1 views

Uploaded by exams_sbs

Evolution of quality over time and manufacturing stages discussed

- Chapter4Exercises135789111215171819203234353639444546and50.
- Measures of Central Tendency
- GRE Quant-Distribution
- The stat 43 assignment
- Day 4 - Measures of Central Tendency
- Statistics
- Week 9
- Lesson Plan : Statistics
- Mean Median Mode
- OrginalSchmidt Operating Instructions Multilingual High
- math lesson plan
- Laws of Mean, Median and Mode
- RES341 Team+C Whitner+Autoplex+Project Presentation Final
- lesson 1 high level
- Student 1 PPT MAT Manuscript Science Achievement in Trece Martinez Final Edited
- 1 Sample Sign
- grade 5-6 lesson plan
- 8th_meanmedian and Mode
- FAQs
- Stats_X R. N. Yadav

You are on page 1of 83

University of Trento

nicola.mezzetti@gmail.com

A.A. 2014/2015

”You can’t fake quality any more than you can fake a good meal.”

(W. Burroughs)

Nicola Mezzetti, Ph.D. An Excurse on the Evolution of Quality

Oxford Dictionary

Quality:

1 The standard of something as measured against other things

of a similar kind; the degree of excellence of something.

2 A distinctive attribute or characteristic possessed by someone

or something.

Oxford Dictionary

Quality:

1 The standard of something as measured against other things

of a similar kind; the degree of excellence of something.

2 A distinctive attribute or characteristic possessed by someone

or something.

for us!

Let’s look for some hints in ancient times!

Historians have traced the

concept of quality as far back as

3000 B.C. in Babylonia.

Among the references to quality

from the code of Hammurabi,

ruler of Babylonia, is the

following excerpt: ”The mason

who builds a house which falls

down and kills the inmate shall

be put to death.”

In 1500 B.C. Papyrus, the Egyptian medical, used to describe

precisely how to prepare pharmaceutical drugs.

Archimedes of Syracuse (287 - 212 B.C.) tested the gold

proportion of a votive crown of King Hiero II.

The Roman architect Vitruvius

wrote ”The ten books on

architecture” to describe his

personal experience about the

quality of construction work and

customers’ needs.

Since 13th century craftsmen

began organizing into unions

called guilds to develop

knowledge about production

stages and materials.

Apprenticeship has been

instituted at those times, within

guilds.

The evolution of quality in the 20th century

Identifies reasons for inefficiency:

1 Belief that a material increase in the output of each man or

each machine in the trade would result in throwing a large

number of men out of work

2 Defective systems of management which make it necessary for

each workman to soldier, or work slowly

3 Rule-of-thumb methods

Four principles of scientific management:

1 Develop a science for each element of a man’s work, which

replaces the old rule-of-thumb method.

2 Scientifically select and then train, teach, and develop the

workman, whereas in the past he chose his own work and

trained himself as best he could.

3 Heartily cooperate with the men so as to ensure all of the work

being done in accordance with the principles of the science

which has been developed.

4 Equal division of the work and the responsibility between the

management and the workmen.

Nicola Mezzetti, Ph.D. An Excurse on the Evolution of Quality

Application of Taylor’s scientific management to the

production of Model T.

Allowed to break down the assembly process in 84 steps

The chassis of the car was pulled down the line by a chain

conveyor and then 140 workers applied their assigned parts to

the chassis.

Interchangeable parts were produced in mass quantities and

... brought directly to the workers who were trained to work at

specific assembly stations.

Impact:

The use of interchangeable parts allowed for continuous work

flow and more time on task by laborers.

The production time for a single car dropped from over twelve

hours to 93 minutes.

These concepts allowed Ford to increase his profit margin and

lower the cost of the vehicle to consumers ($260 in 1924).

SQC was pioneered by Walter Shewhart at Bell Laboratories in the

early 1920s.

Shewhart’s principle was that bringing a process into a state

of statistical control would allow the distinction between

assignable and chance cause variations and the prediction of

future output

Assignable cause: the variation is due to

differences in quality of raw material

differences in machines or in operators

difference of time

Chance cause:

the variation is negligible and inevitable

the variation appears randomly and is due to chance and not

to defects in machines or materials or to any other factor

the process is said to be under the state of statistical control

Nicola Mezzetti, Ph.D. An Excurse on the Evolution of Quality

Nicola Mezzetti, Ph.D. An Excurse on the Evolution of Quality

The Pareto principle states that, for many events, roughly 80% of

the effects come from 20% of the causes.

Four step model for carrying out change

Do: test the change.

Check: review the test, analyze the results and identify what

you have learned.

Act: take action based on what you learned in the study

if the change did work, incorporate what you learned from the

test into wider changes

otherwise, go through the cycle again with a different plan

Nicola Mezzetti, Ph.D. An Excurse on the Evolution of Quality

Quality control: an effective system for coordinating the quality

maintenance and quality improvement efforts of the various groups

in an organization so as to enable production at the most

economical levels which allow for full customer satisfaction.

Idea of modern quality control as management-based: Increasing

operator efficiency by educating them on quality in

order to enhance overall quality.

Increasing quality awareness throughout the organization.

Involving the entire organization in each and every quality

initiative undertaken.

Quality control as an administrative effort for continuous

quality improvement:

Providing a channel for knowledge integration and

communication, thus enhancing product quality.

Encouraging employee participation in organizational quality

control initiatives.

Nicola Mezzetti, Ph.D. An Excurse on the Evolution of Quality

At the end of WWII, Toyota had to compete with American

manufacturers in order to survive.

Due to large lot production, an American worker produced

approximately nine times as much as a Japanese worker.

Given the small demand in Japan, Toyota could not rely on the

same methods. Taiichi Ohno had the idea of making profit by

cost reduction.

Absolute elimination of waste

Just-in-time (JIT): getting the number of parts needed, when

they are needed.

Autonomation: automating a process to include inspection.

Multi-skilled workers: one worker assigned to more

machines.

Minimizing transportation: putting the machines of a

process together.

Quality Loss Function

Quality is the loss imparted by L(y ) = k(y − T )2

the product to society from the where:

time the product is shipped. y is the observed value

T is the target value

Taguchi and Wu proposed the k is the cost coefficient

absolute quality conformance

approach encouraging to pay

more attention to upstream

activities (i.e., product design

and planning of processes):

any variation from the exact

specifications entails a cost

or loss to the organization.

Quality as the conformance to the requirements which the company itself

has established for its products based upon the customers needs.

Do It Right First Time: People are not prevented from making

mistakes, but encouraged to improve continuously.

The ultimate goal is to train all the staff and give them the tools for

quality improvement, to apply the basic precept of Prevention

Management in every area.

process. Four principles:

1 Quality is conformance to requirements.

2 Quality prevention is preferable to quality inspection and correction.

3 Standard of quality is always be zero defect, not close enough.

4 Quality is measure on the basis of Price, Price of Non

Conformance (PONC).

Juran is widely credited for adding the human dimension to

quality management.

For Juran, human relations problems are the cause of isolation and

resistance to change and thus the root cause of quality issues.

outlining the issue.

The book established the know-how about capitalizing on the

benefits of quality. It was organized in two parts:

Breakthrough: creation of beneficial change.

Control: prevention of adverse change.

Primary Elements of TQM:

Umbrella methodology based on Customer-focused

principles and practices of:

Total employee involvement

behavioural sciences,

Process-centered

analysis of quantitative and

Integrated system

nonquantitative data,

Strategic and systematic

economics theories,

approach

process analysis

Continual improvement

... to continually improve the

Fact-based decision making

quality of all processes.

Communications

1 Create constancy of purpose for improving products and services.

2 Adopt the new philosophy.

3 Cease dependence on inspection to achieve quality.

4 End the practice of awarding business on price alone; instead, minimize

total cost by working with a single supplier.

5 Improve constantly and forever every process for planning, production and

service.

6 Institute training on the job.

7 Adopt and institute leadership.

8 Drive out fear.

9 Break down barriers between staff areas.

10 Eliminate slogans, exhortations and targets for the workforce.

11 Eliminate numerical quotas for the workforce and numerical goals for

management.

12 Remove barriers that rob people of pride of workmanship, and eliminate

the annual rating or merit system.

13 Institute a vigorous program of education and self-improvement for

everyone.

14 Put everybody in the company to work accomplishing the transformation.

Six Sigma is...

A strategy to improve process quality by identifying and eliminating

defects and minimizing variation in process outputs.

A data driven approach based on measurement of the process

variation using Statistical Process Control.

A structured implementation approach based on a DMAIC cycle and

certified experts.

Nicola Mezzetti, Ph.D. An Excurse on the Evolution of Quality

The ISO 9000 standard family comprises:

ISO 9000:2005 Quality management systems: Fundamentals

and vocabulary

Eight Principles of Quality Management

ISO 9001:2008 Quality management systems

Requirements of a quality management system

ISO 9004:2009 Managing for the sustained success of the

organization - A quality management approach

ISO 19011:2011 Guidelines for quality and/or environmental

management systems auditing

Eight Principles of Quality Management:

1 Customer focus

2 Leadership

3 Involvement of people

4 Process approach

5 System approach to management

6 Continual improvement

7 Factual approach to decision-making

8 Mutually beneficial supplier relationships

Lean

A systematic approach to identifying and eliminating waste

through continuous improvement, flowing the product at the pull

of the customer in pursuit of perfection.

