Improving Performance Using

Statistical Thinking

MN ASQ Section 1203

post-Dinner Meeting
September 14, 1999
presented by Robert Mitchell
Chair, ASQ Statistics Division
 MISSION
 Promote Statistical Thinking for quality and productivity
improvement.

 Serve ASQ, business, industry, academia, and government as a

resource for effective use of Statistical Thinking for quality and
productivity improvement.
 Our primary customers are Statistics Division members.
 Other key customers are:
 Management
 Users and potential users of Statistical Thinking


Educators of the above customers
 Provide a focal point within ASQ for application-driven
development and effective use of new statistical methods.

 Support the growth and development of ASQ Statistics Division

members.
 VISION

Statistical Thinking Everywhere

Statistical Thinking Statistical Methods

Process Variation Data Improvement

Philosophy Analysis Action

 DESIRED END STATE

 Our members will be proud to be part of the Statistics

Division
 Our Division’s operations will be a model for other
organizations.
 We will be a widely influential authority on scientific
approaches to quality and productivity improvement.
 PRINCIPLES
 Our customers’ needs will be continuously anticipated
and met (i.e. Customer focused rather than customer
driven).
 Our market focus for products and services is
weighted as follows:
 Greatest weight on intermediate level

 Nearly as much weight on basic level

 Much less weight on advanced level.

 Focus on a few key things.

 Balance short-term and long-term efforts.

 Value diversity (including geographical and

occupational) of our membership.
 Be proactive.
 PRINCIPLES

 Recognize that we exist for our customers.

 View statistics from the broad perspective of quality

management.
 Apply Statistical Thinking ourselves; that is, practice
what we preach.
 Uphold professional ethics.

 Continuously improve.
 STRATEGY
 Design and deliver selected useable products.

 Have a strong and vibrant Division infrastructure.

 Demonstrate the broad effectiveness of Statistical Thinking.

 Integrate Statistical Thinking into educational curricula.

 Develop a vibrant information communication system.

 Influence key decision makers.

Improving Performance Through
Statistical Thinking

• Why ST is important?
• What is ST?
• How do you apply ST?
• How to begin implementing ST
capability in an organization
• Summary
• Appendix - tools, case studies
 Definition

Statistical Thinking is a philosophy of learning and

action based on the following fundamental principles:

 All work occurs in a system of interconnected

processes

 Variation exists in all processes

 Understanding and reducing variation are keys

to success.

Glossary and Tables for Statistical Quality Control - Quality Press, 1996
Evaluation Based on Past Experiences
or Perceptions
 We’ve always done it this way!
 We work with what we get!
 We do the best we can!
 I never thought about it!
 I’m so busy with the day to day troubles, I
don’t have time
to try that!
 That has no effect on the finished product
(or service)!
 Evaluation Based on Data
THE KEY IS TO ASK WHY!!!

 Why is the material from the fabrication

department machine so inconsistent?
 Why do so we constantly firefight?

 Why does our daily output vary so much?

 Why does every job result in a monumental

task that takes forever to complete?
 Definition

Statistical Thinking is a philosophy of learning and

action based on the following fundamental principles:

 All work occurs in a system of interconnected

processes,
 Process
Transforms Inputs into Outputs
Inputs Transformation Outputs

Suppliers Accounts Payable Finished

Information Car Repair Goods
Historical Data Blending or Services
Approvals

External Factors
- Materials - People
- Methods - Machines
- Environment - Measures
 Processes and Sub-Processes
Supplier Supplier Supplier Supplier

External External
Sub- Sub- Sub-
Supplier Customer
Process A Process B Process C

Customer Customer Customer Customer

 Definition

Statistical Thinking is a philosophy of learning and

action based on the following fundamental principles:

 All work occurs in a system of interconnected

processes

 Variation exists in all processes

 Like it or not, variation is everywhere!

 Our drive time to work each day.

 The quantity of production each shift.
 Departure time of our plane.
 Measurement system
 ………
Know it, accept it, learn to deal with it!
 Definition

Statistical Thinking is a philosophy of learning and

action based on the following fundamental principles:

 All work occurs in a system of interconnected

processes

 Variation exists in all processes

 Understanding and reducing variation are keys

to success.
Deming once said:
“If I had to reduce my message for management
to just a few words, I’d say it all had to do with
reducing variation.”
 Variation and Targets
Variation can be thought of as:

1. Deviations around the overall

average or
Ave
2. A deviation of the overall average
from a desired target.

