ANALYZE

www.asklexph.com
Analyze Phase
Goal: To identify potential root causes and to confirm real causes
using data.

Output Deliverables:

✓ List of potential root causes

✓ Confirmed root causes
✓ List of causes for improvement

Generate List validated

Screen potential Validate using
potential root root causes for
root causes data analysis
causes improvement

www.asklexph.com
BRAINSTORMING

www.asklexph.com
What is Brainstorming?
▪ A technique used to generate ideas from a group of individuals in any
stage of any problem solving activity.

▪ It is a means of getting a large number of ideas from a group of people in

a short time.

▪ The more alternative choices we have, the better our decision is.

www.asklexph.com
Brainstorming Methods

✓Structured

• participants go in turns, one idea per turn until all ideas

are exhausted

✓Unstructured

• participant who has an idea can raise it and in no

specific turn of order

www.asklexph.com
Brainstorming Rules

1) Give participants a time to think about the topic to be

discussed.

2) No killer phrase and no pet ideas

3) Crazy ideas are welcome

4) Build up on ideas previously raised

5) The more the ideas, the better.

6) Make the ideas visual.

7) Summarize and combine ideas,

if necessary.
www.asklexph.com
WHY – WHY
ANALYSIS

www.asklexph.com
5 Whys (Why-Why) Analysis

▪ It is an iterative, interrogative technique used to explore the cause-and-

effect relationship underlying a particular problem.

▪ Its primary goal is to establish the root cause of an effect by repeating

the question “why?”; each answer becomes the basis of the next
question.

▪ The technique is developed by Sakichi Toyoda and was used within

Toyota

www.asklexph.com
Why – Why Analysis in Action #1

www.asklexph.com
Why – Why Analysis in Action #2

www.asklexph.com
Why – Why Analysis in Action #3

www.asklexph.com
Why – Why Analysis in Action #4

www.asklexph.com
How to Get Started with Why - Why

1. Form the Team

2. Define the Problem

3. Ask Why

4. Validate

5. Take Action
www.asklexph.com
FISHBONE
DIAGRAM

www.asklexph.com
What is a Fishbone (Cause & Effect) Diagram?

▪ A tool used to analyze cause and effect relationship.

▪ Use your team's process knowledge about the entire process to create
an effective cause-and-effect diagram.

▪ This is created by Kaoru Ishikawa, 1968.

www.asklexph.com
How to Construct a Fishbone Diagram?
1) Generate potential causes of a problem through brainstorming.

2) Draw a horizontal arrow pointing to the right. This is the spine.

3) Decide upon the major cause categories of the event, problem, or key
characteristics.

4) Write the major cause categories on the left-hand side of paper and draw
lines to them off the main horizontal line.

5) Conduct a Why – Why Analysis by asking why questions that can lead to
the potential root causes.

6) Look for causes that appear repeatedly. These may represent root causes.
Look for what you can measure in each cause so you can quantify the
effects of any changes you make. Most importantly, identify and circle
the causes that you can take action to.

www.asklexph.com
Fishbone Diagram Sample #1

www.asklexph.com
Fishbone Diagram Sample #2

www.asklexph.com
Fishbone Diagram Sample #3

www.asklexph.com
Root Cause Analysis Tips

1. How relevant are the questions and answers to the original

X or Y you are investigating?

2. Did you find a root cause that helps you control or avoid the
situation?

3. Are the questions and answers significant enough,

considering your project scope?

4. Can we validate the potential root cause using

relevant data?

www.asklexph.com
GRAPHICAL
ANALYSIS

www.asklexph.com
Graphical Analysis Tools

▪ These are set of tools used to validate the hypotheses or potential root
causes identified from the root cause analysis.

▪ Graphical Analysis Tools are used to examine the distribution of data,

track data over time and examine potential relationship between data.

