Inc. In the capability histogram. your data need to come from a normal distribution. In Subgroup size. 301 . To perform a capability sixpack (between/within) 1 2 Choose Stat > Quality Tools > Capability Sixpack > Between/Within Do one of the following: • When subgroups are in one column.

Quality Control of them. 3-parameter Weibull. . 2parameter exponential. and distribution parameters To confirm normality. Gamma. Subgroup size (use a constant or an ID column): Enter the subgroup size (for equal-size subgroups) or a column of subscripts (for unequal-size subgroups). Lower spec: Enter the lower specification limit.Nonnormal Capability Sixpack (Nonnormal Distribution) Stat > Quality Tools > Capability Sixpack > Nonnormal Use to produce process capability report when your data do not or transformed (using Johnson distribution system) to follow a normal distribution. the report includes: When using a nonnormal model. Smallest extreme value. Lognormal. the control limits will not be straight lines but will vary with the subgroup size. 3-parameter lognormal. then perform capability on the transformed data. <Estimate> <Options> Capability Sixpack (Nonnormal Distribution) − Estimate Stat > Quality Tools > Capability Sixpack < Nonnormal > choose Distributions > Estimate Distributions Weibull. 4 If you like. Johnson transformation: Choose to transform your data to follow a normal distribution using the Johnson transformation. use any of the available dialog box options. 3-parameter gamma. To confirm control. and 3parameter loglogistic Use to control the method for determining distribution parameters. If the subgroups are not equal. To select the best fitting distribution. Subgroups across rows of: Choose if subgroups are arranged in rows across several columns. Loglogistic. You can also use this feature to correct nonnormality in your data using a Johnson transformation. Largest extreme value. the report includes: • • • • • • • An Xbar chart (or I chart for individual observations) An R chart (or MR chart for individual observations) A run chart of the last 25 subgroups (or last 25 observations) A histogram of the process data A probability plot A process capability plot Overall capability statistics Pp. the report includes: To assess capability. Fit data with Distribution: Choose a distribution. Exponential. Ppk. Pp and Ppk. Minitab bases the calculations on maximum likelihood estimates of the distribution parameters. Minitab only calculates the overall capability statistics. Enter the column. Enter the columns. Capability Sixpack . Upper spec: Enter the upper specification limit. then click OK. rather than mean and variance estimates as in the normal case. 302 2003 Minitab Inc. Dialog box items Data are arranged as Single column: Choose if data are in one column. Note You must enter at least a lower and/or upper specification limit. see individual distribution identification.

Choose Stat > Quality Tools > Capability Sixpack > Nonnormal. choose Single column. Calculate statistics using:____sigma tolerance: Enter a number to calculate the capability statistics (3 on each side of the process mean). This may cause the control chart limits and the center line to have different values for that subgroup. type 10. Title: Type a new title to replace the default title. Under Data are arranged as. there is a gap in the chart. Minitab calculates the statistics for that subgroup based on non-missing observations. Under Fit data with. Use historical estimates (Use the order: Location/Shape. You can set preferences for the sigma tolerance for capability statistics using Tools > Options > Control Charts and Quality Tools > Capability Analysis. Example of Capability Sixpack (Nonnormal) Suppose you work for a company that manufactures floor tiles. then select Weibull. If an entire subgroup is missing. Note 1 2 3 4 5 You can set preference for a method to calculate the capability statistics using Tools > Options > Control Charts and Quality Tools > Capability Analysis. shape. type 8. Click OK . The data do not follow a normal distribution. Also allows you to enter a new title to replace the default title. 2003 Minitab Inc. and threshold. Data − Capability Sixpack (Nonnormal Distribution) You can enter individual observations or data in subgroups. Scale. and threshold.Process Capability Dialog box items Estimate parameters of distributions: Choose to estimate the distribution parameters from the data. The default is 0. you measured warping in ten tiles each day for ten days. and are concerned about warping in the tiles. For more information. then enter Warping. or in rows across several columns. In Subgroup size. 303 . Note Threshold parameter is applicable to 3-parameter distributions only. Open the worksheet TILES. Set shape (Weibull or gamma) or scale (other dists) at: Enter the shape or scale parameter depending on your choice of distribution. You must enter parameter values in the following order: location. enter values of the parameters in the following order: location. Minitab estimates any parameters not specified below. Dialog box items P-Value to select best fit: Type a p-value for selecting the best transformation. so you decide to analyze your data using capability sixpack based on a Weibull probability model. Dialog box items Break out data for last___subgroups: Specifies the number of subgroups to use for the Run chart. Individual observations should be structured in one column. The default is 25. Capability Sixpack (Nonnormal Distribution) − Options Stat > Quality Tools > Capability Sixpack > Nonnormal > Options Use to control the sigma tolerance level and break out data for the final subgroups. For example.MTW. Subgroup data can be structured in one column. choose Distribution. scale. If a single observation in a subgroup is missing. In Upper spec. Enter a value for each parameter of the distribution. To ensure production quality. for 3-parameter loglogistic distribution. Capability Sixpack (Nonnormal Distribution) − Estimate Stat > Quality Tools > Capability Analysis > Nonnormal > choose Johnson transformation > Estimate You can enter a p-value for selecting the best transformation. create a second column of subgroup indicators. scale. Set threshold at: Enter the threshold parameter. see Data. Threshold): Choose to specify historical estimates of parameters. For subgroups of unequal size arranged in a column.10.

