Six Sigma is a robust problem-solving method. However, training is often expensive due to time and software requirements. Over the course of my career, have often had to deliver easy-to-use, how-to, training on the Six Sigma method. These slides teach how to begin, execute and complete a project using key tools, such as CTQ tree, Process Mapping, Measurement System Analysis, Capability, Hypothesis Testing, Design of Experiments, Brainstorming, Control Charts, etc.  Including exercises, this training can be delivered in 3 days; and with some mentorship, close to green belt level performance can be achieved. thanks for your interest.

1. Stand
Problem Solving using
Six Sigma Methodology

2. © alpha2infinity
variation
2
Variation leads to Waste
DENTIST
1. Waiting
2. Over Production
3. Over Processing
4. Inventory
5. Motion
6. Transportation
7. Defects
8. Talent
Six Sigma Lean

3. © alpha2infinity
six sigma phases
3
Measure
Y = f(x1, x2, …, xn)
Analyze
Y = f(x1, x2, …, x9)
Improve
Y = f(x1, x2, x3)
Control
Define
Project Y
• Define project scope. Begin & End
• Define Project Y.
How will we measure success?
• Ensure good Measurement System
• Measure current performance
• Map the process and identify ALL
variables that can affect performance
• Analyze existing process data
• Identify significant process variables
affecting performance
• Experiment to determine set points of
critical process inputs to control process
• Implement new improved process
• Demonstrate performance of new
improved process
• Develop plans to sustain improvement
6Sigma aka DMAIC is a data-driven, variation reduction methodology

4. © alpha2infinity
typical six sigma project
4
O special cause
0
5
10
15
20
25
30
35
40
-4 -2 0 2 4 6 8 10 12 14 16
DefectRate
Time, weeks
Define Improve ControlMeasure & Analyze
Defects
Defects
Characterization & Understanding Improving & Optimizing

5. © alpha2infinity
why define?
5
“If I had only one hour to save
the world, I would spend fifty-
five minutes defining the
problem, and only five
minutes finding the solution”
Define

6. © alpha2infinity
why define?
6
Define
Without Define, we spin our wheels and waste time & money

7. © alpha2infinity
agenda
7
DEFINE
 Identify the vital few
• Bound the problem
• Understand customer requirements
• Prioritize customer requirements
• Identify project stakeholders
• Complete Project Charter
MEASURE
• Map the process (Fishbone)
• Prioritize process inputs
• Basic Statistics
• Verify measurement system is good
• Determine current process capability
ANALYZE
• Conduct Hypothesis Tests
• Chi Square
• t-Test & ANOVA
• Linear Regression
IMPROVE
• Design of Experiments
• Brainstorming
• Idea Box
• Solution Selection
CONTROL
• Demonstrate improvement
• Operating Instructions
• Develop control plan
• Monitor new process

8. © alpha2infinity
identify the vital few
Trivial Many Vital Few

9. © alpha2infinity
pareto analysis
9
1858 - 1923
Vilfredo Pareto was an Italian engineer, sociologist,
economist and philosopher.
Observed that 80% of the land in Italy was owned
by 20% of the population.
Focus on the Vital Few

10. © alpha2infinity
pareto chart
Pareto is a
Sorting Tool
Things to work on

11. © alpha2infinity
agenda
11
DEFINE
 Identify the vital few
 Bound the problem
• Understand customer requirements
• Prioritize customer requirements
• Identify project stakeholders
• Complete Project Charter
MEASURE
• Map the process (Fishbone)
• Prioritize process inputs
• Basic Statistics
• Verify measurement system is good
• Determine current process capability
ANALYZE
• Conduct Hypothesis Tests
• Chi Square
• t-Test & ANOVA
• Linear Regression
IMPROVE
• Design of Experiments
• Brainstorming
• Idea Box
• Solution Selection
CONTROL
• Demonstrate improvement
• Operating Instructions
• Develop control plan
• Monitor new process

12. © alpha2infinity
a process view
• Everything is a process.
• The quality of a product or service is directly related to
the quality of the process.
• As change agents, we must address the process as a
whole in order to deliver real change.
• We start with SIPOC – a high level view of the process
Suppliers Inputs OutputsProcess Customers

13. © alpha2infinity
SIPOC
OS PI C
Step A Step B Step C Step D Step E
START STOP
List ~5 High Level Steps to describe entire process
• Helps define project scope. Begin & End boundaries
• Helps define potential areas of improvement
• Helps identify customers impacted by your project
• Helps identify suppliers of inputs affecting your process

14. © alpha2infinity
SIPOC
1. Define First & Last step.
Break process into ~5 steps
2. List the major inputs of each
step.
3. Identify suppliers of each
input. Identify major suppliers.
4. List the major outputs of each
step.
5. Finally identify customers for
each output. People Impacted

15. © alpha2infinity
agenda
15
DEFINE
 Identify the vital few
 Bound the problem
 Understand customer requirements
• Prioritize customer requirements
• Identify project stakeholders
• Complete Project Charter
MEASURE
• Map the process (Fishbone)
• Prioritize process inputs
• Basic Statistics
• Verify measurement system is good
• Determine current process capability
ANALYZE
• Conduct Hypothesis Tests
• Chi Square
• t-Test & ANOVA
• Linear Regression
IMPROVE
• Design of Experiments
• Brainstorming
• Idea Box
• Solution Selection
CONTROL
• Demonstrate improvement
• Operating Instructions
• Develop control plan
• Monitor new process

