Root Cause Analysis- An Overview

This training event is funded by the Irish Medtech Association Skillnet

The Irish Medtech Association Skillnet is a national network which delivers training, upskilling
and professional development programme for the medical technology & pharmachemical sectors
in Ireland.

The Network is overseen by Irish Medtech Association, and Ibec as contracting organisation.

Contact [email protected] or Phone 061 – 43 1802 for further info.

The Irish Medtech Association Skillnet is funded by member companies and the Training Networks Programme, an
initiative of Skillnets funded from the National Training Fund through the Department of Education and Skills.
www.skillnets.ie

2
Course Overview

Overview of Various Root Cause Analysis Tools

 Problem Solving Architecture
 Tool Selection
 When and How to Use Tools
 Investigation Progression
 Agenda
 Day 1: (2.5 Hours)
 Problem Solving Architecture
 Overview of DMAIC/ PDCA
 Define the Problem
 Day 2: (2.5 Hours)
 Overview of CAPA
 Preventive Action Tools
 Corrective Action Tools
 Day 3: (2.5 Hours)
 Poka Yoke
 Bias
 Summary

©Loftus Consulting 4
Problem Solving Architecture
 Process Control
Capability Conformance
Analysis Analysis

Eliminate Investigate for

Assignable Cause Assignable Cause

Capability analysis
• What is the currently "inherent" capability of my process when it is "in control"?
• “Have I got a fighting chance to make good product?”

Conformance analysis
• Quality Management System controls daily activities
• Data monitoring identifies when control has likely been lost and special cause variation has occurred

Investigate for assignable cause

• Find “Root Cause(s)” of special variation
• Use one of our Problem Identification tools or Six Sigma

Eliminate assignable cause

• Implement corrective actions to improve process capability
 The Meaning and Objectives
of Continuous Improvement
 Lean is a high touch/low tech methodology
focused on reducing waste in processes

 Six Sigma is a systematic methodology that

focuses on reducing variability and defects

In combination, they deliver a proven method for

Continuous Improvement

7
 Video- Toast Kaizen

An Introduction to Continuous Improvement & Lean

Principles

8
 Choose Your Battles Carefully!
Value Non-value Added
Added (Waste or “Muda”)

It Is Much MORE: It Is Much:

• Dangerous • Safer
• Risky • Less Risky
• Time Consuming • Faster
• High Tech • Lower Tech
• Expensive • Cheaper
• Customer Impacting • Customer Friendly
• Demanding Of • Less Demanding Of
Management Time Management Oversight
To Speed Up Or Improve To Speed Up Or Improve
Value Added Activity Non-value Added Activity
9
What Is Six Sigma?
 Often referred to as the DMAIC process of
process improvement.
 The simple Six-Sigma formula is Y=f(X), where
Y represents the key process outputs, X
represents key process inputs that strongly
effect the output, and f represents the
relationship between the inputs and outputs.

10
Variation – Simple Exercise
 Bus 1 avg. time = 55 mins
 Bus 2 avg. time = 50 mins
 How many would take bus 1? Bus 2?

 More information: 1 weeks data

 Bus 1: 51, 55, 58, 52, 59 mins
 Bus 2: 90, 40, 35, 45, 40 mins
 How many would take bus 1? Bus 2?

11
Variation
Variation is the fluctuation seen in the output of a process. There are two
types of variation:

 Common Cause Variation: (Chance Cause)

 Irregular variation within historical norms
 The ‘noise’ within the system, natural pattern
 Usually not significant relative to the product tolerance

 Special Cause Variation: (Assignable Cause)

 Usually unpredictable or sudden variation
 Often a cause of out of specification product

Monitor variation through the use of Control Charts

12
Major Sources of Variation
 Poor Design
 Skills and Behaviors
 Unstable Parts and Materials
 Measurement Systems
 Insufficient Process Capability

13
Overview of DMAIC/ PDCA
 Six Sigma- Five Steps
Six Sigma Funnel
The DMAIC (D-MAY-IK) Model
Y
Define the problem and project Process Input Variables
Define
x x
Measure Measure & gather current baseline data

x x
Analyse Analyse the data, identify root cause(s) x
x x x x
Improve Improve by addressing root cause(s) x x
x
x
Control Control to sustain gains

Key Variables
X X X
Remember Y=f(X)
 Six Sigma DMAIC Cycle Summary
DEFINE Identify, scope, and justify the project(s)

MEASURE Experiment with key process settings to

identify their impact on key product
characteristics

ANALYSE Review data from Measure phase to

identify the key process inputs (KPIVs)
and their impact on the output (KPOVs)

IMPROVE Carry out improvements based on data

analysis and know-how to optimise
performance

CONTROL Assure Long-Term process stability.