If you like. The shape (1. enter the data column in Single column. In Lower spec and Upper spec. 3 4 5 304 2003 Minitab Inc. In Subgroup size. • To transform your data to follow a normal distribution. Also.247553 and p = 0. The capability plot. however.33. You can enter one or both specification limits. For individual observations. A generally accepted minimum value for the indices is 1. Note To choose the best fitting distribution using the same data. To make a Capability Sixpack with nonnormal data 1 2 Choose Stat > Quality Tools > Capability Sixpack > Nonnormal. and enter the columns containing the data in the box. choose Subgroups across rows of. then click OK. Pp (∗) as no lower specification limit is specified. do one of the following: • When subgroups or individual observations are in one column. enter lower and upper specification limits. the points approximately follow a straight line and fall within the confidence interval suggesting that Weibull distribution fits our data well. Industry guidelines determine whether the process is capable. on the Weibull probability plot (AD = 0. .33. • When subgroups are in rows. enter a subgroup size of 1.25). Under Data are arranged as. choose Distribution and select a distribution. see example of individual distribution identification. use any dialog box options. Under Fit data with. choose Johnson transformation. For other examples using this data. so your process does not appear to be capable. shows that the process is not meeting specifications. Ppk (0.Quality Control Graph window output Interpreting the results The capability histogram does not show evidence of any serious discrepancies between the fitted Weibull distribution and the data. choose one of the following: • To perform capability analysis using a nonnormal distribution.27812) are the estimated parameters for the Weibull distribution based on the maximum likelihood method.69368) and scale (3. enter a subgroup size or column of subgroup indicators. see Example of Capability Analysis for Nonnormal Data and Example of Capability Sixpack using a Box-Cox Transformation.72869) falls below the guideline of 1.