16. © alpha2infinity
customer requirements
16
All customer requirements are not equal. They fall into 3 categories:
Basic → Performance → Delighter. And change over time.
Delighter:
These needs provide satisfaction when met, but do not cause dissatisfaction when
unmet. Since these attributes are unexpected, they delight customers, and are often
unimagined and unspoken.
Performance:
These needs result in satisfaction when met and dissatisfaction when unmet.
Delivering more of these needs result in increased customer satisfaction.
These are attributes that are verbalized and the ones in which companies compete.
Basic:
These needs are taken for granted when met, but result in dissatisfaction when unmet.
Meeting these needs do not result in increased customer satisfaction. Since customers
expect these attributes and view them as basic, they are unlikely to talk about them.

17. © alpha2infinity
KANO model
17
Time
Delighter
(Unknown)
Basic
(Unspoken)
What happens
over time??
Performance
(Spoken)
Did it
Poorly
Did it
Very
well

18. © alpha2infinity
hotel example
18
KANO example for hotel selection

19. © alpha2infinity
gathering Voice of Customer
19
Reactive Approach
• Looks at historical data to
determine issues, unmet
needs, problems &
interests of former
customers.
o Customer complaints
o Customer service calls
o Product return information
o Warranty claims
o Web page activity
Proactive Approach
• Verify conclusions of
reactive research
• Determine interests &
unmet needs of future
customers
o Surveys
o Interviews
o Market research
o Benchmarking
o Observe customers
 Takes more effort but is
often more powerful than
reactive research!

20. © alpha2infinity
VOC technique: listen
LISTEN
to Customers
 Survey
 Interviews
 Focus Groups

21. © alpha2infinity
VOC technique: observe
OBSERVE
Customer
behavior
 Focus Groups
 Go & See
 Ethnography
Studies

22. © alpha2infinity
affinity diagram
22
The affinity diagram is a tool used to
organize ideas and language data. This
method was developed by a noted
Japanese anthropologist in the 1950s
called Jiro Kawakita.
He was confronted with a need to
summarize large quantities of factual
language data from his field research.
He had an “aha” moment about grouping
the data and distilling it into useful
fundamental messages.

23. © alpha2infinity
voice(s) of customer
Normally my husband puts
appliances together, but not
this one. It was very easy to
assemble the item.
After meeting with the customers, write down one
key customer insight that you heard or saw per Post-it
Vote to select top 20

24. © alpha2infinity
affinity diagram
Ease of Assembly Category 2
Category 4Category 3
Lone Wolf
Group images into similar categories
Work until ≤ 5 categories
Category Titles are Drivers in CTQ tree

25. © alpha2infinity
agenda
25
DEFINE
 Identify the vital few
 Bound the problem
 Understand customer requirements
 Prioritize customer requirements
• Identify project stakeholders
• Complete Project Charter
MEASURE
• Map the process (Fishbone)
• Prioritize process inputs
• Basic Statistics
• Verify measurement system is good
• Determine current process capability
ANALYZE
• Conduct Hypothesis Tests
• Chi Square
• t-Test & ANOVA
• Linear Regression
IMPROVE
• Design of Experiments
• Brainstorming
• Idea Box
• Solution Selection
CONTROL
• Demonstrate improvement
• Operating Instructions
• Develop control plan
• Monitor new process

26. © alpha2infinity
critical to quality
26
A CTQ (critical-to-quality) is a specific measurable characteristic of a
product or service that fulfils the expressed requirement of the customer.
CTQ performance must exceed specifications to satisfy the customer.
Hard to measure Easy to measure
Need Drivers CTQs Specification
Rich 1.1 < TDS > 1.3
Not Acidic Extraction > 18%
Not Bitter Extraction < 22%
> 140°F
< 176°F
Price < $1.99 / medium
Taste
Temperature
Cost
Customer wants
"good coffee"
Optimum drinking
temperature
We need to translate VOC to CTQs

27. © alpha2infinity
CTQ tree
27
• Only 1 CTQ / Project
• Cannot improve
Project CTQ at
expense of other
CTQs
Cannot rob Peter to pay Paul

28. © alpha2infinity
agenda
28
DEFINE
 Identify the vital few
 Bound the problem
 Understand customer requirements
 Prioritize customer requirements
 Identify project stakeholders
• Complete Project Charter
MEASURE
• Map the process (Fishbone)
• Prioritize process inputs
• Basic Statistics
• Verify measurement system is good
• Determine current process capability
ANALYZE
• Conduct Hypothesis Tests
• Chi Square
• t-Test & ANOVA
• Linear Regression
IMPROVE
• Design of Experiments
• Brainstorming
• Idea Box
• Solution Selection
CONTROL
• Demonstrate improvement
• Operating Instructions
• Develop control plan
• Monitor new process

29. © alpha2infinity
stakeholders
Stakeholder:
Anyone impacted by your projectTeachers
Principal
Team Members
You
Stakeholders
can be inside
& outside the
organization

30. © alpha2infinity
why stakeholder analysis?
• Provide resources
• Give advice
• ‘lobby’ for your cause
• Break down barriers
• Accelerate and improve
results
• Withhold resources
• Withhold expertise
• Convince others to
withhold support
• Create barriers
• Delay and sabotage