Eight Wastes: Some Lean Building Blocks:

Waiting Work Cells

Transportation Total Productive Maintenance

Non-Value-Added-Processing Total Quality Management

Excess Inventory Point-Of-Use-Storage

Defects Batch Size Reduction

Excess Motion Visual Controls

Underutilized People Concurrent Engineering

Nicola Mezzetti, Ph.D. An Excurse on the Evolution of Quality

1 Performance improvement

Scientific Management

Statistical Process Control

Total Quality Control

2 Cost reduction (since 1950s)

Toyota Production System

Quality Loss Function

Zero Defects

3 People Management (since 1960s)

Managerial Breakthrough

Total Quality Management

ISO 9000:2005

ISO defines Corporate Social

Responsibility (CSR) as follows:

organizations to address economic,

social and environmental issues in a

way that aims to benefit people,

communities and society.

The article K3 of the Treaty on

European Union (1998) obliges all

the member states to adopt

anti-corruption measures.

”The total composite product and service characteristics of

marketing, engineering, manufacture, and maintenance through

which the product and service will meet the expectations of the

customer” (A. Feigenbaum, 1961)

”Quality is the loss imparted by the product to society from the

time the product is shipped.” (G. Taguchi, 1962)

”Conformance to requirements” (P. Crosby, 1979)

”The degree to which a set of inherent characteristics fulfills

requirements” (ISO 9000:2005, 2005)

”Fitness for purpose” (J. Juran, 2010)

”Quality is a dynamic state associated with products, services,

people, processes, and environments that meets or exceeds

expectations and helps produce superior value” (D. Goetsch and

S. Davis, 2010)

Building and managing organizations capable of

meeting customer requirements

protecting the business reputation by identifying and

eliminating defects

reducing waste in production

understanding what’s important for the customer (current

and future needs)

rapidly adapting in order to provide flexible and fast

responses to market opportunities

... and of continuously improving itself.

In 1990s Organizational Excellence has been proposed as an

alternative to Quality Management. The main reasons:

Quality Management has no theoretical foundations;

There is quite a confusion over the very definition of

”Quality”;

It remains unclear whether quality management is simply:

a collection of essentially independent techniques;

a management philosophy;

a coherent management methodology;

a strategy;

a theory for managing only the quality of services and process;

a master theory for managing the entire enterprise.

Quality identified as an end in itself rather than a means of

contributing to enterprise success.

Adding value for customers;

Creating a sustainable future;

Developing organizational capability;

Harnessing creativity & innovation;

Leading with vision, inspiration and integrity;

Managing with agility;

Succeeding through the talent of people;

Sustaining outstanding results.

Adding value for customers ⇒ Customer Orientation;

Creating a sustainable future ⇒ Corporate Social Responsibility

Developing organizational capability ⇒ Process approach & System

Approach to Management;

Harnessing creativity & innovation ⇒ Continuous quality

improvement;

Leading with vision, inspiration and integrity ⇒ Leadership;

Managing with agility ⇒ Breakthrough;

Succeeding through the talent of people ⇒ Total employee

involvement;

Sustaining outstanding results ⇒ Continuous quality improvement

& Zero Defects.

Quality Engineering is a technology that deals with the

improvement of the organizations’ performance in understanding

and meeting customers expectations.

gr. TECHNOLOGÍA from TÉCHNÊart and LOGÍA for LÒGOS

discourse, treatise.

Propr. Exposure of rules for dealing with art and scientifically a

subject; Doctrine on immediate application of physics, chemistry

and mathematics to arts and crafts, so that that the architect may

not act against the real scientific principles.

”Quality is not an act, it is a habit.” (Aristotle)

”It is most important that top management be quality-minded. In

the absence of sincere manifestation of interest at the top, little

will happen below.” (J. Juran)

”Quality is everyone’s responsibility.” (W. Deming)

”Quality is when your customers come back and your products

don’t.” (E. Naumann, P. Shannon)

Basic Statistical Tests Click here to start

Basic Statistical Tests

Compare groups

Focus of test

with each other

Data type

Compare one

group with

a target

Variables data Attribute data

Data type

Continuous

Attribute data

data

Focus of Type of

comparison attribute data

Center of the Spread of the

distribution distribution (standard

(mean or median) deviation or variance)

Binary data Count data

(yes/no) (occurrences)

Data

distribution

Normal Nonnormal

distribution distribution

Shape of

distribution

1-Sample 1-Sample

1-Sample t 1-Sample Sign 1 Variance 1 Proportion

Wilcoxon Poisson Rate

Compare groups

with each other

groups than two groups

Compare two

groups

Data type

data

sampled

Paired

Independent

(dependent)

samples

observations

Type of

Focus of attribute

comparison data

(mean or median) (standard deviation (yes/no) (occurrences)

or variance)

Data

distribution

Normal Nonnormal

distribution distribution

2-Sample

2-Sample t Mann-Whitney 2 Variances Paired t 2 Proportions

Poisson Rate

Compare more

than two groups

Data type

Focus of

comparison

distribution distribution (standard

(mean or median) deviation or variance)

Data

distribution

Normal Nonnormal

distribution distribution

Kruskal-Wallis and Chi-Square

ANOVA Variances

Do you want to compare

one group with a target or Compare one group with a target Compare groups with each other

groups with each other? Compare a characteristic value of a group, Compare two or more groups or the same

such as a mean or standard deviation, with a group under different conditions. Often used

target value. Often used to evaluate whether a to evaluate whether a process is affected by

process meets a performance goal. different factors, changes, or conditions.

Example Example

Focus A quality analyst wants to determine whether A sales analyst tracks customer purchases

of test the mean fill weight of a cereal product differs at various store locations. The analyst wants

from value stated on the package label (500 to determine whether the average cost of a

g). The analyst weighs a sample of cereal customer purchase differs between locations.

Compare one Compare boxes from a single production line.

group with a groups with

target each other

To estimate a characteristic value of a group, you must analyze a representative sample of data from

the group. Then, you can use hypothesis tests to determine whether that value differs from a target or

from another group. A target value may be an industry standard, a past performance level, or a process

specification.

Click a shape to move on any page to Basic Statistical Tests

through the decision tree return to Start.

Do you have continuous

data or attribute data? Continuous data Attribute data

Measures a characteristic of a part or process, Counts the presence of a characteristic or

such as length, weight, or temperature. The condition, such as a physical trait, a type of

data often includes fractional (or decimal) defect, or a rating, such as pass/fail. The data

values. are whole numbers.

Focus of test:

Compare one group Example Example

with a target A quality analyst wants to determine whether Managers at a bank want to know whether the

the mean fill weight differs from the value proportion of customers who are interested

stated on the package label (500 g). The in student loans is high enough (at least 5%)

analyst weighs a sample of cereal boxes from to warrant offering the service to customers.

Data type a single production shift. They survey 3500 customers and count how

many of them are interested in student loans.

Continuous Attribute

data data

If possible, collect continuous data because they provide more detailed information. However,

sometimes attribute data adequately describe the quality of a part or a process. For example, if you

track broken light bulbs, you don’t need to measure a characteristic of the bulb to evaluate whether

it’s broken or not. What matters is only the number of bulbs that are broken (counts).

Do you want to evaluate

the center or the spread of Center of the distribution Spread of the distribution

the distribution? Determines whether the central value of the Determines whether the variation of the data

data equals a target value, such as a process equals a target value, such as an industry

specification. Often used to evaluate how standard. Often used to evaluate how

accurate a process is. consistent a process is.

Focus of test:

Compare one group Example Example

with a target An inspector measures a shipment of wheel The manager of a lumberyard wants to

rims and records their diameters. The inspector analyze how consistently a sawmill produces

Data type:

Continuous

evaluates whether the mean diameter of beams that are on average 100 cm long. The

the rims equals the specification diameter of manager measures the beams and analyzes

16 in. whether the standard deviation of the lengths

is less than 1 mm.

Focus of

comparison

distribution distribution

You can evaluate the central value of the data using the mean (the arithmetic average) or the median

(the middle value when the data are arranged from lowest to highest). The median is less sensitive

than the mean to outliers.

You can evaluate the variation, or spread, of the data using the standard deviation or the variance.

The standard deviation is often easier to interpret because it uses the same units as the data.

Does your data follow

a normal or nonnormal Normal distribution Nonnormal distribution

distribution? Data that follow a symmetric, bell-shaped Data that follow a distribution that either is

distribution. not symmetric or is symmetric but not bell-

shaped.

Example

Focus of test: An auto parts company manufactures wheel Example

Compare one group rims to be 16 inches in diameter. Inspectors A shipping company guarantees delivery of

with a target randomly sample wheel rims and measure domestic packages within 72 hours. Inspectors

their diameters. They determine that the randomly sample orders and record the delivery

Data type:

Continuous

measurements follow a normal distribution. times. They determine that the times are not

symmetrically distributed and, therefore, are

Focus of comparison: not normal.

Center of the distribution

Data

distribution

Normal Nonnormal

distribution distribution

The distribution of your data often depends on your process. For example, data that track cycle

time for service processes, such as the time needed to process an application, serve a customer, or

deliver a product, often do not follow a normal distribution.