Ave Target
Reduced deviation around the
Average and on Target

Voice of Voice of
Process Customer

Target
 Improvement
using
Statistical Thinking

Process Variation Data Improvement

Statistical Thinking Statistical Methods

Philosophy Analysis Action
 Pitfall
Valuing only quantitative data

Qualitative data are just as important.

Example: Idea data!!

 Benefits of Statistical Thinking
Provides a theory and methodology for improvement:
 Helps identify where improvement is needed.
 Provides a general approach to take.
 Suggests tools to use.

A complete improvement approach includes all

elements of Statistical Thinking:
Process - Variation - Data
 Tips for Statistical Thinking
CAPDo Diagnostic
Check
Act
 Assess major problems and
 Fix the obvious.
root causes
 Decide how to
 Define the scope of the project
 Identify the people who will approach the
be affected by the project issue or
problem.

Do Plan
 Implement the Check Act
 Set project goals
plan. and expectations.
 Assure continued  Develop the
Do Plan
process project plan.
performance.
 Use of Statistical Thinking
Depends on levels of activity and job responsibility.

Where we're
Strategic Executives
headed

Systems & processes

Managerial Managers
to guide us

Where the work

Operational Workers
gets done
 Process

Process
Inputs Outputs
Suppliers Customers

S I P O C

A series of activities that converts inputs into outputs

 Understand Customer Needs

Substitute Tech. Char. Product Planning Matrix

End
User
Req.. Mfg. Tests Qualification Matrix
Data
Converting Package
Cu Matrix
st
Pr om
oc er Coating
es Re
s/ qu
Pr ire
od m
uc en
tF ts Extruding
lo
w
Examples of Operational Processes

 Manufacturing
 Order Entry
 Delivery
 Distribution
 Billing
 Collection
 Service
 Examples of Statistical Thinking
at the Operational Level
 Work processes are flowcharted and
documented
 Key measurements are identified.
 Time series plots displayed
 Process management and improvement utilize:
 Knowledge of variation, and
 Data
 Improvement activities focus on the process,
not blaming employees.
 Examples of Managerial Processes

1. Employee Selection 6. Setting Objectives

2. Training and and Goals
Development 7. Project
3. Performance Management
Management 8. Communications
(including coaching)
9. Management
4. Recognition and Reporting
Reward
10. Planning
5. Budgeting
 Examples of Managerial Processes
Management
Employee Reporting Training &
Selection Development

Planning
Core Business Processes
Objectives
& Goals
Budgeting
Customers

Communication Recognition
& Rewards
Performance
Management
 Examples of Statistical Thinking
at the Managerial Level
 Managers use meeting management techniques
 Standardized project management systems are
in place.
 Both project process and results are reviewed.
 Process variation is considered when setting
goals.
 Measurement is viewed as a process.
.
 The number of suppliers is reduced
 A variety of communication media are used.
Examples of Strategic Processes

 Strategic Plan Development

 Strategic Plan Deployment
 Acquisitions
 Corporate Budget Development
 Communications - Internal and
External
 Succession Planning and Deployment
 Organizational Improvement
 Examples of Statistical Thinking
at the Strategic Level
 Executives use systems approach.
 Core processes have been flow charted
 Strategic direction defined and deployed.
 Measurement systems in place (i.e. BSC).
 Employee, customer, and benchmarking
studies are used to drive improvement.
 Experimentation is encouraged.
 Learning is valued.
Robustness - An Underused Concept

 Key aspect of Statistical Thinking

 Reduce the effects of uncontrollable
variation in:
 Product design
 Process design
 Management practices
 Anticipate variation and reduce its effects
Robustness of Product and Process
Design
 A third way to reduce variation
 Anticipate variation
 Design the process or product to be insensitive to
variation
 A robust process or product is more likely to
perform as expected
 Improved reliability
 100% inspection cannot provide robustness
Robust Products are Designed in
Anticipation of Customer Use
 Unintended Use
 Washing machine tops
 Craftsman tools
 User-friendly computers and software
 Low-maintenance automobiles
 5 mph bumpers
 Medical instruments for home use
 Product and Process Robustness

 Product Performance is insensitive to

variations in conditions of manufacture,
distribution, use and disposal.