31. © alpha2infinity
stakeholder analysis
Stakeholder analysis is a critical because it influences:
Resources, Solutions and ultimately Project success!
So we need to be Proactive about identifying and managing Stake Holders

32. © alpha2infinity
agenda
32
DEFINE
 Identify the vital few
 Bound the problem
 Understand customer requirements
 Prioritize customer requirements
 Identify project stakeholders
 Complete Project Charter
MEASURE
• Map the process (Fishbone)
• Prioritize process inputs
• Basic Statistics
• Verify measurement system is good
• Determine current process capability
ANALYZE
• Conduct Hypothesis Tests
• Chi Square
• t-Test & ANOVA
• Linear Regression
IMPROVE
• Design of Experiments
• Brainstorming
• Idea Box
• Solution Selection
CONTROL
• Demonstrate improvement
• Operating Instructions
• Develop control plan
• Monitor new process

33. © alpha2infinity
before project begins
Needs
Approval
&
Sponsor
Team

34. © alpha2infinity
project charter
Project Charter is
a good executive
communication
tool

35. © alpha2infinity
agenda
35
DEFINE
 Identify the vital few
 Bound the problem
 Understand customer requirements
 Prioritize customer requirements
 Identify project stakeholders
 Complete Project Charter
MEASURE
 Map the process (Fishbone)
• Prioritize process inputs
• Basic Statistics
• Verify measurement system is good
• Determine current process capability
ANALYZE
• Conduct Hypothesis Tests
• Chi Square
• t-Test & ANOVA
• Linear Regression
IMPROVE
• Design of Experiments
• Brainstorming
• Idea Box
• Solution Selection
CONTROL
• Demonstrate improvement
• Operating Instructions
• Develop control plan
• Monitor new process

36. © alpha2infinity
measure phase
36
Measure
Y = f(x1, x2, …, xn)
Analyze
Y = f(x1, x2, …, x9)
Improve
Y = f(x1, x2, x3)
Control
Define
Project Y
• Define project scope. Begin & End
• Define Project Y.
How will we measure success?
• Ensure good Measurement System
• Measure current performance
• Map the process and identify ALL
variables that can affect performance
• Analyze existing process data
• Identify significant process variables
affecting performance
• Experiment to determine set points of
critical process inputs to control process
• Implement new improved process
• Demonstrate performance of new
improved process
• Develop plans to sustain improvement

37. © alpha2infinity
process mapping
Everything is a process.
What is Process Mapping?
Graphical illustration of the work process
When do you apply process mapping?
Always
What are benefits of Process Mapping?
Identify major tasks / activities that are performed
Identify sequence of tasks / activities
Identify accountability of task /activities
Identify process inputs that affect process output
Tools Required
1. Room with
large wall
2. X-functional
team (6 – 8)
3. Experts on
the process
4. Sharpie
Markers
5. Large colored
Post-it
Start with
SIPOC
Complete
Flow chart
Complete
Process Map

38. © alpha2infinity
flowchart symbols
Start / Stop
Action
(verb / noun)
Decision Y
N
Delay or
Wait
Documentation
Or Paperwork
Movement

39. © alpha2infinity
flowchart example
Start Drive to Airport Park the Car Walk to Terminal
Wait in
Line
Security Check Walk to Gate
Wait for
Boarding
Stop
ID Check
e-ticket?
bags?Walk to security
Walk to Counter
Print
B. Pass
Board Plane
N
Y
N
Check bags
Wait in
Line
Wait in
Line
List + Sequence of Actions, Delays, Decisions and Movement

40. © alpha2infinity
Fishbone
Originally developed by Prof. Kaoru Ishikawa to show the
causes of a specific effect, this tool encourages the team to
think of inputs in categories of:
Personnel, Measurement, Environment, Material, Machine &
Method (6Ms).

41. © alpha2infinity
fishbone
41
Fishbone Diagram is
useful in identifying
all inputs to a
process step.
Ideally, create a
fishbone for every
major process step

42. © alpha2infinity
process mapping
42
• List Inputs and Outputs for each Action Step
• Goal is to identify ALL process inputs.
• Some of these are the cause of undesired
variation

43. © alpha2infinity
agenda
43
DEFINE
 Identify the vital few
 Bound the problem
 Understand customer requirements
 Prioritize customer requirements
 Identify project stakeholders
 Complete Project Charter
MEASURE
 Map the process (Fishbone)
 Prioritize process inputs
• Basic Statistics
• Verify measurement system is good
• Determine current process capability
ANALYZE
• Conduct Hypothesis Tests
• Chi Square
• t-Test & ANOVA
• Linear Regression
IMPROVE
• Design of Experiments
• Brainstorming
• Idea Box
• Solution Selection
CONTROL
• Demonstrate improvement
• Operating Instructions
• Develop control plan
• Monitor new process

44. © alpha2infinity
cause & effect matrix
44
Major process outputs
& their importance
Major
process
inputs
Higher Scores are
more important
Collect data on Project Y
& process inputs (Xs) for
subsequent analysis