Although the normal distribution is always bell-shaped, not all bell-shaped distributions are normal. To

determine whether your data follow a normal distribution, use normal probability plots, histograms, or

normality tests such as the Anderson-Darling test. You can perform these evaluations using Minitab’s

Normality Test or Graphical Summary.

If you have 30 or more observations, you can generally treat your data as being normal and perform

a basic statistical test to compare the mean of the group with a target value.

Basic Statistical Tests 1-Sample t

1-Sample t

The 1-sample t-test determines whether the mean differs from a target value and provides a range of

values that is likely to include the true mean.

Example

Focus of test: An inspector samples a shipment of wheel rims and measures their diameters. The inspector uses a

Compare one group 1-sample t-test to determine whether the mean diameter of the rims differs from the specification of

with a target 16 in.

Data type: To perform a 1-sample t-test in Minitab, choose Stat > Basic Statistics > 1-Sample t.

Continuous

One-Sample T: Diameter

Focus of comparison:

Center of the distribution Test of mu = 16 vs not = 16

Data distribution: Diameter 20 15.9992 0.0034 0.0008 (15.9976, 16.0007) -1.12 0.278

Normal

1-Sample t

Use Minitab’s 1-sample t-test to evaluate whether the mean is greater than, less than, or not equal to

a target value. If you choose a one-sided test (less than or greater than), the test has more power to

detect a difference in the direction of interest. However, it cannot detect a difference in the opposite

direction.

For example, you want to know whether the mean diameter of the wheel rims differs from 16 in. You

could increase the power of the test by testing only whether the mean is greater than 16 in. However,

this test does not detect a difference if the mean is less than 16 in.

Typically, you use a 1-sample t-test to evaluate continuous data that are normally distributed. You

can also use a 1-sample t-test to evaluate data that is nonnormal but meets one of these conditions:

has 30 or more observations, is symmetric with at least 5 observations, or has relatively few extreme

outliers.To quickly evaluate normality and outliers, use Minitab’s Graphical Summary. If your data

contains any extreme outliers, make sure they are valid measurements.

Use Power and Sample Size for 1-sample t to determine how much data you need to detect an

important difference between the mean and the target value.

Is the data symmetric or

not symmetric? Symmetric Not symmetric

The data are distributed evenly on both sides The data are not distributed evenly on both

of the central value. sides of the central value and are “skewed” in

one direction.

Example

Focus of test: A pharmaceutical company wants to Example

Compare one group determine whether a newly developed drug A stainless steel manufacturer wants to

with a target relieves symptoms in less than 12 minutes. A determine whether the percentage of

researcher administers the drug to 16 patients chromium in its products equals 18%. An

Data type:

Continuous

and records the time elapsed until symptoms inspector measures the chromium content in

abate. Although the data are not normal, they a sample of products and determines that the

Focus of comparison: are symmetrically distributed. data do not follow a symmetric distribution.

Center of the distribution

Data distribution:

Nonnormal

Shape of

distribution

To evaluate symmetry, display your data in a histogram or boxplot to determine whether it is similarly

distributed on both sides of the central data value. If the data are symmetric, the mean and the

median are approximately equal, so you can use either the mean or the median to describe the

center of the data.

You can also use Minitab’s Symmetry Plot to evaluate the symmetry of your data. Choose Stat >

Quality Tools > Symmetry Plot.

Basic Statistical Tests 1-Sample Wilcoxon

1-Sample Wilcoxon

The 1-sample Wilcoxon test determines whether the median differs from a target value and provides a

range of values that is likely to include the true median.

Example

Focus of test: A researcher at a pharmaceutical company administers a newly developed drug to 16 patients and records

Compare one group the time elapsed until symptoms abate. The data do not follow a normal distribution but are symmetric.

with a target The researcher uses a 1-sample Wilcoxon test to determine whether the drug relieves symptoms in less

than 12 minutes.

Data type:

Continuous To perform a 1-sample Wilcoxon test in Minitab, choose Stat > Nonparametrics > 1-Sample

Wilcoxon.

Focus of comparison:

Center of the distribution Wilcoxon Signed Rank Test: Time

Nonnormal N for Wilcoxon Estimated

N Test Statistic P Median

Shape of distribution: Time 15 15 19.0 0.011 9.350

Symmetric

Wilcoxon Signed Rank CI: Time

Confidence

1-Sample Wilcoxon Estimated Achieved Interval

N Median Confidence Lower Upper

Time 15 9.35 95.0 7.30 11.40

The median is the middle value of the data when they are arranged from lowest to highest. The

median is less sensitive to outliers than the mean.

You can determine whether the median is greater than, less than, or not equal to a target value. If you

choose a one-sided test (less than or greater than), the test has more power to detect a difference in

the direction of interest. However, it cannot detect a difference in the opposite direction.

You can also use a 1-sample t-test to evaluate data that is nonnormal but meets one of these

conditions: has 30 or more observations, is symmetric with at least 5 observations, or has relatively

few extreme outliers. To quickly evaluate normality and outliers, use Minitab’s Graphical Summary. If

your data contains any extreme outliers, make sure they are valid measurements.

Basic Statistical Tests 1-Sample Sign

1-Sample Sign

The 1-sample sign test determines whether the median differs from a target value and provides a range

of values that is likely to include the true median.

Example

Focus of test: An inspector for a stainless steel manufacturer measures the chromium content in 12 product samples

Compare one group and determines that the data do not follow a symmetric distribution. The inspector uses a 1-sample sign

with a target test to determine whether the median chromium content differs from the specification of 18%.

Data type: To perform a 1-sample sign test in Minitab, choose Stat > Nonparametrics > 1-Sample Sign.

Continuous

Sign Test for Median: %Chromium

Focus of comparison:

Center of the distribution Sign test of median = 18.00 versus not = 18.00

Data distribution:

%Chromium 12 8 0 4 0.3877 17.70

Nonnormal

Sign CI: %Chromium

Shape of distribution:

Not symmetric Sign confidence interval for median

Confidence

Achieved Interval

1-Sample Sign N Median Confidence Lower Upper Position

%Chromium 12 17.70 0.8540 17.50 18.10 4

0.9500 17.43 18.76 NLI

0.9614 17.40 19.00 3

The median is the middle value of the data when they are arranged from lowest to highest. The

median is less sensitive to outliers than the mean and can be a more informative measure of the

center when the data do not follow a symmetric distribution.

You can evaluate whether the median is greater than, less than, or not equal to a target value. If you

choose a one-sided test (less than or greater than), the test has more power to detect a difference in

the direction of interest. However, it cannot detect a difference in the opposite direction.

You can also use a 1-sample t-test to evaluate data that is nonnormal but meets one of these

conditions: has 30 or more observations, is symmetric with at least 5 observations, or has relatively

few extreme outliers. To quickly evaluate normality and outliers, use Minitab’s Graphical Summary. If

your data contains any extreme outliers, make sure they are valid measurements.

Basic Statistical Tests 1 Variance

1 Variance

The 1 variance test determines whether the variation (standard deviation or variance) differs from a

target value and provides a range of values that is likely to include the true variation.

Example

Focus of test: A manager of a lumberyard measures the length of 50 beams cut by a sawmill. The manager uses a 1

Compare one group variance test to evaluate whether the standard deviation of the beam lengths is less than 1 mm.

with a target

To perform a 1 variance test in Minitab, choose Stat > Basic Statistics > 1 Variance.

Data type:

Continuous Test and CI for One Standard Deviation: Length

Method

Focus of comparison:

Spread of the distribution Null hypothesis Sigma = 1

Alternative hypothesis Sigma not = 1

The adjusted method is for any continuous distribution.

Statistics

Length 50 0.871 0.759

CI for

Variable Method CI for StDev Variance

Length Standard (0.728, 1.085) (0.529, 1.178)

Adjusted (0.712, 1.122) (0.507, 1.259)

Tests

Length Standard 37.17 49.00 0.215

Adjusted 28.89 38.09 0.282

You can use Minitab’s 1 Variance test on normal or nonnormal continuous data.

You can evaluate whether the standard deviation (or variance) is greater than, less than, or not equal

to a target value. If you choose a one-sided test (less than or greater than), the test has more power

to detect a difference in the direction of interest. However, it cannot detect a difference in the opposite

direction.

Do you have binary data

or count data? Binary data Count data

Classifies items into one of two categories, Counts the presence of a characteristic,

such as pass/fail or yes/no. Often used to result, or activity. Often used to compare an

compare a proportion. occurrence rate.

Compare one group A direct-mail firm wants to track the proportion Inspectors at a bus company count the number

with a target of customers who respond to a direct-mail of bus breakdowns each day for 30 days. The

advertisement by purchasing the advertised company wants to determine the daily rate of

Data type:

Attribute

item. Marketing analysts randomly sample bus breakdowns.

1000 customers who received the mailer and

record whether or not the customer bought the

item.

Type of

attribute

data

For count data, you count the number of occurrences in a given amount of time, area, volume, or other

observation space. If a process has a constant rate of occurrence, use data from any observation

space to make an inference about the entire process, such as the number of defects per day, per

month, or per year. Otherwise, make sure that the observation space over which you collect the data

is appropriate for the question you want to answer.