Source: Minitab Ltd.

www.asklexph.com
 DATA
DISTRIBUTION

www.asklexph.com
Graphical tools for the distribution of data

www.asklexph.com
HISTOGRAM

www.asklexph.com
What is a Histogram?
A histogram is a plot that lets you discover, and show, the
underlying frequency distribution (shape) of a set
of continuous data. This allows the inspection of the data for its
underlying distribution (e.g., normal distribution), outliers,
skewness, etc.

www.asklexph.com
Anatomy of a Histogram
Histogram of None, GrowFast, SuperPlant
Normal Central Tendency
Vari abl e
0.09 N on e
Grow Fast
0.08 SuperPl ant

0.07 Mean StDev N

18.42 5.091 50
0.06 25.16 4.287 50
20.41 5.500 49
Density

0.05

0.04

0.03

0.02

0. 01

0.00
8 12 16 20 24 28 32 36
Data
Dispersion

www.asklexph.com
Typical Types of Histogram

The skewed distribution is

A common pattern is the bell-
asymmetrical because a
shaped curve known as the
natural limit prevents
"normal distribution."
outcomes on one side.

www.asklexph.com
Typical Types of Histogram

The outcomes of two Several processes with

processes with different normal distributions are
distributions are combined in combined.
one set of data.

www.asklexph.com
Histogram Using Minitab #1

Source Sheet: Energy

Graph > Histogram > Simple with Fit

www.asklexph.com
Histogram Using Minitab #2
GrowFast, a plant fertilizer manufacturer, is attempting to develop a formula that
yields the greatest plant growth. In a controlled greenhouse environment, they set up
3 groups of 50 identical seedlings, a control group with no fertilizer, a group using their
product, and another group using their closest competitor, SuperPlant. After 3 months,
they measured the plants' heights in centimeters.

Source Sheet: Plant Histogram of None, GrowFast, SuperPlant

Normal
Vari abl e
0.09 N on e
Grow Fast
0.08 SuperPl ant

0.07 Mean StDev N

Graph > Histogram > 0.06
18.42
25.16
5.091
4.287
50
50
20.41 5.500 49
Density

With Fit and Groups 0.05

0.04

0.03

0.02

0. 01

0.00
8 12 16 20 24 28 32 36
Data

www.asklexph.com
BOXPLOT

www.asklexph.com
What is a Boxplot?
A box and whisker plot is defined as a graphical method of
displaying variation in a set of data. In most cases, a histogram
analysis provides a sufficient display, but a box and whisker plot
can provide additional detail while allowing multiple sets of data
to be displayed in the same graph.

Target

A B
www.asklexph.com
 Boxplot
Anatomy of aofBox
None,Plot
GrowFast, SuperPlant

* Outlier

4th Quartile

3rd Quartile
Inter Quartile Range 2nd Quartile | Median
(IQR)
1st Quartile
IQR = Q3 – Q1

www.asklexph.com
Boxplot Using Minitab #1

Source Sheet: Energy

Graph > Boxplot > Simple Y
www.asklexph.com
 Boxplot Using Minitab #2
GrowFast, a plant fertilizer manufacturer, is attempting to develop a formula that yields
the greatest plant growth. In a controlled greenhouse environment, they set up 3
groups of 50 identical seedlings, a control group with no fertilizer, a group using their
product, and another group using their closest competitor, SuperPlant. After 3 months,
they measured the plants' heights in centimeters.

Source Sheet: Plant Boxplot of None, GrowFast, SuperPlant

40

35

30

Graph > Boxplot > 25

Data

Multiple Y's 20

15

10

None GrowFast SuperPlant

www.asklexph.com
PARETO CHART

www.asklexph.com
What is a Pareto Chart?
▪ A bar chart where the bars are arranged in descending order of magnitude.
The bars may represent defect categories, locations, departments, and so on.

▪ A problem-solving tool that involves ranking all potential problem areas or

sources of variation according to their contribution to cost or total variation.

▪ Typically, 80% of the effects come from 20% of the possible causes, so efforts
are best spent on these “vital few” causes, temporarily ignoring the “trivial
many” causes.