45. © alpha2infinity
agenda
45
DEFINE
 Identify the vital few
 Bound the problem
 Understand customer requirements
 Prioritize customer requirements
 Identify project stakeholders
 Complete Project Charter
MEASURE
 Map the process (Fishbone)
 Prioritize process inputs
 Basic Statistics
• Verify measurement system is good
• Determine current process capability
ANALYZE
• Conduct Hypothesis Tests
• Chi Square
• t-Test & ANOVA
• Linear Regression
IMPROVE
• Design of Experiments
• Brainstorming
• Idea Box
• Solution Selection
CONTROL
• Demonstrate improvement
• Operating Instructions
• Develop control plan
• Monitor new process

46. © alpha2infinity
types of data
DATA
Attribute Variable
Binomial
Data can have only 2 values
• pass/fail, yes/no
Poisson
Data can have any discrete value
• A+, A, A-, B+, B, B-,… F
Data can have any value
on continuous scale
Normal
t-distribution
lognormal
Weibull
triangular
uniform
& many more!
COUNT MEASURE

47. © alpha2infinity
sample v population
Population
estimate
Study the sample to
make inferences
about the population
Sample
mini-population

48. © alpha2infinity
common statistics
DATA
Attribute Variable
Location
• mean
• median
• mode
Spread
• range
• standard deviation
• variance
Shape
• skewness
• kurtosis
Counts (Poisson)
Proportions (binomial)

49. © alpha2infinity
location
Mean: the total of values divided by the number of values;
• (2 + 1 + 2 + 3 + 4 + 5) / 6 = 2.83
• Is sensitive to extreme scores
Median: the middle value after a set of values has been sorted (low to high)
• 1, 2, 2, 3, 4, 5; the median is 2.5
• Is robust to extreme scores
Mode: the value that occurs most frequently;
• 1, 2, 2, 3, 4, 5; the mode is 2
• A data set may contain multiple modes; multimodal
Consider a sample: 2, 1, 2, 3, 4, 5

50. © alpha2infinity
spread
Range: max value – min value
• Range = 5 -1 = 4
• Is a poor estimator of variation; relies only on extreme values; ignoring
most values
Standard Deviation: the average distance of a data point from the mean
• Most commonly used; has same units as the mean
• Average distance from the mean
• Less sensitive to outliers than the range
Variance: (standard deviation)2
Consider a sample: 2, 1, 2, 3, 4, 5

51. © alpha2infinity
normal distribution
Many natural phenomena and man-made processes follow a
“normal” distribution. It is a continuous distribution and can be
described by its mean and standard deviation.
Different Mean
Same Standard Deviation
Same Mean
Different Standard Deviation

52. © alpha2infinity
shape
Positive Negative
SKEWNESS
symmetry
Platykurtic Leptokurtic
KURTOSIS
‘peakedness’
Mean
Median
Mode

53. © alpha2infinity
property of normal distribution
The area under the normal curve can be used to estimate the probability
of an “event” occurring.
If data is NOT normal (i.e. misshapen), these probabilities change.
3210-1-2-3
0.4
0.3
0.2
0.1
0.0
X
Density
95%
99.73%
68%
Ordinary UnusualUnusual

54. © alpha2infinity
normality check
54
If probability of
observing
normality test
statistic > 5%,
data is normal

55. © alpha2infinity
other continuous distributions
55

56. © alpha2infinity
agenda
56
DEFINE
 Identify the vital few
 Bound the problem
 Understand customer requirements
 Prioritize customer requirements
 Identify project stakeholders
 Complete Project Charter
MEASURE
 Map the process (Fishbone)
 Prioritize process inputs
 Basic Statistics
 Verify measurement system is good
• Determine current process capability
ANALYZE
• Conduct Hypothesis Tests
• Chi Square
• t-Test & ANOVA
• Linear Regression
IMPROVE
• Design of Experiments
• Brainstorming
• Idea Box
• Solution Selection
CONTROL
• Demonstrate improvement
• Operating Instructions
• Develop control plan
• Monitor new process

57. © alpha2infinity
why evaluate measurement system?
57
Observed Variation
Variation from Process Variation from Measurement
Due to Gage
- accuracy
Due to Operator
- precision
Need to ensure that at
least 90% of what we
are observing is from
process.
𝑉𝑎𝑟𝑖𝑎𝑡𝑖𝑜𝑛 𝑜𝑏𝑠𝑒𝑟𝑣𝑒𝑑 = 𝑉𝑎𝑟𝑖𝑎𝑡𝑖𝑜𝑛 𝑝𝑟𝑜𝑐𝑒𝑠𝑠 + 𝑉𝑎𝑟𝑖𝑎𝑡𝑖𝑜𝑛 𝑚𝑒𝑎𝑠𝑢𝑟𝑒𝑚𝑒𝑛𝑡
Repeatability
Reproducibility

58. © alpha2infinity
accuracy & precision
58
Measurement quality is described by accuracy & precision
Accuracy = Average value from target Calibration Study
Precision = Variation amongst data points Gage R&R Study