Suppose the rate of phone calls at a call center varies greatly each hour. If you track only the calls

received between 8 am and 9 am, the rate will not reflect the average hourly rate over the entire day.

Conversely, if you track the calls over the entire day, the average hourly rate will not reflect the rate

of calls received between 8 am and 9 am.

Basic Statistical Tests 1 Proportion

1 Proportion

A 1 proportion test determines whether a proportion differs from a target value and provides a range of

values that is likely to include the true proportion.

Example

Focus of test: A marketing analyst at a direct-mail firm randomly samples 1000 customers who received an advertising

Compare one group mailer and records whether or not the customer bought the item. The analyst uses a 1 proportion test

with a target to determine whether the proportion of customers who respond to the mailer differs from the national

average of 6.5%.

Data type:

Attribute To perform a 1 proportion test in Minitab, choose Stat > Basic Statistics > 1 Proportion.

Binary

Test of p = 0.065 vs p not = 0.065

1 Proportion Exact

Sample X N Sample p 95% CI P-Value

1 87 1000 0.087000 (0.070268, 0.106208) 0.008

You can evaluate whether the proportion is greater than, less than, or not equal to the target value. If

you choose a one-sided test (less than or greater than), the test has more power to detect a difference

in the direction of interest. However, it cannot detect a difference in the opposite direction.

Suppose you want to know whether the proportion of customers who respond to the mailer differs

from 6.5%. You could increase the power of the test by testing only whether the proportion is greater

than 6.5%. However, this test does not detect a difference if the proportion is less than 6.5%.

Use Power and Sample Size for 1 proportion to determine how much data you need to detect an

important difference between the proportion and the target value.

Basic Statistical Tests 1-Sample Poisson Rate

1-Sample Poisson Rate

A 1-sample Poisson rate determines whether a rate differs from a target value and provides a range of

values that is likely to include the true rate.

Example

Focus of test: Inspectors at a bus company count the number of bus breakdowns each day for 30 days. The inspectors

Compare one group use a 1-sample Poisson rate to determine the daily rate of bus breakdowns and see whether it differs

with a target from the daily breakdown rate from the previous year (2.1 buses/day).

Data type: To perform a 1-sample Poisson rate in Minitab, choose Stat > Basic Statistics > 1-Sample Poisson

Attribute Rate.

Count

Test of rate = 2.1 vs rate not = 2.1

1-Sample Poisson Rate Sample Occurrences N Occurrence 95% CI P-Value

1 94 30 3.13333 (2.53205, 3.83441) 0.000

“Length” of observation = 1.

You can evaluate whether the rate is greater than, less than, or not equal to the target value. If you

choose a one-sided test (less than or greater than), the test has more power to detect a difference in

the direction of interest. However, it cannot detect a difference in the opposite direction.

Suppose you want to know whether the rate of bus breakdowns differs from 2.1. You could increase

the power of the test by testing only whether the daily rate of bus breakdowns is less than 2.1 per day.

However, this test does not detect a difference if the rate is greater than 2.1 per day.

How many groups are you

comparing? Two groups More than two groups

Compare two groups or the same group under Compare more than two groups or the same

two conditions. group under more than two conditions.

Example Example

Focus of test: Hospital researchers want to compare post- A quality engineer wants to compare the mean

Compare groups surgery recovery times in two groups of strength of nylon fiber from 5 different suppliers

with each other patients: patients who receive follow-up care to see whether the strength of the nylon differs

at the hospital and patients who receive follow- between the suppliers.

up care at home.

Number

of groups

Two groups

groups

Do you have continuous

data or attribute data? Continuous data Attribute data

Measures a characteristic of a part or process, Counts the presence of a characteristic or

such as length, weight, or temperature. The condition, such as a physical trait, a type of

data often includes fractional (or decimal) defect, or a rating, such as pass/fail. The data

values. are whole numbers.

Focus of test:

Compare groups Example Example

with each other A quality engineer wants to evaluate whether An analyst wants to compare the quality of

the mean weight of cereal boxes is the same light bulbs from two different suppliers. The

Number of groups: across two production shifts. The analyst analyst inspects a sample of bulbs from each

Two samples boxes from each shift and records supplier and counts the number of broken

their weights. bulbs in each sample.

Data type

Continuous Attribute

data data

If possible, collect continuous data because they provide more detailed information. However,

sometimes attribute data adequately describe the quality of a part or a process. For example, if you

track broken light bulbs, you don’t need to measure a characteristic of the bulb to evaluate whether

it’s broken or not. What matters is only the number of bulbs that are broken (counts).

Do you have independent

samples or paired Independent samples Paired observations

(dependent) observations? Each sample is selected randomly, so the The sample is a set of matched pairs, so that

observations in one sample do not affect the one observation is dependent on (related to)

observations in the other sample. the other observation in the pair.

Compare groups A health management company wants to To evaluate a computer training course, a

with each other compare satisfaction ratings from former human resource analyst tests the computer

patients at two hospitals. Because the two skills of each trainee before and after the

Number of groups: patient groups are composed of different training. Therefore, each post-training test

Two individuals who are randomly selected, the score is associated (paired) with a pre-training

samples are independent. test score from the same trainee.

Data type:

Continuous

sampled

Independent Paired

samples observations

• Randomly sample items or subjects from two groups (populations). For example, a quality

engineer takes a random sample of bolts that are produced by different machines. The bolts

from each machine are considered independent samples.

• Randomly split a sample into groups and apply distinct conditions to each group. For example,

an analyst randomly assigns a sample of customers to two groups and sends each group a

different promotional email. The analyst records whether each customer responds to the email.

Paired (dependent) observations occur when you:

• Measure a characteristic of the same subject under different conditions. For example, a

researcher measures cholesterol in the same patients before and after they follow a low-fat diet.

• Apply the same conditions to two items or subjects. For example, a quality engineer measures

the tread wear in two brands of tires that were put on the same car. The tread wear for the two

tires is likely to depend on the car.

Do you want to compare

the center or the spread Center of distribution Spread of distribution

of each distribution? Compares the central value in the Compares the variation in the measurements

measurements for each group. You want to for each group. You want to determine whether

determine whether the groups differ. the data vary more in one group than in the

Example other.

Focus of test:

Compare groups A quality analyst at a health management Example

with each other company records satisfaction ratings from A manager of a lumberyard measures the

former patients at two hospitals. The analyst length of beams that are cut by two different

Number of groups: wants to compare the average satisfaction sawmills. The manager wants to compare the

Two ratings at each hospital to determine whether spread of the measurements for each sawmill

they differ. to determine whether the consistency of the

Data type: beam lengths is the same.

Continuous

Independent samples

Focus of

comparison

Center of Spread of

distribution distribution

You can evaluate the center of the data using the mean (the arithmetic average) or the median (the

middle value when the data are arranged from lowest to highest). The median is less sensitive than

the mean to outliers.

You can evaluate the variation, or spread, of the data using the standard deviation or the variance.

The standard deviation is often easier to interpret because it uses the same units as the data.

Does your data follow

a normal or nonnormal Normal distribution Nonnormal distribution

distribution? Data that follow a symmetric, bell-shaped Data that follow a distribution that either is

distribution. not symmetric or is symmetric but not bell-

shaped.

Example

Focus of test: A quality analyst at a health management Example

Compare groups company records the patient satisfaction A public transportation department uses two

with each other ratings at two hospitals. The analyst brands of paint for road stripes. For each

determines that the satisfaction ratings for paint brand, an analyst records the number of

Number of groups: each hospital follow a normal distribution. months that stripes last on the highway. The

Two analyst determines that the durations for each

paint brand are not symmetrically distributed

Data type: and, therefore, are not normal.

Continuous

Independent samples

Focus of comparison:

Center of distribution

Data

distribution

The distribution of your data often depends on your process. For example, data that track cycle

time for service processes, such as the time needed to process an application, serve a customer, or

Normal Nonnormal

distribution distribution

deliver a product, often do not follow a normal distribution.

Although the normal distribution is always bell-shaped, not all bell-shaped distributions are normal. To

determine whether your data follow a normal distribution, use normal probability plots, histograms, or

normality tests such as the Anderson-Darling test. You can perform these evaluations using Minitab’s

Normality Test or Graphical Summary.

If you have 30 or more observations in each sample, you can generally treat your data as being

normal and perform a 2-sample t-test to compare the means of two groups.

Basic Statistical Tests 2-Sample t

2-Sample t

The 2-sample t-test determines whether the means of two groups differ and provides a range of values

that is likely to include the true difference between the means.

Example

Focus of test: A quality analyst at the health management company wants to compare patient satisfaction ratings at

Compare groups two hospitals. The analyst uses a 2-sample t-test to evaluate whether the average satisfaction ratings

with each other from former patients differ between hospitals.

To perform a 2-sample t-test in Minitab, choose Stat > Basic Statistics > 2-Sample t.