▪ Most commonly used for prioritization of resources and also for validation of
root causes.

www.asklexph.com
Anatomy of a Pareto Chart

Cumulative
%

Categories

Frequency

www.asklexph.com
How to Create a Pareto Chart?

1. Decide what categories you will use to group items.

2. Decide what measurement is appropriate. Common

measurements are frequency, quantity, cost and time.

3. Decide what period of time.

4. Collect the data, recording the category each time, or

assemble data that already exist.

5. Subtotal the measurements for each category.

www.asklexph.com
How to Create a Pareto Chart?
6. Determine the appropriate scale for the measurements you have collected. The
maximum value will be the largest subtotal from step 5. (If you will do optional steps 8
and 9 below, the maximum value will be the sum of all subtotals from step 5.) Mark the
scale on the left side of the chart.

7. Construct and label bars for each category. Place the tallest at the far left, then the next
tallest to its right, and so on. If there are many categories with small measurements, they
can be grouped as “other.”

8. Calculate the percentage for each category: the subtotal for that category divided by the
total for all categories. Draw a right vertical axis and label it with percentages. Be sure
the two scales match. For example, the left measurement that corresponds to one-half
should be exactly opposite 50% on the right scale.

9. Calculate and draw cumulative sums: add the subtotals for the first and second
categories, and place a dot above the second bar indicating that sum. To that sum add
the subtotal for the third category, and place a dot above the third bar for that new sum.
Continue the process for all the bars. Connect the dots, starting at the top of the first bar.
The last dot should reach 100% on the right scale.

www.asklexph.com
Pareto Chart Sample Using Minitab #1
A quality engineer for an automotive supply company wants to decrease the
number of car door panels that are rejected because of paint flaws. The
engineer wants to determine whether a relationship exists between the type
of paint flaws and the shift during which the door panels are made.
Source Sheet: Flaws

Stat > Quality

Tools > Pareto
Chart

Defects: Flaws

www.asklexph.com
Pareto Chart Sample Using Minitab #2

A clothing manufacturer tracked the number and type of defects in a line of clothing.

Source Sheet: ClothingDefect

Based on defect count Based on total cost

Stat > Quality Tools > Pareto Chart Stat > Quality Tools > Pareto Chart

Defects: Defect Defects: Defect

Frequencies: Count Frequencies: Count*Cost

www.asklexph.com
Things to consider when you use a Pareto chart, p1

▪ Data collected during a short period of time, especially from

an unstable process, may lead to incorrect conclusions.
Because the data may not be reliable, you may get a
misleading idea of the distribution of defects and causes.
When the process is not in control, the causes may be
unstable and the vital few problems may change from week
to week. Short periods of time may not be representative of
your process as a whole.

▪ Data collected during long periods of time may include

changes. Examine the data for stratification or changes in
the problem distribution over time.

www.asklexph.com
Things to consider when you use a Pareto chart, p2

▪ Choose categories carefully. If your initial Pareto analysis

does not yield useful results, you may want to ensure that
your categories are meaningful and that your "other" category
is not too large.

▪ Choose weighting criteria carefully. For example, cost may be

a more useful measure for prioritization than number of
occurrences, especially when the costs of various defects
differ.

www.asklexph.com
Things to consider when you use a Pareto chart, p3

▪ Concentrating on the problems with the highest frequency

should decrease the total number of items needing rework.
Concentrating on the problems with the highest cost should
increase the financial benefits of the improvement.

▪ The goal of a Pareto analysis is to obtain maximum reward

from the quality efforts, but that doesn't mean that small,
easily solved problems should be ignored until the larger
problems are solved.

www.asklexph.com
TIME SERIES

www.asklexph.com
Graphical tools for variables over time

www.asklexph.com
TIME SERIES

www.asklexph.com
What is a Time Series?

A time series is a sequence of observations over regularly

spaced intervals of time.