59. © alpha2infinity
repeatability
59
Repeatability = agreement with self (attribute data)
variation from one operator (variable data)
Good Repeatability
1
3
2
Poor Repeatability
1
3
2
Test Conditions
- same part
- same gage
- same operator
𝑉𝑎𝑟𝑖𝑎𝑡𝑖𝑜𝑛 𝑚𝑒𝑎𝑠𝑢𝑟𝑒𝑚𝑒𝑛𝑡 = 𝑽𝒂𝒓𝒊𝒂𝒕𝒊𝒐𝒏 𝒓𝒆𝒑𝒆𝒂𝒕𝒂𝒃𝒊𝒍𝒊𝒕𝒚 + 𝑉𝑎𝑟𝑖𝑎𝑡𝑖𝑜𝑛 𝑟𝑒𝑝𝑟𝑜𝑑𝑢𝑐𝑖𝑏𝑖𝑙𝑖𝑡𝑦

60. © alpha2infinity
reproducibility
60
Jerry
Bob
Mike
Good Reproducibility
Bob
Mike
Jerry
Poor Reproducibility
Reproducibility = agreement between operators (attribute data)
variation from different operators (variable data)
Test Conditions
- same part
- same gage
- different operators
𝑉𝑎𝑟𝑖𝑎𝑡𝑖𝑜𝑛 𝑚𝑒𝑎𝑠𝑢𝑟𝑒𝑚𝑒𝑛𝑡 = 𝑉𝑎𝑟𝑖𝑎𝑡𝑖𝑜𝑛 𝑟𝑒𝑝𝑒𝑎𝑡𝑎𝑏𝑖𝑙𝑖𝑡𝑦 + 𝑽𝒂𝒓𝒊𝒂𝒕𝒊𝒐𝒏 𝒓𝒆𝒑𝒓𝒐𝒅𝒖𝒄𝒊𝒃𝒊𝒍𝒊𝒕𝒚
Measurement System Analysis should be performed
for all Process CTQs (outputs of customer value)

61. © alpha2infinity
gage R&R attribute data
61
70%
90%
Kappa

62. © alpha2infinity
gage R&R variable data
62
28%
14%
%GRR

63. © alpha2infinity
agenda
63
DEFINE
 Identify the vital few
 Bound the problem
 Understand customer requirements
 Prioritize customer requirements
 Identify project stakeholders
 Complete Project Charter
MEASURE
 Map the process (Fishbone)
 Prioritize process inputs
 Basic Statistics
 Verify measurement system is good
 Determine current process capability
ANALYZE
• Conduct Hypothesis Tests
• Chi Square
• ANOVA
• Linear Regression
IMPROVE
• Design of Experiments
• Brainstorming
• Idea Box
• Solution Selection
CONTROL
• Demonstrate improvement
• Operating Instructions
• Develop control plan
• Monitor new process

64. © alpha2infinity
capability
64
Capability Studies determine process performance from the
point of view of the customer!
Voice of The Process
Voice of The Customer
1USLLSL
Specifications
From Customer
𝐶𝑎𝑝𝑎𝑏𝑖𝑙𝑖𝑡𝑦 𝐶𝑝 =
𝑉𝑂𝐶 (𝑇𝑜𝑙𝑒𝑟𝑎𝑛𝑐𝑒)
𝑉𝑂𝑃 (𝑃𝑟𝑜𝑐𝑒𝑠𝑠 𝑆𝑝𝑟𝑒𝑎𝑑)
=
(𝑈𝑆𝐿 − 𝐿𝑆𝐿)
𝟔𝜎 99.73%

65. © alpha2infinity
capability concept
65
TOLERANCE
PROCESS
TOLERANCE
PROC.
Capability = 1
3210-1-2-3
0.4
0.3
0.2
0.1
0.0
Capability = 2
3210-1-2-3
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
Cpk is a metric used to measure process capability

66. © alpha2infinity
capability metric, Cpk
66
𝐶𝑝𝑘 =
𝐷𝑖𝑠𝑡𝑎𝑛𝑐𝑒 𝑓𝑟𝑜𝑚 𝑋 𝑡𝑜 𝑛𝑒𝑎𝑟𝑒𝑠𝑡 𝑠𝑝𝑒𝑐
3𝜎
𝐶𝑝𝑘 = 𝑀𝑖𝑛 𝐶 𝑝𝐿, 𝐶 𝑝𝑈
𝐶 𝑝𝐿 =
𝑋 − 𝐿𝑆𝐿
3𝜎
𝐶 𝑝𝑈 =
𝑈𝑆𝐿 − 𝑋
3𝜎
Variable Data Attribute Data
𝐶𝑝𝑘 = 𝑃𝑟𝑜𝑝𝑜𝑟𝑡𝑖𝑜𝑛 𝑂𝑢𝑡𝑠𝑖𝑑𝑒 𝑠𝑝𝑒𝑐
0.20
0.15
0.10
0.05
0.00
Probability
11
0.01147
0
𝐹𝑎𝑖𝑙𝑢𝑟𝑒 𝑅𝑎𝑡𝑒 =
𝑁𝑢𝑚𝑏𝑒𝑟 𝐹𝑎𝑖𝑙𝑒𝑑
𝑁𝑢𝑚𝑏𝑒𝑟 𝐼𝑛𝑠𝑝𝑒𝑐𝑡𝑒𝑑
The failure rate can be converted to Cpk
Customer Requirement
Specification
Capability Study should be performed for all Process CTQs

67. © alpha2infinity
capability calculation
67
Cpk for Attribute Data
1.3
2.0
Cpk
Cpk of 2 is 6 sigma process
3.4 defects in 1,000,000
Cpk for Variable Data