Number of groups:

Two

Two-Sample T-Test and CI: A, B

Two-sample T for A vs B

Data type:

Continuous N Mean StDev SE Mean

A 10 80.7 10.6 3.4

B 15 59.0 14.2 3.7

How are data sampled:

Difference = mu (A) - mu (B)

Independent samples

Estimate for difference: 21.70

95% CI for difference: (10.79, 32.61)

T-Test of difference = 0 (vs not =): T-Value = 4.11 P-Value = 0.000 DF = 23

Focus of comparison: Both use Pooled StDev = 12.9213

Center of distribution

Data distribution:

Use Minitab’s 2-sample t-test to evaluate whether the mean of one group is greater than, less than, or

Normal

not equal to the mean of the other group. If you choose a one-sided test (less than or greater than),

the test has more power to detect a difference in the direction of interest. However, it cannot detect a

2-Sample t difference in the opposite direction.

Typically, you use a 2-sample t-test to evaluate continuous data that are normally distributed. You

can also use a 2-sample t-test to evaluate data that is nonnormal but meets one of these conditions:

has 30 or more observations, is symmetric with at least 5 observations, or has relatively few extreme

outliers.To quickly evaluate normality and outliers, use Minitab’s Graphical Summary. If your data

contains any extreme outliers, make sure they are valid measurements.

Use Power and Sample Size for 2-sample t to determine how much data you need to detect an

important difference between the means of the two groups.

Basic Statistical Tests Mann-Whitney

Mann-Whitney

The Mann-Whitney test determines whether the medians of two groups differ and provides a range of

values that is likely to include the true difference between the medians.

Example

Focus of test: A public transportation department uses two brands of paint for road stripes. An analyst records the

Compare groups number of months that each brand of paint lasts on the highway. Because the samples are relatively

with each other small and the data is not normal, the analyst uses a Mann-Whitney to determine whether the duration of

the two paints differs.

Number of groups: To perform a Mann-Whitney test in Minitab, choose Stat > Nonparametrics > Mann-Whitney.

Two

Data type:

N Median

Continuous

Brand A 11 36.00

Brand B 10 37.50

Independent samples 95.5 Percent CI for ETA1-ETA2 is (-3.30,0.20)

W = 94.5

Test of ETA1 = ETA2 vs ETA1 not = ETA2 is significant at 0.0671

The test is significant at 0.0668 (adjusted for ties)

Focus of comparison:

Center of distribution

Data distribution: The median is the middle value of the data when they are arranged from lowest to highest. The

Nonnormal median is less sensitive to outliers than the mean and can be be a more informative measure of the

center when the data do not follow a symmetric distribution.

Mann-Whitney You can evaluate whether the median of one group is greater than, less than, or not equal to the

median of the other group. If you choose a one-sided test (less than or greater than), the test has

more power to detect a difference in the direction of interest. However, it cannot detect a difference

in the opposite direction.

For the Mann-Whitney test, both groups should have approximately the same variance. To compare

the variances of the two groups, use Minitab’s 2 Variances.

You can also use the 2-sample t-test to evaluate data that is nonnormal but meets one of these

conditions: has 30 or more observations, is symmetric with at least 5 observations, or has relatively

few extreme outliers.To quickly evaluate normality and outliers, use Minitab’s Graphical Summary. If

your data contains any extreme outliers, make sure they are valid measurements.

Basic Statistical Tests 2 Variances

2 Variances

The 2 variances test determines whether the variances (or standard deviations) of two groups differ and

provides a range of values that is likely to include the true variation in each group.

Example

Focus of test: A manager of a lumberyard wants to compare the consistency of the length of beams cut by two different

Compare groups sawmills. The manager uses a 2 variances test to determine whether the variation of the beam lengths

with each other differs between the sawmills.

To perform a 2 variances test in Minitab, choose Stat > Basic Statistics > 2 Variances.

Number of groups:

Two

Data type:

Continuous

Independent samples

Focus of comparison:

Spread of distribution

2 Variances

Basic Statistical Tests Paired t

Paired t

The paired t-test examines the mean difference between paired observations and provides a range of

likely values for the difference.

Example

Focus of test: A human resource analyst evaluates the computer skills of each trainee before and after training. The

Compare groups analyst uses a paired t-test to determine whether the mean test scores before and after the training

with each other differ.

To perform a paired t-test in Minitab, choose Stat > Basic Statistics > Paired t.

Number of groups:

Two Paired T-Test and CI: Before, After

Data type:

Continuous N Mean StDev SE Mean

Before 15 73.20 15.73 4.06

After 15 83.47 10.23 2.64

Difference 15 -10.27 8.52 2.20

How are data sampled:

Paired observations 95% CI for mean difference: (-14.99, -5.55)

T-Test of mean difference = 0 (vs not = 0): T-Value = -4.67 P-Value = 0.000

Paired t

Typically, you use a paired t-test to evaluate the mean differences between paired observations when

the differences follow a normal distribution.

You can also use a paired t-test to evaluate paired data that is nonnormal but meets one of these

conditions: has 30 or more observations, is symmetric with at least 5 observations, or has relatively

few extreme outliers.To quickly evaluate normality and outliers, use Minitab’s Graphical Summary. If

your data contains any extreme outliers, make sure they are valid measurements.

If the paired differences are nonnormal and contain many extreme outliers, use the 1-sample

Wilcoxon test (for symmetric data) or the 1-sample sign (for nonsymmetric data) to evaluate the

paired differences.

Do you have binary data

or count data? Binary data Count data

Classifies items into one of two categories, Counts the presence of a characteristic,

such as pass/fail or yes/no. Often used to response, or activity. Often used to compare

compare proportions. occurrence rates.

Compare groups To evaluate a new delivery process, a delivery An automotive manufacturer wants to compare

with each other company counts the number of mishandled the number of scratches on each car door that

packages per day before and after it implements is produced by two production lines. Because

Number of groups: the new process. Because the total number of the car doors have the same surface area,

Two packages handled per day varies, proportions inspectors can easily compare the production

provide a better way to compare the number lines by comparing the rate of scratches per

Data type: of mishandled packages. door.

Attribute

Type of

attribute

data

To collect count data, you count the number of occurrences in a given amount of time, area, volume,

or other observation space. If a process has a constant rate of occurrence, use data from any

observation space to make an inference about the entire process, such as the number of defects per

day, per month, or per year. Otherwise, make sure that the observation space over which you collect

the data is appropriate for the question you want to answer.

Suppose the rate of phone calls at a call center varies greatly each hour. If you track only the calls

received between 8 am and 9 am, the rate will not reflect the average hourly rate over the entire day.

Conversely, if you track the calls over the entire day, the average hourly rate will not reflect the rate

of calls received between 8 am and 9 am.

Basic Statistical Tests 2 Proportions

2 Proportions

The 2 proportions test determines whether the proportions of two groups differ and provides a range of

values that is likely to include the true difference.

Example

Focus of test: A delivery company tracks how many packages are mishandled each day before and after it implements

Compare groups a new delivery process. A quality analyst uses a 2 proportions test to determine whether the proportion

with each other of mishandled packages under the old process differs from the proportion of mishandled packages under

the new process.

Number of groups: To perform a 2 proportions test in Minitab, choose Stat > Basic Statistics > 2 Proportions.

Two

Data type:

Attribute Sample X N Sample p

1 18 867 0.020761

2 11 794 0.013854

Type of attribute data: Difference = p (1) - p (2)

Binary Estimate for difference: 0.00690734

95% CI for difference: (-0.00558971, 0.0194044)

Test for difference = 0 (vs not = 0): Z = 1.08 P-Value = 0.279

2 Proportions Fisher’s exact test: P-Value = 0.349

You can evaluate whether one proportion is greater than, less than, or not equal to the other

proportion. If you choose a one-sided test (less than or greater than), the test has more power to

detect a difference in the direction of interest. However, it cannot detect a difference in the opposite

direction.

Suppose you want to know whether a process change affects the proportion of mishandled packages.

You could increase the power of the test by testing only whether the proportion for the new process

is less than the proportion for the old process. However, this test will not detect a difference if the

proportion for the new process is greater than the proportion for the old process.

Use Power and Sample Size for 2 Proportions to determine how much data you need to detect an

important difference between two proportions.

Basic Statistical Tests 2-Sample Poisson Rate

2-Sample Poisson Rate

The 2-sample Poisson rate test determines whether the rates of two groups differ and provides a range

of values that is likely to include the true difference.

Example

Focus of test: An inspector for an automotive manufacturer wants to compare the number of scratches per car door

Compare groups for two production lines. The inspector uses a 2-sample Poisson rate to evaluate whether the rate of

with each other scratches per door is the same for both production lines.

To perform a 2-sample Poisson rate test in Minitab, choose Stat > Basic Statistics > 2-Sample Poisson

Number of groups: Rate.

Two

Data type: Total Rate of

Attribute Variable Occurrences N Occurrence

Line A 6 25 0.24

Line B 16 25 0.64

Type of attribute data:

Count Difference = rate(Line A) - rate(Line B)

Estimate for difference: -0.4

95% CI for difference: (-0.767722, -0.0322782)

Test for difference = 0 (vs not = 0): Z = -2.13 P-Value = 0.033

2-Sample Poisson Rate

Exact Test: P-Value = 0.052

* NOTE * The normal approximation may be inaccurate for small total number of occurrences.

You can evaluate whether one rate is greater than, less than, or not equal to the other rate. If you

choose a one-sided test (less than or greater than), the test has more power to detect a difference in

the direction of interest. However, it cannot detect a difference in the opposite direction.