. 123...............Normal (Stat Menu) . 226 Gage Run Chart ............ 302 Capability Sixpack (Normal)................ 123........................ 157. 129....................... 124 For defects ............ 90............................ 27................................ 78....................................................... 149.......................................... 226 Gage R&R Study (Nested) (Stat menu)......... 197.... 278 Capability Analysis (Multiple Variables (Normal)) .......................... 217 Gage R&R Study (Nested) ............ 176 CUSUM Chart (Stat menu) ....... 9 Cohen's kappa .................................................. 45 I-MR-R/S (Between/Within) Chart......... 125................ 285 Capability Analysis (Multiple Variables (Nonnormal))278 Capability Analysis ..................................................................Poisson (Stat menu) ............... 236 Attribute Gage Study (Analytic Method) (Stat menu) . 109......................... 45 Individual Distribution Identification .... 264 Capability Analysis (Normal)................... 214 Gage Linearity and Bias Study (Stat menu) .......Binomial (Stat menu) ................................. 165................................ 289 Capability Sixpack (Between/Within) ...... 90 I-MR-R/S (Between/Within) Chart ................ 185.................... 105........................................................ 245 Individual Distribution Identification (Stat menu).. 269 Capability Analysis .......................... 140 C Chart (Stat menu) ............... 193......... 124 Control charts for ...................................... 293 Capability Sixpack ........... 20 Historical stages ...... 245 Individuals Chart................ 232 Control Charts (Stat menu) ................................ 69................... 139...................................................................... 105 Individuals Chart (Stat menu) . 33.............. 189................................................ 19 Attributes ....................................... 194 Control limits.....................................Multiple Variables (Normal) (Stat menu)............................................... 169................ 161 EWMA Chart (Stat menu) ................. 260 Capability Analysis ..Between/Within (Stat menu) 260 Capability Analysis (Binomial)........... 229 Attribute Gage Study (Analytic Method) ........................................... 264 Capability Analysis ........... 124 Destructive testing ..................................... 187 Defectives.. 132. 293 Cause-and-Effect ...Normal (Stat menu) .................................... 145 Attributes data.................... 169 Variables for individuals ......... 194 Genearlized Variance Chart (Stat menu)............. 179............ 119.................................................................. 204 Historical values ................................... 9 Cause-and-Effect (Stat menu)...... 173......... 253 Capability Analysis ................... 211 Gage Run Chart (Stat menu) ......................................... 236 Attributes control charts ....... 9 Fleiss' kappa.......................... 214 Gage R&R Study (Crossed) .........Between/Within (Stat menu) .......................... 177.................................. 305 ....................... 124 Control charts for ....... 173. 253 Capability Analysis (Poisson)................... 99.................Multiple Variables (Nonnormal) (Stat menu) .. 298 Capability Sixpack ............ 73..........Index A Attribute Agreement Analysis...... 63.. 105 Interpreting capability indices ........................... 181................ 211 Generalized Variance Chart .. 243 Ishikawa diagram........ 160.... 42........ 239 B Bias ....... 245 EWMA Chart... 25 Box-Cox Transformation (Stat menu)........ 114 Variables for subgroups .............. 269 Capability Analysis (Nonnormal).. 200 Time-weighted......................Nonnormal (Stat menu) ......................................... 54........................... 23........................ 90 I-MR Chart (Stat menu)............................................ 60....... 302 Capability Sixpack ....................................... 143.................................................. 36.........................................Nonnormal (Stat menu) ........... 298 Capability Sixpack (Nonnormal) ................ 171.. 169 2003 Minitab Inc........................ 145 For defectives.... 9 J Johnson Transformation....................................................... 285 Capability Analysis ......... 232 G Gage Linearity and Bias Study ....... 140 Capability Analysis (Between/Within) ......... 289 Capability Analysis ............ 124 Defects ............... 82 H Historical charts ............................................................ 25 C C Chart.......................... 95................................................................... 229 Attribute Agreement Analysis (Stat menu) ................................................. 45................ 20 I I-MR Chart .... 132................................ 136............................................... 86............. 244.................. 19 CUSUM Chart..... 125...... 250 D Decomposed Tsquared statistic .............................................................. 152............................................ 217 Gage R&R Study (Crossed) (Stat menu).. 226 E Estimating process standard deviation.............. 139..... 214 Box-Cox Transformation .............. 161 F Fishbone diagram......... 124 Multivariate ...................................... 50...................................................................................