68. © alpha2infinity
analyze phase
68
Measure
Y = f(x1, x2, …, xn)
Analyze
Y = f(x1, x2, …, x9)
Improve
Y = f(x1, x2, x3)
Control
Define
Project Y
• Define project scope. Begin & End
• Define Project Y.
How will we measure success?
• Ensure good Measurement System
• Measure current performance
• Map the process and identify ALL
variables that can affect performance
• Analyze existing process data using
appropriate hypothesis test
• Identify significant process inputs
• Experiment to determine set points of
critical process inputs to control process
• Implement new improved process
• Demonstrate performance of new
improved process
• Develop plans to sustain improvement

69. © alpha2infinity
agenda
69
DEFINE
 Identify the vital few
 Bound the problem
 Understand customer requirements
 Prioritize customer requirements
 Identify project stakeholders
 Complete Project Charter
MEASURE
 Map the process (Fishbone)
 Prioritize process inputs
 Basic Statistics
 Verify measurement system is good
 Determine current process capability
ANALYZE
 Conduct Hypothesis Tests
• Chi Square
• t-Test & ANOVA
• Linear Regression
IMPROVE
• Design of Experiments
• Brainstorming
• Idea Box
• Solution Selection
CONTROL
• Demonstrate improvement
• Operating Instructions
• Develop control plan
• Monitor new process

70. © alpha2infinity
hypothesis testing
70
Our Decision
X has NO
effect on Y
X has an
effect on YTheAbsoluteTruth
XhasNO
effectonY 
a error
Seeing
something
not present
Xhasan
effectonY
b error
Miss seeing
something
that is present

In the measure phase, we identified all
‘potential’ causes of problem.
In analyze phase we will examine data
to identify the ‘significant’ causes.

71. © alpha2infinity
hypothesis test selection
71
SingleY
Single X
Attribute Variable
Attribute
Chi2 Logistic
Regression
Variable
t-test
ANOVA
Medians
Linear
Regression
1. Based on data type select the
correct Hypothesis test
2. All Hypothesis tests calculate
the probability of a error and
report it as p-value.
3. IF p-value is < 0.05, we are at
least 95% confident that the
X has a significant effect on Y

72. © alpha2infinity
scenarios
72
Is there a difference in burrito size between Chipotle & Moe’s?
Parameter Description Type of Data
Y
X
Hypothesis Test
Weight of Burrito Variable
Burrito Shop Attribute
ANOVA (t-test)
Does the age of male customer affect the type of car they buy?
Parameter Description Type of Data
Y
X
Hypothesis Test
Type of Car Attribute
Age Variable
Logistic Regression

73. © alpha2infinity
scenarios
73
Does the air pressure effect how far TB12 can throw a football?
Parameter Description Type of Data
Y
X
Hypothesis Test
Distance Thrown Variable
Air Pressure Variable
Regression
Parameter Description Type of Data
Y
X
Hypothesis Test
Hire: yes, no Attribute
Age: young, old Attribute
Chi-Square
The HR department wants to see if there is a link between age
(old and young) and whether that person gets hired

74. © alpha2infinity
agenda
74
DEFINE
 Identify the vital few
 Bound the problem
 Understand customer requirements
 Prioritize customer requirements
 Identify project stakeholders
 Complete Project Charter
MEASURE
 Map the process (Fishbone)
 Prioritize process inputs
 Basic Statistics
 Verify measurement system is good
 Determine current process capability
ANALYZE
 Conduct Hypothesis Tests
 Chi Square
• t-Test & ANOVA
• Linear Regression
IMPROVE
• Design of Experiments
• Brainstorming
• Idea Box
• Solution Selection
CONTROL
• Demonstrate improvement
• Operating Instructions
• Develop control plan
• Monitor new process

75. © alpha2infinity
chi2 test
- Enter Data
- Evaluate p-value
- Make Conclusion

76. © alpha2infinity
agenda
76
DEFINE
 Identify the vital few
 Bound the problem
 Understand customer requirements
 Prioritize customer requirements
 Identify project stakeholders
 Complete Project Charter
MEASURE
 Map the process (Fishbone)
 Prioritize process inputs
 Basic Statistics
 Verify measurement system is good
 Determine current process capability
ANALYZE
 Conduct Hypothesis Tests
 Chi Square
 t-Test & ANOVA
• Linear Regression
IMPROVE
• Design of Experiments
• Brainstorming
• Idea Box
• Solution Selection
CONTROL
• Demonstrate improvement
• Operating Instructions
• Develop control plan
• Monitor new process

77. © alpha2infinity
t-test
77
Use to compare
means of TWO
groups

78. © alpha2infinity
mann-whitney test
78
When data is NOT
normal, compare
medians.