Suppose you want to detect whether a process change affects the rate of scratches per door. You

could increase the power of the test by testing only whether the rate for the new process is lower

than the rate for the old process. However, this test will not detect a difference if the rate for the new

process is higher than the rate for the old process.

Do you have continuous

data or attribute data? Continuous data Attribute data

Measures a characteristic of a part or process, Counts the presence of a characteristic or

such as length, weight, or temperature. The condition, such as a physical trait, a type of

data often includes fractional (or decimal) defect, or a rating, such as pass/fail. The data

values. are whole numbers.

Focus of test:

Compare groups Example Example

with each other A quality engineer wants to evaluate whether An automated inspection process examines

the mean weight of cereal boxes is the same samples of bolts produced by four machines

Number of groups: across three different production shifts. The for severe cracks that make the bolts unusable.

More than two analyst samples boxes from each shift and For each sample, analysts record the number

records their weights. of bolts that are inspected and the number of

bolts that are rejected.

Data type

Continuous Attribute

data data

If possible, collect continuous data because they provide more detailed information. However,

sometimes attribute data adequately describe the quality of a part or a process. For example, if you

are tracking broken light bulbs, you don’t need to measure a characteristic of the bulb to evaluate

whether it’s broken or not. What matters is only the number of bulbs that are broken (counts).

Do you want to compare

the center or the spread of Center of distribution Spread of distribution

each distribution? Compares the central value of each group to Compares the variation of each group to

determine whether at least one group differs determine whether the data vary more (or

from the others. less) in at least one group than the others.

Compare groups A quality engineer wants to compare the A manager of a lumberyard measures the

with each other hardness of four different paint blends after length of beams that are cut by three different

applying them to a metal surface. She wants sawmills. He wants to compare the spread

Number of groups: to determine whether the hardness of at least of the measurements for each sawmill and

More than two one paint blend is different from the others. determine whether the consistency of the

beam lengths is the same.

Data type:

Continuous

Focus of

comparison

distribution distribution

You can evaluate the center of the data using the mean (the arithmetic average) or the median (the

middle value when the data are arranged from lowest to highest). The median is less sensitive than

the mean to outliers.

You can evaluate the spread, or variation, of the data using the standard deviation or the variance.

The standard deviation is often easier to interpret because it uses the same units as the data.

Does your data follow

a normal or nonnormal Normal distribution Nonnormal distribution

distribution? Data that follow a symmetric, bell-shaped Data that follow a distribution that either is

distribution. not symmetric or is symmetric but not bell-

shaped.

Example

Focus of test: A quality engineer measures the hardness of Example

Compare groups four different paint blends after applying them A health administrator records the unoccupied

with each other to a metal surface. The engineer determines bed space for three hospitals located in the

that the hardness of each blend follows a same city. The administrator determines that

Number of groups: normal distribution. the data are not symmetrically distributed and,

More than two therefore, are not normal.

Data type:

Continuous

Focus of comparison:

Center of the distribution

Data

distribution

Normal Nonnormal

distribution distribution The distribution of your data often depends on your process. For example, data that track cycle

time for service processes, such as the time needed to process an application, serve a customer, or

deliver a product, often do not follow a normal distribution.

Although the normal distribution is always bell-shaped, not all bell-shaped distributions are normal. To

determine whether your data follow a normal distribution, use normal probability plots, histograms, or

normality tests such as the Anderson-Darling test. You can perform these evaluations using Minitab’s

Normality Test or Graphical Summary.

If you have 30 or more observations in each sample, you can generally treat your data as being

normal when you perform a basic statistical test to compare the means of two or more groups.

Basic Statistical Tests One-Way ANOVA

One-Way ANOVA

A one-way ANOVA determines whether the means of two or more independent groups differ and provides

a range of values that is likely to include the true mean of each group.

Example

Focus of test: A quality engineer measures the hardness of four different paint blends after applying them to a metal

Compare groups surface. The engineer uses a one-way ANOVA to determine whether the mean hardness differs among

with each other the blends.

To perform a one-way ANOVA in Minitab, choose Stat > ANOVA > One-Way. To determine which groups

Number of groups: differ from one another, check Comparisons and select a method for comparing the groups.

More than two

Continuous

Source DF SS MS F P

Paint 3 281.7 93.9 6.02 0.004

Error 20 312.1 15.6

Focus of comparison: Total 23 593.8

Center of the distribution

S = 3.950 R-Sq = 47.44% R-Sq(adj) = 39.56%

Data distribution: Individual 95% CIs For Mean Based on Pooled StDev

Level N Mean StDev +---------+---------+---------+---------

Normal

Blend A 6 14.733 3.363 (-----*------)

Blend B 6 8.567 5.500 (------*------)

Blend C 6 12.983 3.730 (------*------)

One-Way ANOVA Blend D 6 18.067 2.636 (------*------)

+---------+---------+---------+---------

5.0 10.0 15.0 20.0

Typically, you use one-way ANOVA to evaluate the means of two or more groups if you have

continuous data that are normally distributed. The test assumes that the data from each group have

about the same spread (equal variance). To test the variances of the groups, use Minitab’s Test for

Equal Variances.

You can also use one-way ANOVA to evaluate data that is nonnormal but meets one of these

conditions: has 30 or more observations, is symmetric with at least 5 observations, or has relatively

few extreme outliers.To quickly evaluate normality and outliers, use Minitab’s Graphical Summary. If

your data contains any extreme outliers, make sure they are valid measurements.

Use Power and Sample Size for one-way ANOVA to determine how much data you need to detect an

important difference between the group means.

Basic Statistical Tests Kruskal-Wallis

Kruskal-Wallis

The Kruskal-Wallis test determines whether the medians of two or more groups differ.

Example

A health administrator wants to compare the unoccupied bed space in three hospitals located in the

Focus of test: same city. Because the data are not normal and contain extreme outliers, she uses the Kruskal-Wallis

Compare groups test to determine whether the median bed space in each hospital differs.

with each other

To perform a Kruskal-Wallis test in Minitab, choose Stat > Nonparametrics > Kruskal-Wallis.

Number of groups:

Kruskal-Wallis Test: Beds versus Hospital

More than two

Kruskal-Wallis Test on Beds

Continuous 1 11 16.00 14.0 -1.28

2 11 31.00 23.3 2.65

3 11 17.00 13.7 -1.37

Overall 33 17.0

Focus of comparison:

Center of the distribution H = 7.05 DF = 2 P = 0.029

H = 7.05 DF = 2 P = 0.029 (adjusted for ties)

Data distribution:

Nonnormal

You can also use one-way ANOVA to evaluate data that is nonnormal but meets one of these

Kruskal-Wallis conditions: has 30 or more observations, is symmetric with at least 5 observations, or has relatively

few extreme outliers.To quickly evaluate normality and outliers, use Minitab’s Graphical Summary. If

your data contains any extreme outliers, make sure they are valid measurements.

If you have many extreme outliers, you may want to use Mood’s median test to compare the central

values of the groups. Although Mood’s median test is less powerful than the Kruskal-Wallis test, its

results are not as strongly affected by outliers.

Basic Statistical Tests Test for Equal Variances

Test for Equal Variances

A test for equal variances determines whether the variation (variance or standard deviation) of two

or more independent groups differs, and it provides a range of values that is likely to include the true

variation of each group.

Compare groups The manager of a lumberyard wants to compare the consistency of the length of beams cut by three

with each other different sawmills. He uses a test for equal variances to evaluate whether the variation of the beam

lengths differs for any of the sawmills.

Number of groups: To perform a test for equal variances in Minitab, choose Stat > ANOVA > Test for Equal Variances.

More than two

Data type:

Continuous

Focus of comparison:

Spread of the distribution

You can use the test for equal variances on normal or nonnormal data.

Basic Statistical Tests Cross Tabulation and Chi-Square

Cross Tabulation and

Chi-Square Cross tabulation and chi-square analysis tests whether the distribution of observations for an attribute

variable differs for any of the groups.

Example

Focus of test: An automated inspection process examines samples of bolts from four suppliers and determines whether

Compare groups each bolt contains a severe crack that makes it unusable. Analysts use cross tabulation and chi-square

with each other to determine whether the distribution of cracked bolts differs among the suppliers.

To perform cross tabulation and chi-square in Minitab, choose Stat > Tables > Cross Tabulation and

Number of groups: Chi-Square. Click Chi-Square and check Chi-Square analysis.

More than two

Tabulated statistics: Supplier, Cracked bolt

Attribute

No Yes All

Chi-Square B 92 8 100

C 90 10 100

D 100 0 100

All 378 22 400

t

Pearson Chi-Square = 11.352, DF = 3, P-Value = 0.010

Likelihood Ratio Chi-Square = 16.026, DF = 3, P-Value = 0.001

For a more in-depth analysis of the relationship between the groups and their levels, consider using

logistic regression.

Contacts

Minitab World Headquarters

Minitab Inc.