...... 21 Time-weighted control charts ... 269.............................................. 23 T Tests for special causes ........ 185........... 253............................................................................................. 152 Moving Range Chart .......... 232 Kendall's coefficient of concordance.......... 289.................................. 84 Special causes .. 210 P P Chart................................................. 54........... 285 To conduct capability analysis for multiple variables 278 Transformation ................................................................................. 13 Multivariate control charts ..................... 45...... 239....... 193..... 4 306 2003 Minitab Inc............................................... 54 Xbar Chart (Stat menu).. 146 Unequal-size subgroups...................... 214 Gage............................................................................................................ 73 Shewhart chart ........................................... 178 Tsquared ............................................. 278..................................... 152.................... 285..................... 132 NP Chart (Stat menu).......... 82 U U Chart ........................................................................... 152 Moving Average Chart (Stat menu).......................................... 146 U Chart (Stat menu)........................................................... 5 Process Capability .............................................................................................................................Generalized Variance Chart . 250 K Kappa........................................................... 211 Moving Average Chart ........... 21 Subgroups ..................... 64 Run Chart............................................................. 5 Pareto Chart (Stat menu) .. 232 L Linearity ........................................... 114 Multi-Vari Chart ...................................................... 82 Zone Chart (Stat menu) ........ 36..... 28 Xbar-R Chart (Stat menu) .............................. 27.................................................................................................................................. 25................... 214 Gage R&R Study (Crossed) .................................................................. 14 Symmetry Plot (Stat menu).....Generalized Variance Chart......................................... 169 To conduct a capability analysis for multiple variables when data are non-normal . 217 Gage R&R Study (Nested) .... 28 Xbar-S Chart.. 2 Run Chart and Runs Tests comparison ..................................... 73.................... 298.............. 73 S Chart (Stat menu) .......................................................... 232 Cohen's kappa ............. 293.................................. 201... 21 Tests for ............................................... 226 Gage Run Chart .......... 64 R Chart (Stat menu) ........ 63......................................... 245........................................................................................................ 54 Xbar-R Chart .............................. 214 M Measurement systems analysis....................... .......... 84 Comparing with zone chart ........................... 232 Fleiss' kappa ...................................... 132 Number of distinct categories ............................................................................................ 229 Attribute Gage Study (Analytic Method) ............................................. 177.......... 236 Gage Linearity and Bias Study.................. 209 Attribute Agreement Analysis . 114 Moving Range Chart (Stat menu)...... 250............ 232 Kendall's correlation coefficient ... 250 Tsquared .............................................. 82 X Xbar Chart ........ 260...................... 14 Run Chart (Stat menu)....... 125 Pareto Chart...... 186 S S Chart ..................... 186 Tsquared Chart (Stat menu) ............................. 13 Multi-Vari Chart (Stat menu)... 37 Xbar-S Chart (Stat menu) . 23 Unequal sizes ................................................................ 114 Variables control charts for subgroups ................................................................................................ 90......................................................................... 2 V Variables control charts for individuals .................................................. 200 Multivariate EWMA Chart..... 105......... 37 Z Z-MR Chart.... 99 Zone Chart............................. 23 Symmetry Plot ............. 99............ 202 Multivariate EWMA Chart (Stat menu) ..................................................................................................... 125 P Chart (Stat menu) ............................. 178 Tsquared Chart............Quality Control Johnson Transformation (Stat menu)......................... 242 NP Chart ......... 264............................................... 201 N Nonnormal data ......................................................... 302 Q Quality Tools (Stat menu) .................................................... 1 R R Chart.... 160..................................... 99 Z-MR Chart (Stat menu) ................................

- 6 Ifrs Financial Stat
- thinkcell manual.pdf
- MSA.pdf
- 13-FR Amendment Slides by Parveen Sharma Sir for CA Final-MAY 2016
- CA Final Allied Law Amendment Charts for May 2016 Y91LA9XM
- 1.TIME VALUE OF MONEY.pdf
- SFM_Workbook_for_May_2015.pdf
- CA Final SFM Solve Questions on Bond Valuation by Prof GDTPE140
- SFM Workbook for May 2015
- Session-8D Lean Metrics & Future State VSM.pdf
- University_Risk_Assessment_and_Customs_Update_by_Dave_Zavitz.ppt
- PowerPoint Macros
- PMP 54Formulas
- Reccomendation for Attrition Project
- Accelerating Innovation Using Process Simulation
- DMAIC-kit
- 142 All Accounting Standards
- Hypothesis Testing

Sign up to vote on this title

UsefulNot useful- Juran's Quality Handbook- The Complete Guide to Performance Excellence by Joseph M. Juran- Joseph Defeo- Joseph a. de Feoby Sandeep GM
- Lean Six Sigma for Engineers and Managers - With Applied Case Studies - Matthew John Franchetti (CRC, 2015)by Khai Huynh
- FMEA Handbookby Ernesto Neri
- ASQ Six Sigmaby dmrpanda9940

- Quantitative+Data+Analysis+With+Minitab
- MINITAB User's Guide Chap_11
- Minitab Tutorial
- MINITAB Training
- Design of Experiment Using Minitab Book_1
- Manual Minitab
- Manual Minitab 16 Eng
- GEMS Minitab Handbook 13
- 6. QualityProcessandImprovementAllTopics
- 6Sigma Perparatory Material
- QC minitab.pdf
- Design of Experiments With MINITAB
- Juran's Quality Handbook- The Complete Guide to Performance Excellence by Joseph M. Juran- Joseph Defeo- Joseph a. de Feo
- Lean Six Sigma for Engineers and Managers - With Applied Case Studies - Matthew John Franchetti (CRC, 2015)
- FMEA Handbook
- ASQ Six Sigma
- FMEA
- Quality Control and Realiability_all in One
- Statistical Process Control Sixth Edition - John Oakland
- Minitab Samples
- yan MINITAB
- Digital Electronic Devices Logic Families
- MeetMinitab
- Tutorial Minitab
- Six Sigma Stats Minitab
- SPC
- Introduction to Statistical Process Control
- Control Chart
- Design of Experiments With MINITAB Escrito Por Paul G. Mathews
- Minitab - Quality Control