79. © alpha2infinity
between
ANOVA
79
𝑊𝑖𝑡ℎ𝑖𝑛 𝑉𝑎𝑟𝑖𝑎𝑡𝑖𝑜𝑛 = 𝑋𝑖 − 𝑋𝑖𝑗
2
𝐵𝑒𝑡𝑤𝑒𝑒𝑛 𝑉𝑎𝑟𝑖𝑎𝑡𝑖𝑜𝑛 = 𝑋𝑖 − 𝑋
2
𝐹 =
𝐵𝑒𝑡𝑤𝑒𝑒𝑛
𝑊𝑖𝑡ℎ𝑖𝑛
=
𝑆𝑖𝑔𝑛𝑎𝑙
𝑁𝑜𝑖𝑠𝑒
is F = 1?
X2
𝑟1
𝑟2
𝑟3
X1
𝑏3
𝑏2
𝑏1
between
Called ANalysis Of VAriance, but we
make conclusions about Means

80. © alpha2infinity
ANOVA
80
Use to determine if
means of > TWO
groups are all equal

81. © alpha2infinity
mood’s median test
81
When data is NOT
normal, analyze
medians.

82. © alpha2infinity
agenda
82
DEFINE
 Identify the vital few
 Bound the problem
 Understand customer requirements
 Prioritize customer requirements
 Identify project stakeholders
 Complete Project Charter
MEASURE
 Map the process (Fishbone)
 Prioritize process inputs
 Basic Statistics
 Verify measurement system is good
 Determine current process capability
ANALYZE
 Conduct Hypothesis Tests
 Chi Square
 t-Test & ANOVA
 Linear Regression
IMPROVE
• Design of Experiments
• Brainstorming
• Idea Box
• Solution Selection
CONTROL
• Demonstrate improvement
• Operating Instructions
• Develop control plan
• Monitor new process

83. © alpha2infinity
Regression
83
Variation in Y
explained by
Model
1. Is there a linear relationship between X and Y?
2. Can the X predict the Y value?? i.e Equation
3. & if yes, how good is this prediction???

84. © alpha2infinity
improve phase
84
Measure
Y = f(x1, x2, …, xn)
Analyze
Y = f(x1, x2, …, x9)
Improve
Y = f(x1, x2, x3)
Control
Define
Project Y
• Define project scope. Begin & End
• Define Project Y. How will we measure
success?
• Ensure good Measurement System
• Measure current performance
• Map the process and identify ALL
variables that can affect performance
• Analyze existing process data using
appropriate hypothesis test
• Identify significant process inputs
• Experiment to determine set points of
critical process inputs to control process
• Implement new improved process
• Demonstrate performance of new
improved process
• Develop plans to sustain improvement

85. © alpha2infinity
agenda
85
DEFINE
 Identify the vital few
 Bound the problem
 Understand customer requirements
 Prioritize customer requirements
 Identify project stakeholders
 Complete Project Charter
MEASURE
 Map the process (Fishbone)
 Prioritize process inputs
 Basic Statistics
 Verify measurement system is good
 Determine current process capability
ANALYZE
 Conduct Hypothesis Tests
 Chi Square
 t-Test & ANOVA
 Linear Regression
IMPROVE
 Design of Experiments
• Brainstorming
• Idea Box
• Solution Selection
CONTROL
• Demonstrate improvement
• Operating Instructions
• Develop control plan
• Monitor new process

86. © alpha2infinity
DOE
86
In the ANALYZE phase we looked at existing data.
In the IMPROVE phase, we conduct experiments using a
systematic method to determine cause-and-effect
relationships between inputs into a process (factors) and
the output of that process (response).
Process Inputs (Factors)
– Change brand of gas
 Octane rating (85, 90)
 Driving Speed (55, 60)
– Tune-up Car
– New tires
 Tire Pressure (30, 35)
 Significant Inputs identified in analyze phase
Process Output (Response)
– Fuel efficiency
Specification: > 30 mpg
Driving
(Process)

87. © alpha2infinity
DOE setup
87
- Enter set points of factors to
setup experimental conditions
- Enter response values of
experimental runs

88. © alpha2infinity
DOE
88
- Evaluate p-value to
identify critical factor(s)
- Determine contribution
of each factor (practical)

89. © alpha2infinity
agenda
89
DEFINE
 Identify the vital few
 Bound the problem
 Understand customer requirements
 Prioritize customer requirements
 Identify project stakeholders
 Complete Project Charter
MEASURE
 Map the process (Fishbone)
 Prioritize process inputs
 Basic Statistics
 Verify measurement system is good
 Determine current process capability
ANALYZE
 Conduct Hypothesis Tests
 Chi Square
 t-Test & ANOVA
 Linear Regression
IMPROVE
 Design of Experiments
 Brainstorming
• Idea Box
• Solution Selection
CONTROL
• Demonstrate improvement
• Operating Instructions
• Develop control plan
• Monitor new process

90. © alpha2infinity
brainstorming
90
Brainstorming
is a tool for generating as many ideas or solutions to a problem.
Ideas should be as broad and wide as possible. There are no bad ideas.
However it is not a tool for determining the best solution to the problem.
Brainstorming should be done with a team

91. © alpha2infinity
brainstorming Methods
91
Free-wheeling brainstorm:
• Team members call out their ideas spontaneously.
• The scribe records the ideas as they are suggested.
Round-robin brainstorming:
• The facilitator asks each member, in turn, for an idea.
• Members may pass on any round.
• The facilitator records the ideas. The session continues until team
members have no more ideas to add.
The slip method:
• The facilitator asks members to write down their ideas on Post-its.
• Post-its are placed on wall for everyone to read.
• Allow team to write new ideas after reading the Post-its
• All ideas are then collected and organized (No duplication, Clarity).