DMAIC
16
 Phase 1: DEFINE
Purpose: To accurately describe the problem, define project
success, and highlight resources required to complete the project.
Collect process and customer background information.

Possible Deliverables:
• Project charter
• Project team
• Financial Assessment- Approved COPQ
• Understand the CTQ Characteristics
• Project plan and timeline
DMAIC
17
 Project Problem Statements
As part of the Charter, an effective problem
description is:
 Specific, explaining exactly what is wrong
 Measurable, indicating the scope of the problem in quantifiable
terms by answering questions like “How much?” “How many?” or
“How often?”
 Achievable, meaning there is a realistic chance to solve the
problem
 Relevant, meaning it can be resolved with the assigned resources
 Timely, meaning it will be delivered within an acceptable
timeframe
DMAIC
18
Typical Definition Sheet

19
 Phase 2: MEASURE
Purpose: Challenge current measurement systems and ensure data is credible.
Gather existing and/or new data to help characterise the process capability and
to focus the project by narrowing the range of potential causes.
Possible Deliverables:
• Measurement system analyses (MSA)
Define
• Data collection plan
• Baseline measurement data

Me
Con

asu
• Process Map Breakthrough

t ro l

re
• Cause and Effect Diagram Methodology
• Process FMEA
Y=f(X)

• List of Theories to test

Im z e
pr aly
• First draft of preliminary control plan o ve An

DMAIC 20
Where does Variation Occur?
- Beware of Measurement System Variation!

System Variation

Measurement
Measurement Variation
Part to Part (Actual)
Variation
Variation

Long Term Short Term Due to Operator Due to Gauge

Process Variation Process Variation Issues

Lack of Lack of
Reproducibility Repeatability

DMAIC Carry Out a Gauge R&R Study

21
Six Sigma Defect Levels
Processes that operate to ‘6 sigma quality’ levels
produce less that 3.4 defects per million
opportunities (DPMO).

The goal of 6σ is to make all defects unacceptable

and to strive for Zero defects.

22
Measuring the Level of Control
-Capability Index

If you shoot darts at a board and your shots are falling in

the same spot forming a tight group, this gives a high Cp
score, and if the grouping is around the target (bullseye),
you also have a high CpK score.

Cpk of 1.33 [4 sigma] is usually sufficient to to satisfy most

customers. CpK of <1 is usually unacceptable; and CpK
of 1.5 is approaching 6 Sigma standard

Cp and CpK are calculated using statistical techniques

23
PROCESS CAPABILITY
Capability spectrum

Low Capability Medium Capability High Capability

 Phase 3: ANALYSE
Purpose: Through data analysis, narrowing down the trivial many
process variables (X's) to the significant few Key Process Input Variables
(KPIV's).

Possible Deliverables:
• List of Key Process Input Variables (KPIV's / Vital X's)
• Updated Preliminary Set of Validation Settings
• Updated List of Improvements to Test

Toolbox to be used by Black/ Green Belts: Histograms, Boxplots,

Scatterplots, Multivari Charts, Pareto Charts, Run Charts, Dotplots, Defect
Concentration Plots, Main Effects Plots, Interaction Plots, Interval plots,
Factor Pareto Charts, Marginal Plots, Correlation and Regression
DMAIC 25
 Not all inputs are important
Only a few will drive the process
- Focus on Process Inputs - Focus on Process Outputs

Y, X1, …….., X15

Characterisation

Y’s DEFINE & MEASURE

X’s
Optimisation

ANALYSE

IMPROVE
Unimportant X’s
CONTROL removed

Important X’s (KPIVs)

discovered and controlled Y = f ( X1, X5, X9 )
26
 ACCURACY
• Bias is the difference between the observed average measurement
value and the “targeted” or “true” value. Bias is a measure of “lack of
accuracy”.
• It is typically the most common and easiest problem to fix.

TRUE Bad Accuracy, Good

VALUE Precision

DMAIC 27
 PRECISION
• PRECISION is the ability to replicate measurements time after time.
LACK OF PRECISION implies VARIABILITY.
• Typically a number of issues are at fault, making resolution more
difficult

TRUE Accurate but not

VALUE Precise

DMAIC 28
 STABILITY
 This may be a measurement issue. All
instruments need to be periodically re-calibrated.
 A lack of stability could imply a “creeping bias”
over time. Trend analysis needed on a regular
basis to monitor for performance creep.