▪ Monthly unemployment rates for the previous five years

▪ Daily production at a manufacturing plant for a month
▪ Decade-by-decade population of a state of the previous
century

www.asklexph.com
Anatomy of a Time Series Chart

Data
Points

Time
Elements

www.asklexph.com
Why use time series chart?
To check for patterns:

▪ Trend

▪ Shift

▪ Cycle

www.asklexph.com
Time Series Sample #1
The manager of a shipping yard wants to study the amount of cargo that is
transported. The manager collects the weight of all the cargo that passes through
the shipping yard each month.

Source Sheet: Shipping

Graph > Time Series

Plot > Simple

Series: Weight
Time/Scale: Stamp >
Month

www.asklexph.com
Time Series Sample #2
The administrator of a hospital wants to examine the number of cardiac patients
admitted over the past 24 months to analyze trends in the data.

Source Sheet: Patient

Time Series Plot of Year 1

460

440

Graph > Time Series 420

Plot > Simple Year 1

400

380

360
Series: No. of Patients
340

Time/Scale: Stamp > 320

Month 300
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Month

www.asklexph.com
 RELATIONSHIP
BETWEEN VARIABLES

www.asklexph.com
Graphical tools for relationships between variables

www.asklexph.com
SCATTER PLOT

www.asklexph.com
What is a Scatter Plot?
▪ It is a problem solving tool and is also known as the correlation diagram.

▪ It is used to uncover possible cause-and-effect relationships.

▪ It cannot prove that one variable causes another, but it does show
how a pair of variables is related and the strength of that relationship.
Statistical tests quantify the degree of correlation between the variables.

▪ It is a tool used in regression modelling or predicting a close estimate of

the dependent variable given a value of the independent variable by the
factor of relationship present between the two.

www.asklexph.com
Anatomy of a Scatter Plot

Best
Fit
Line

Paired
Data
Points

www.asklexph.com
How to Interpret a Scatter Plot

www.asklexph.com
Scatter Plot Sample Using Minitab #1
A research team at a fitness assessment company is looking for a method to predict a
person's body fat percentage. This health measure is difficult and expensive to measure
directly. In its model, the team wants to include a predictor variable that is easier to measure,
and is considering the use of Body Mass Index (BMI).

Source Sheet: BMI

Scatterplot of %Fat vs BMI
50

45

Stat > Graph > 40

Scatter Plot > 35

%Fat

With Regression 30

25

Y Variables: %Fat 20

X Variables: BMI 15
15 20 25 30 35
BMI

www.asklexph.com
Scatter Plot Sample Using Minitab #2
A food scientist wants to determine the optimal time for heating a frozen dinner. The
scientist prepares 14 samples at various times and temperatures, and then has
professional food tasters rate each sample for overall quality.

Source Sheet: Frozen

Stat > Graph >

Scatter Plot >
With Regression

Y Variables: Quality
X Variables: Time

www.asklexph.com
Scatter Plot Sample Using Minitab #3
A quality engineer for a camera manufacturer wants to shorten the flash recovery
time. Flash recovery time is the least amount of time that is required between flashes.
The engineer wants to determine whether a relationship exists between the voltage
that remains in the camera battery immediately after a flash and the flash recovery
time.
Source Sheet: Flash

Stat > Graph >

Scatter Plot >
With Regression

Y Variables: Flash
X Variables: Volts

www.asklexph.com
VALIDATION
TABLE

www.asklexph.com
When to Best Use Each Graphical Tool

Y Continuous Categorical

Histogram
Box Plot
Continuous Scatter Plot
Pareto (prioritization)
Time Series (over time)

Data Tables > Bar Charts > Summary

Categorical
Statistics

www.asklexph.com
What is a Validation Table?
A validation table is a simple table that summarizes the efforts done in
generating and validating the potential root causes. It will give the team an
idea of which has been valid root causes among all the generated potential
root causes.

Validation Method
(Observation, Records
Potential Root Causes Conclusion
Checking,
(Valid or Not Valid)
Data/Graphical
Analysis, etc.)

There is high variation in preparation

Box Plot Valid
time affecting the lead time

Lead time is higher during peak hours

Time Series Valid
due to surge of customer

www.asklexph.com