Quality Plaza Minitab is a global company with subsidiaries and representatives around the world. To

1829 Pine Hall Road find a Minitab partner in your country, visit www.minitab.com/contacts.

State College, PA 16801-3008

USA

Training

Phone: +1 814.238.3280 x3236 Training by Minitab™ maximizes your ability to improve quality. It helps you make more

Fax: +1 814.238.4383 effective business decisions by teaching you how to analyze your data with Minitab®

Statistical Software and manage your projects using Quality Companion by Minitab™.

Email: training@minitab.com

www.minitab.com/training

Technical Support

Phone: +1 814.231.2682 Our specialists are highly skilled in Minitab software, statistics, quality improvement, and

www.minitab.com/support; computer systems. Minitab subsidiaries and Independent Local Representatives around

the world offer technical support by phone in their local language.

customer.minitab.com to log a question

Mentoring

Phone: +1 814.238.3280 x3236 Mentoring by Minitab™ makes it easier to implement cost-saving quality improvement

Email: mentoring@minitab.com initiatives by providing the statistical support you need, just when you need it. We even

begin with a free consultation.

www.minitab.com/training/mentoring/

© 2009 Minitab, Inc. All rights reserved. The contents of this publication may not be reproduced without permission.

MINITAB® and all other trademarks and logos for the Company’s products and services are the exclusive property of Minitab, Inc. All other marks referenced remain the property of their respective

owners. See www.minitab.com for more information.

182 Six Sigma – Concepts and Cases

Mumbai Dabawala

and take the empty Tiffin boxes back. They make one

mistake in two months. This means there is one error in

every 16 million transactions!

”Six Sigma” means a failure rate of 3.4 parts per million or 99.9997 percent perfect;

however, the term in practice is used to denote more than simply counting defects.

Six Sigma can now imply a whole culture of strategies, tools, and statistical

methodologies to improve the bottomline of companies. In all, six sigma is a rigorous

analytical process for anticipating and solving problems. The objective of six sigma

is to improve profits through defect reduction, yield improvement, improved

consumer satisfaction and best-in-class product/process performance.

Why is it important?

perfection. To get to the six-sigma level means cutting down on huge costs and

thereby the wasted dollars. For example, if you are four sigma – you would be

producing products at the rate of 6200 defectives for every million you produce vs

3.4 defectives if you are at the six sigma level. Moreover, six sigma improvement

projects typically return in excess of $150k to $250k per project with a Black Belt

returning as much as $1 million to the bottomline each year.

Mumbai Dabawala 183

Texas Instruments (1988), IBM (1990), Asea Brown Boveri (1993), AlliedSignal/

Kodak (1994), GE (1995), Whirlpool, PACCAR, Invensys, & Polaroid (1996/98),

and many other companies worldwide have successfully implemented Six Sigma.

Recently Ford, DuPont, Dow Chemical, Microsoft and American Express have

started working on instituting Six Sigma processes.

Bottomline drives management action. What is your Cost of (poor) Quality? First

you need to determine that. Properly implemented, six sigma implementation can

become a profit-center for the company. Jack Welch at GE claims that the returns on

six sigma implementation amount to about $500 million as of 1998. Remember that

six sigma is complementary to other initiatives such as ISO or QS 9000 (which is

mainly procedural), Total Quality Management (which is mainly cultural) and

Statistical Process Control (which is primarily statistical process monitoring).

Six Sigma focuses on process quality. As such, it falls into the category of a process

capability (Cp) technique. Traditionally, a process is considered capable if the

natural spread, plus and minus three sigma (a yield of 99.73 percent), was less

than the engineering tolerance. A later refinement considered the process location

as well as its spread (Cpk) and tightened the minimum acceptable so that the

process was at least four sigma from the nearest engineering requirement. Six

Sigma requires that processes operate such that the nearest engineering requirement

is at least plus or minus six sigma from the process mean. This requires considerable

scientific and testing actions – often thousands of tests are run on multiple variables

to get an understanding of what’s going on. Once you determine the process

variables and using the other process analysis techniques, you need to consider

the ones causing the major losses and work on making them more capable.

etc.

184 Six Sigma – Concepts and Cases

A person trained in the Six Sigma methodology who is a team member of six sigma

process improvement action teams.

A person that is part of the leadership structure for process improvement teams are

called “Black Belts” (just as Total Quality utilized “Quality Improvement Team

Leaders” to provide structure). Black Belts are highly-regarded, technically-oriented

product or line personnel who have an ability to lead teams as well as to advise

management.

A person trained in the six sigma methodology who acts as the organization-wide

Six Sigma director or a program manager. He oversees Black belts and process

improvement projects and provides guidance to Black belts as required. A Master

Black belt teaches other six sigma students and helps them achieve Green belt and

Black belt status.

Usually a top executive or senior manager who “talks the talk” and “walks the

walk” of six sigma. This person is the sponsor, a catalyst and the driving force

behind the organization’s six sigma implementation.

For the legendary Jack Welch, a self-proclaimed cynic of quality programmes, Six

Sigma was not another quality tool that was heavy on slogans and light on results.

Under Welch’s regime, GE splurged $450 million (nearly Rs.2,205 crore) in two

years on training its employees for Master Black Belts, Black Belts and Green Belts

(leaders for Six Sigma initiatives within the company).

Mumbai Dabawala 185

So what was special for Six Sigma? Simply speaking, it is a vision, a tool for

quality improvement, a benchmark and a profit improvement methodology all

rolled into one. Started by Motorola in the US in 1985, it was used, among other

things, to manufacture a virtually defect-free pager. The methodology has been

extensively used for achieving defect-free products.

chose Six Sigma as a corporate vision and applied it for all their service outputs

(non-manufacturing) to achieve the Six Sigma level of defect-free outputs (99.99967

percent defect-free products). The Six Sigma methodology is based on the paradigm

that a zero defect in any product or process is possible. It believes that the Six

Sigma level of quality – just 3.4 defects per million opportunities – would be

attainable if the products and processes are designed properly.

The Six Sigma approach to quality ensures that the defects are eliminated

progressively by identifying the root causes and eliminating the source of variation.

As the defects are eliminated, the yield improves, work in process comes down,

customer satisfaction improves and the profitability of the company goes up.

While implementing Six Sigma, the first task at hand is identifying the factors

critical to quality and pinning down defects that put a question mark on quality.

As C R Nagaraj, corporate vice president, Mission:Quality, Wipro Limited, explains,

“Six Sigma converts a business problem into a statistical problem and finds a

statistical solution. It then converts the statistical solution into a business solution.”

This is the basic template for all the Six Sigma project methodologies. There are

tools and techniques at different stages that help one understand the problem,

diagnose root causes, validate critical root causes and implement corrective action.

For example, a tool like the Design Customer Satisfaction and Manufacturing

(DCSM) would be used for designing and manufacturing new products, while the

cross-functional process mapping (CFPM) would be used for large processes that

run through the business.

Variation that induces defects is caused by two factors: chance factors and

assignable factors. Chance factors are those that are generated by the system and

over which the operator has little control. For instance, in a manufacturing process,

186 Six Sigma – Concepts and Cases

when a lot of bearings in a machine wear out, they produce variation in the product

under manufacture. Similarly, deviations in the quality of raw material or power

supply produce variations. These are chance factors that the management has to

control.

factors are resolved by acting on the system and by management action alone. No

amount of operator skill can control these,” says Nagaraj.

On the other hand, the assignable factors that constitute 15 percent of the

variation is dependent on the person who is operating the system and factors such

as skills, diligence begin their role at this stage. “The management is most

responsible for the variation that is produced, which includes putting up the right

processes, putting up metrics for understanding process behaviour and plan

corrective action,” says Nagaraj.

Six Sigma differs from other quality initiatives in terms of its structured approach

to achieve profitability improvement through the competitive advantage.

Importantly, while other quality initiatives take an operations point of view, Six

Sigma approaches problems from the customer’s side.

Sigma is the statistical measurement for variation in any output, and when

companies talk of + - 6 times the Sigma within a specification, it means that 99.99967

percent of the products manufactured are within specification. The higher the

sigma level, the better the quality. For example, a Three Sigma level of quality

means 93.32 percent of the products or processes are within specification. Among

Indian companies the average Sigma level is estimated by consultants to be in at

2.5 Sigma level.

So how does the Six Sigma work? The methodology focuses mainly on the

strategically important outputs of an organisation that affect customer satisfaction.

The most critical to quality features are attacked first and the rest follow in order of

importance. A Six Sigma scale provides a means of establishing a measure of

performance for any tangible and intangible outputs.

breakthrough strategy to achieve Six Sigma quality in a process, division or a

Mumbai Dabawala 187

company. These eight stages are: Recognize, Define, Measure, Analyse, Improve,

Control, Standardise and Integrate. The highly skilled teams known as Black Belts

work full time on Six Sigma projects and lead teams through each of the core four

phases Measure, Analyse, Improve and Control (MAIC) that affect key process.

They make one Error on every 16 million transactions. The world renowned Forbes

magazine has selected them as a colossal example of six sigma’s success..

The Mumbai tiffinwallahs are international figures now thanks to Forbes Global.

The Forbes story details the efficiency with which they deliver the Tiffins of

their customers. Around 5000 tiffinwallahs deliver 175,000 lunches everyday and

take the empty Tiffin back. They make One Mistake in 2 months.