Week 3

Week 2

Week 1

29

DMAIC
The Importance of Robust Data
 Robust data highlights ‘unseen’ problems like accuracy,
precision and stability
 Analysis of robust data will almost always reveal a
previously unknown source of process variation
 Tracking robust data can predict problems before they
become significant
 Caution: Measurement systems are an often overlooked
source of poor data- challenge your data sources (Gauge
R&R)
30
Predicting Distance can be Important !!
 Phase 4: IMPROVE
Purpose: Optimise the process output by making improvements to
gain control of the key inputs. You may need to Measure and Analyse
again to confirm expected improvement.
Define

Possible Deliverables:

Me
Con

as
Breakthrough

ure
• Engineering improvements

trol
Methodology
• Reliable measurement system Y=f(X)

• Optimal settings for X’s Im e

z
pr
o aly
ve An
• Delivery of goal from define phase

DMAIC

32
 Phase 5: CONTROL
Purpose: To maintain the gains through control of the Key Process Input
Variables (KPIV's).
Possible Deliverables:
Define
 Error Proofing / Poka-Yoke
 Safety Improvements

Me
Con

asu
Breakthrough

r
rol

e
 5S / TPM / Updated PM Methodology
Y=f(X)
 Validation (IQ, OQ, PQ)
 Updated FMEA Im
pr aly
ze
o ve An

 Updated Value Stream Map (Improved Process)

 Updated SOP’s
 Training
 Control Plan e.g. SPC (Statistical Process Control)

DMAIC  Project Impact & Summary

e.g. updated COPQ and documented project report (A3 storyboard)
33
The Importance of Robust Control
 Controlling key input variables gives a high degree
of assurance that the output (product) will be
good.
 Early detection of problems is far less expensive
 The process (equipment, habits, acceptance levels,
etc) will revert to the previous ‘normal’ levels if
specific control mechanisms are not put in place

34
 Maintaining Robust Control
- Standard Work
 Lots of ways of carrying out a task- find the best way
 Document the best way in SOP’s
 Train all personnel on SOP’s, with demonstrated evidence
of understanding
 Audit ongoing application of procedures to ensure they are
strictly adhered to.
 Update as new learnings are discovered

Challenge is to keep procedures up to date and effectively implemented,

especially during a period of change 35
 Plan-Do-Check-Act (PDCA)
(Also known as the Deming Cycle)

Plan: Identify and analyse the problem

Do: Develop and test a potential solution

Check: Measure impact of test solution,

and analyse for improvement

Act: Implement the improved solution fully

©Loftus Consulting
 PDCA
Plan Do Check Act
Pareto Analysis Rank Solutions Measure ‘Just Do It!’
5 Why’s Trial Improve Kaizen Blitz
Brainstorming Pilot Project Cost Analysis Project Flow
Fishbone Project Mgt
Repeat
FMEA

PDCA is used to Identify and Implement the next Continuous Improvement iteration

©Loftus Consulting
Overview of CAPA
13485- CAPA
Corrective Action
The organisation shall take action to eliminate the cause
of nonconformities in order to prevent recurrence.
Documented procedure for:
 Review nonconformities
 Determine causes
 Evaluate need for action to prevent recurrence
 Determine and implement actions, including
documentation
 Record results of investigation and actions taken
 Review effectiveness
13485 CAPA
Preventive Action
The organisation shall determine action to eliminate the
cause of potential nonconformities in order to prevent
their occurrence. Documented procedure for:
 Determine potential nonconformities and their causes
 Evaluate their need for action to prevent occurrence
 Determine and implement action needed
 Record results of any investigations and actions taken
 Review effectiveness
CAPA- Purpose
 Avoid over or under responding to issues
 Set appropriate risk filters into the process
 Supports continuous self-improvement
 50% of FDA actions relate to CAPA
 Important to understand and follow the system
 Poor root cause analysis makes corrective actions a waste
of time
Preventive Action Tools
Failure Mode & Effects Analysis (FMEA)
 An inductive, bottom-up, single point of failure
analysis method aimed at analysing the effects of
failures on equipment or subsystems
 Highly structured, systematic techniques for failure
analysis.
 Good at exhaustively cataloguing initiating faults, and
identifying their local effects.
 It is not good at examining multiple failures or their
effects at a system level.
Failure Modes & Effects Analysis (FMEA)
A tool used to stratify processes for analysis by considering and quantifying all risks
of failure