92. © alpha2infinity
agenda
92
DEFINE
 Identify the vital few
 Bound the problem
 Understand customer requirements
 Prioritize customer requirements
 Identify project stakeholders
 Complete Project Charter
MEASURE
 Map the process (Fishbone)
 Prioritize process inputs
 Basic Statistics
 Verify measurement system is good
 Determine current process capability
ANALYZE
 Conduct Hypothesis Tests
 Chi Square
 t-Test & ANOVA
 Linear Regression
IMPROVE
 Design of Experiments
 Brainstorming
 Idea Box
• Solution Selection
CONTROL
• Demonstrate improvement
• Operating Instructions
• Develop control plan
• Monitor new process

93. © alpha2infinity
ideabox
93
Idea Box is a creativity technique based on
concepts (morphological analysis) developed by
Fritz Zwicky a Swiss astrophysicist in the 1940's.
1. Write down the problem statement
2. Break down the problem into smaller elements
(think SIPOC)
3. Lists options for each element
4. Pick options that combine into ‘best’ solution

94. © alpha2infinity
ideabox
This technique results
in many permutations
The art lies in picking
the best combination
to solve the problem

95. © alpha2infinity
agenda
95
DEFINE
 Identify the vital few
 Bound the problem
 Understand customer requirements
 Prioritize customer requirements
 Identify project stakeholders
 Complete Project Charter
MEASURE
 Map the process (Fishbone)
 Prioritize process inputs
 Basic Statistics
 Verify measurement system is good
 Determine current process capability
ANALYZE
 Conduct Hypothesis Tests
 Chi Square
 t-Test & ANOVA
 Linear Regression
IMPROVE
 Design of Experiments
 Brainstorming
 Idea Box
 Solution Selection
CONTROL
• Demonstrate improvement
• Operating Instructions
• Develop control plan
• Monitor new process

96. © alpha2infinity
selecting a solution
96
Quality S1 S2 S3 S4
Mary 1 2 3 4
Peter 2 1 3 4
Priya 1 2 3 4
Chelsea 1 2 3 4
TOTAL 5 7 12 16
Each person ranks
Each solution against
Each criteria.
The total score is
entered into the
matrix and solutions
are ranked from best

97. © alpha2infinity
agenda
97
DEFINE
• Pareto of problems
• Bounding the problem
• KANO & ‘Voice of Customer’
• Why talk to all customers?
• Project Charter
MEASURE
• Process Mapping
• Fishbone
• Basic Statistics
• Measurement System Analysis
• Capability
ANALYZE
• Hypothesis Testing
• Chi Square
• ANOVA
• Linear Regression
IMPROVE
• Design of Experiments
• Brainstorming
• Idea Box
• Solution Selection
CONTROL
 Demonstrate improvement
• Write Instructions
• Develop control plan
• Monitor new process

98. © alpha2infinity
B v C comparison
98
Analyze data from old & new process
Evaluate if
1. New process is better? (t-test)
2. Is it good enough? (new Cpk)

99. © alpha2infinity
agenda
99
DEFINE
• Pareto of problems
• Bounding the problem
• KANO & ‘Voice of Customer’
• Why talk to all customers?
• Project Charter
MEASURE
• Process Mapping
• Fishbone
• Basic Statistics
• Measurement System Analysis
• Capability
ANALYZE
• Hypothesis Testing
• Chi Square
• ANOVA
• Linear Regression
IMPROVE
• Design of Experiments
• Brainstorming
• Idea Box
• Solution Selection
CONTROL
 Demonstrate improvement
 Write Instructions
• Develop control plan
• Monitor new process

100. © alpha2infinity
examples of SOP
100
Too Wordy
Succinct, Multi-Lingual, Minimal Reading

101. © alpha2infinity101
List
inputs,
required operations &
expected outputs
for each process step

102. © alpha2infinity
agenda
102
DEFINE
• Pareto of problems
• Bounding the problem
• KANO & ‘Voice of Customer’
• Why talk to all customers?
• Project Charter
MEASURE
• Process Mapping
• Fishbone
• Basic Statistics
• Measurement System Analysis
• Capability
ANALYZE
• Hypothesis Testing
• Chi Square
• ANOVA
• Linear Regression
IMPROVE
• Design of Experiments
• Brainstorming
• Idea Box
• Solution Selection
CONTROL
 Demonstrate improvement
 Write Instructions
 Develop control plan
• Monitor new process

103. © alpha2infinity
control plan
103
Complete all sections of
control plan,
particularly
- current controls &
- reaction plans
for key process outputs

104. © alpha2infinity
agenda
104
DEFINE
• Pareto of problems
• Bounding the problem
• KANO & ‘Voice of Customer’
• Why talk to all customers?
• Project Charter
MEASURE
• Process Mapping
• Fishbone
• Basic Statistics
• Measurement System Analysis
• Capability
ANALYZE
• Hypothesis Testing
• Chi Square
• ANOVA
• Linear Regression
IMPROVE
• Design of Experiments
• Brainstorming
• Idea Box
• Solution Selection
CONTROL
 Demonstrate improvement
 Write Instructions
 Develop control plan
 Monitor new process

105. © alpha2infinity
control chart
105
Investigate root cause of any
out-of-control incident.

106. © alpha2infinity
pre-control chart
106
Run in green zone,
Reset in red zone,
Investigate if in between

107. Stand
Thank You