This means there is one Error on every 16 million transactions. This is thus a

6 Sigma performance (a term used in quality assurance if the percentage of

correctness is 99.999999) – the performance which has made companies like

Motorola world famous for their Quality.

can only aspire. “Efficient organisation” is not the first thought that comes to mind

in India, but when the profit motive is given free rein, anything is possible. To

appreciate Indian efficiency at its best, watch the tiffinwallahs at work.

These are the men who deliver 175,000 lunches (or “tiffin”) each day to offices

and schools throughout Mumbai, the business capital of India. Lunch is in a tin

container consisting of a number of bowls, each containing a separate dish, held

together in a frame. The meals are prepared in the homes of the people who commute

into Mumbai each morning and delivered in their own Tiffin carriers. After lunch,

the process is reversed. And what a process – in it’s complexity, the 5,000

tiffinwallahs make a mistake only about once every two months, according to

Ragunath Medge, 42, president of the Mumbai Tiffinmen’s Association. That’s

188 Six Sigma – Concepts and Cases

one error in every 8 million deliveries, or 16 million if you include the return trip.

“If we made 10 mistakes a month, no one would use our service,” says the craggily

handsome Medge.

How do they do it? The meals are picked up from commuters’ homes in suburbs

around central Mumbai long after the commuters have left for work, delivered to

them on time, then picked up and delivered home before the commuters return.

Each tiffin carrier has, painted on its top, a number of symbols which identify

where the carrier was picked up, the originating and destination stations and the

address to which it is to be delivered. After the Tiffin carriers are picked up, they

are taken to the nearest railway station, where they are sorted according to the

destination station. Between 10:15 a.m. and 10:45 a.m. they are loaded in crates

onto the baggage cars of trains. At the destination station they are unloaded by

other tiffinwallahs and re-sorted, this time according to street address and floor.

and cycles are delivered at 12:30 p.m., picked up at 1:30 p.m., and returned where

they came from.

The charge for this extraordinary service is just 150 rupees ($3.33) per month,

enough for the tiffinwallahs, who are mostly self-employed, to make a good living.

After paying Rs.60 per crate and Rs.120 per man per month to the Western Railway

for transport, the average tiffinwallahs clears about Rs.3,250. Of that sum, Rs.10

goes to the Tiffinmen’s Association. After minimal expenses, the rest of the

Rs.50,000 a month that the Association collects go to a charitable trust that feeds

the poor.

Superb service and charity too. Can anyone ask for more?

Comments

What is wonderful about this system is that it extends the design and uses the

tiffinwallahs, the end user and their cognitive and memory structure as well. Since

one tiffinwallahs is not going to pick more than 10-20 tiffincarriers, he can easily

sort recognize at the originating station and deliver it to the owner. Also within a

building, the Tiffinwalah knows which floor to deliver. Within a floor a owner can

Mumbai Dabawala 189

recognize his tiffin amongst others. Thus these Tiffins carry only * A symbol (not

name) of the originating station * A symbol for the destination station * A symbol

for the building where the addressee is.

And what is more amazing is that this is run by people, most of whom are

illiterate.

CMM, ISO9000, ERP, CRM, SCM etc, and it is not a one time process.

improvement, no saving.

- Chapter4Exercises135789111215171819203234353639444546and50.Uploaded bywraith324
- Measures of Central TendencyUploaded byHimraj Bachoo
- GRE Quant-DistributionUploaded byMonjurul Karim Raju
- The stat 43 assignmentUploaded bybajaman1234
- Day 4 - Measures of Central TendencyUploaded byLianne Sedurifa
- StatisticsUploaded byGarudaOzo
- Week 9Uploaded byKelvin Yuen
- Lesson Plan : StatisticsUploaded byNoritah Madina
- Mean Median ModeUploaded byRedmondWA
- OrginalSchmidt Operating Instructions Multilingual HighUploaded byOmar Zambrano Medina
- math lesson planUploaded byapi-298634756
- Laws of Mean, Median and ModeUploaded byDr. Tapan Kr. Dutta
- RES341 Team+C Whitner+Autoplex+Project Presentation FinalUploaded byJuanita Nu'Beauty Martin
- lesson 1 high levelUploaded byapi-297113268
- Student 1 PPT MAT Manuscript Science Achievement in Trece Martinez Final EditedUploaded byDemetrio Barrozo Dela Rosa Jr.
- 1 Sample SignUploaded byDheena Nagarajan
- grade 5-6 lesson planUploaded byapi-284999781
- 8th_meanmedian and ModeUploaded byYetz
- FAQsUploaded byellebannauiii
- Stats_X R. N. YadavUploaded byVibhu Goel
- grade6-42-2-5packUploaded byJdaartienda
- sas_statUploaded byjoydeep786
- Chapter 2b Central TendencyUploaded byPoonam Naidu
- TheoryUploaded byNatarajan Balasubramanian
- BSC Performance ebookUploaded byTri Toro
- Bell Curve ProjectUploaded byJay C
- 2752-8777-1-SM-2Uploaded byVania
- 6aapiUploaded byVijendra Gopa
- HOMOGENITAS 1Uploaded byAnonymous M6wmilWG
- T TestUploaded byFeliciano Bersamira

- Modern Management Theories and PracticesUploaded bySyed Waqar Akbar Tirmizi
- 5705Uploaded byCrystal Murray
- 6605q17ppcUploaded byexams_sbs
- AmulUploaded byhiralnmehta5502
- The Balance on the Balanced ScorecardA Critical Analysis of Some of Its AssumptionsUploaded byক্লান্তপদাতিক
- 1 basicconceptofmacroeconomicsUploaded byexams_sbs
- pom-unit-i-170903142454Uploaded byexams_sbs
- Unit-5 Core- Prof Adwait OakUploaded byexams_sbs
- PMUploaded byYaser Mohamed
- Just in TimeUploaded byexams_sbs
- MATERIAL REQUIREMENTS PLANNING.docUploaded byexams_sbs
- unit III Planning A1.docxUploaded byexams_sbs
- Unit II Management Thoughts AUploaded byexams_sbs
- More ExercisesUploaded byexams_sbs
- six sigmaUploaded byexams_sbs
- Reviewing_Balance_Scorecard_BSC_applicat.pdfUploaded byexams_sbs
- Quality Systems and Six SigmaUploaded byexams_sbs
- managementUploaded byexams_sbs
- HRM-MCQ-Sem-IUploaded bySuren Koundal
- 3 Chap III Ct q fUploaded byexams_sbs
- 2 LESSON 2 Freq Graphs FQUploaded byexams_sbs
- Statistics dispersionUploaded byexams_sbs
- BANA Syllabus 1.docxUploaded byexams_sbs
- 02 Internationaltradepolicy Chap2 -150218235618Uploaded byexams_sbs
- 01 Intro International EconomicsUploaded byexams_sbs
- MBA-Linear Programming.pptUploaded byexams_sbs
- Brand_Valuation.docxUploaded byexams_sbs
- Ch08Uploaded byexams_sbs
- 320EL14Uploaded byTarique Khan
- Accountancyfinalproject 150406045713 Conversion Gate01Uploaded byexams_sbs

- SyllabusUploaded bySujithShivu
- API MPMS 19.3 Evaporative Loss MeasurementUploaded bySandra Viviana Esquivel Benavides
- SPSS Dependent T-testUploaded byRosendo Bernabe
- MicroEconomicsUploaded bymariyam mohammad
- 8D methodUploaded byLaci
- SOCY-321 HomeworkUploaded byDaniel Oshiro
- research paper samstipicUploaded byapi-273337572
- L5_QMBM2362Uploaded byHamis Mohamed
- ANALYSIS BETWEEN THE TOURISM SECTOR AND ECONOMIC GROWTH IN AZERBAIJAN.Uploaded byIJAR Journal
- Ch10 - SimulationUploaded byBambang Ariwibowo
- Kellstedt Whitten Fundamentals Political Science ResearchUploaded byVICKCHAR
- XLendsadUploaded bymahajanamol89
- Testing of HypothesisUploaded byJyoshna Reddy
- Chapter+2+Lecture+NotesUploaded bymy009.tk
- Foundations of StatisticsUploaded bylodderig4466
- Dtsch_Arztebl_Int-106-0184Uploaded byClaudia Navarro
- Weibull FadingUploaded bywiznick
- Research Methodology - Problem Identification & FormulationUploaded byFrancisca Rebello
- Business Statistics Syllabus 8-WkUploaded bylcltxt
- occt 504 term paper-article critiqueUploaded byapi-201999002
- amity MBA (Gen) - 2014 Final[1]Uploaded byNeha Apa
- R abalUploaded byzeroxface
- Personal Income Taxes and Growth of Small Firms in NigeriaUploaded byEditor IJTSRD
- Data Science AcadgildUploaded bypraneet singh
- Analisys DiversityUploaded byJorge Valencia Herverth
- Real Interest ParityUploaded byRamber Sosa
- Ranganathan_Hidden Roots of Classification_chp3Uploaded byRobert Endean Gamboa
- Asymptotic Statistics - Lecture NotesUploaded byAnonymous gUySMcpSq
- Collier 1993 - The Comparative Method_23Sept.pdfUploaded byJavier Tarrega
- Christian Stepanek, Matthias Walter & Andreas Rathgeber (1)Uploaded byEben Mare