 Process Performance is insensitive to

uncontrollable variation in process
 Inputs
 Transformations - activities - steps
 External factors
 Robustness in Management
 Develop strategies that are insensitive to
economic trends and cycles
 Scenario Planning
 Design a project system that is insensitive to
 Personnel changes
 Changes in project scope
 Variations in business conditions
 Respond to differing employee needs
 Adopt flexible work hours
 Provide “cafeteria” benefits package

More ...
 Robustness in Management
(Continued)

 Enable personnel to adapt to changing

business needs
 High Performance Organization (HPO)
 Change Acceleration Process (CAP)
 Technological Acceptance
 Cultural Acceptance
 Ensure meeting effectiveness is not dependent
on facilities, equipment, or participants
 Understanding Human Behavior

 Different people have different methods and

styles of working, learning and thinking
 Different people take in, process and
communicate information in different ways
 People vary --- they are different:
 Day to day
 Person to person
 Group to group
 Site to site
 Organization to organization
 Process Robustness Analysis
 Identify Those Uncontrollable Factors that
Affect Process Performance
 Weather
 Customer Use of Products
 Employee Knowledge, Skills, Experience, Work
Habit
 Age of Equipment
 Design the Process to be Insensitive to the
Uncontrollable Variations in the Factors
 Three Ways to Reduce Variation
and Improve Quality
Control the Process:
Eliminate Special
Cause Variation

Improve the System: Quality

Reduce Common Improvement
Cause Variation

Anticipate Variation:
Design Robust
Processes and Products
 Tips for Statistical Thinking
Start with the customer.
Identify customer requirements and
expectations. Identify market trends.

Start small.
Pick a process that is not very detailed and does
not involve many functions or work groups.
Small “wins” build confidence and teamwork.
Share successes; celebrate failures.
 Tips for Statistical Thinking
7 Quality Tools 7 Management Tools
• Flow Diagram • Affinity Diagram
• Pareto Charts • InterRelationship
• Cause & Effect Diagram
• • Tree Diagram
Histograms
• • Prioritization Matrices
Run Charts
• • Matrix Charts
Control Charts
• • Arrow Diagram
Scatter Diagram
• Process Decision
Program Chart
3 Types of Variation

 Special Cause
 Common Cause
 Structural
 Identify Process State

Threshold State
Process StatesIdeal State
Process in Control Process in Control
NonConformance 100% Conforming
Change process or Product
specs Maintain Control
Sort

State of Chaos Brink of Chaos

Out of Control Out of Control

Out of Spec 100% Conforming

Eliminate Special Investigate Special

Causes Causes
 Tips for Statistical Thinking

If a process or system is not stable -- showing signs

of Special Cause …

 Use the tools of problem-solving to identify the root

cause. Try to identify exactly when, where, how, and
why the process changed.

 If a Special Cause hurts the process, develop

procedures to eliminate the return of the problem. If a
Special Cause is beneficial, develop procedures to make
it a permanent part of the process.
 Tips for Statistical Thinking

If a process or system is stable -- showing only

Common Causes of variation…

 Use the tools of improvement to study all the data

(not just the “good” or the “bad” points) and
identify factors that cause variation. Determine
what needs to be permanently changed to achieve a
different level of quality.
 Stratify the process - sort the data into groups
 Disagregate the process - divide the process into its sub-
processes
 Kano Model of Quality
Satisfied

Requested
Quality

Fulfilled
Not Fulfilled
Expected
Quality

Not Satisfied
Statistical Thinking Everywhere
1999 - 2000 STAT Tactics
1. Demonstrate the Broad Application of Statistical Thinking D1
 Collect new case studies; Market Statistical Thinking
 Expand the theory of Statistical Thinking
 Assist in the discovery process for new applications
 Maintain the integrity of ST (license and trademark)
 Publish new series, “Improving Processes…”

2. Integrate Statistical Thinking into Education & Training B2

 Pilot the “Pegasus” charter school project (Dallas, TX)
 PDCA the process; repeat with new school principals

3. Develop a Vibrant Information & Communication System A2

 Develop and implement a member nurturing process
 Create and staff the Section Liaison position
 Develop a WebPages policy and e-commerce strategy

4. Design & Deliver Selected Useable Products B5

 Develop and disseminate Newsletter technical content
 Create a Clearinghouse of Statistical Stuff
 Beef-up our Short Course offerings; more variety; regional
 Identify and start no more than 3 new publications

5. Have a Strong and Vibrant Division Infrastructure C3

 Division organization re-alignment; change By-Laws
 Develop a Division “Balanced Scorecard”
 Statistical Thinking Everywhere
read more about it...

“Statistical Thinking” Special Publication (Spring 1996)

“Data Sanity” Special Publication (Summer 1998)
available through QIC 1-800-248-1946 or qic@asq.org

“How to Teach Others to Apply Statistical Thinking”

Quality Progress, June 1997, pg. 67 - 79

“Improving Performance through Statistical Thinking”

ASQ Quality Press available November 1999