Each input is given a Risk Priority Number (RPN) based on the following
calculation:

Failure Failure Detection

Effect X Likelihood X Ability = RPN

Rank each element 1-10, yielding a score of 1-1000 RPN

©Loftus Consulting 44
FMEA Template
Function Failure Effects Severity Cause Occurrence Current Detection Critical? RPN Action
Mode Rating Rating Controls Rating
Fill tub High Liquid 8 Sensor 2 Timer 5 No 80 Add 2nd
level overflow failure/ cutout sensor
sensor Discon-
nected

©Loftus Consulting 45
FMEA Example

46
 10 Steps of FMEA
1. Define 1-10 rating scales for the three elements (10 is most severe)
2. List all process steps
3. Identify the failure modes in each step
4. Identify the effects of each failure on the business and/or customer
5. Determine the severity of each failure mode (1-10 scale)
6. List the potential causes of failure
7. Determine the frequency of occurrence (1-10 scale)
8. Identify the controls already in place
9. Access the likelihood of detection (1-10 scale)
10. Multiply the three ratings to get overall Risk Priority Number (RPN)

Focus resources on the area with highest RPN (or greatest severity)

©Loftus Consulting 47
Fault Tree Analysis
 A top-down deductive failure analysis using Boolean
logic to combine a series of lower level events.
 Used to identify best ways to reduce risk and
determine likelihood of failure
 Can be used for creation of diagnostic manuals/
processes
 Can be used to identify and correct causes of an event
 Not generally used to find all possible initiating faults
Fault Tree Logic
FTA Process
 The undesired outcome is the root (top event) of the
tree
 Tree is presented using conventional logic gate
symbols
 Use a single fault tree for every undesired event
 Define the undesired event
 Consider all possible causes, including their probability
of occurrence
 Construct the fault tree (using And and Or gates)
 Evaluate the logic tree
 Control identified hazards to reduce risk
Corrective Action Tools
 The Pareto Principle
- “The 80-20 Rule”
“A small number of causes is responsible for a large percentage of the effect”
- Vilfredo Pareto, 19th century economist and sociologist

E.g.
 80% of land in Italy is owned by 20% of the people (Pareto)
 80% of peas in garden come from 20% of the plants (Pareto)
 For car rentals, the top 0.5% of customers rent 25% of cars.
 In the UK, the top 6% of cola drinkers drink 60% of all colas sold.
 In US, 20% of the population use 80% of the Healthcare Resources
 People wear about 20% of their clothes 80% of the time
 At one Midwest bank (in the US), the top 27% of customers accounted for 100% of profit.

The Pareto principle is used to scope projects down to a manageable size

52
Pareto Example
Customer Complaints

70%

60%

50%

40%

30%

20%

10%

0%
Documents Product Packaging Delivery Other

53
Pareto Example
Document Errors-Occurrence

45%

40%

35%

30%

25%

20%

15%

10%

5%

0%
Quality Cert Quality Cert Invoice Error Packing List IFU Missing
Error Missing Error

What does Pareto tell us to focus on? What should we focus on?
54
5 Why’s
 "... the basis of Toyota's scientific approach ... by repeating why five
times, the nature of the problem as well as its solution becomes clear.“
Taiicho Ohno, Architect of Toyota Production System

 The 5 Whys is a question-asking method used to explore the

cause/effect relationships underlying a particular problem. Ultimately,
the goal of applying the 5 Whys method is to determine a root cause of
a defect or problem.

The key intent is to combine the use of data and knowhow to encourage
the troubleshooter to avoid assumptions
55
 For Want of a Nail!
For want of a nail the shoe is lost;
For want of a shoe the horse is lost;
For want of a horse the rider is lost;
For want of a rider the battle is lost;
For want of a battle the kingdom is lost;
And all for the want of a horseshoe nail.
—George Herbert

©Loftus Consulting
5 Why’s in Practice

Repeat the Question as Often as Necessary to Get to Root Cause

57
 Brainstorming
Brainstorming is a method of efficiently gathering creative ideas by:
 Encouraging lateral thinking
 Using an uninhibited, fun approach to explore a serious topic

Individual Brainstorming:
 For self-employed/ small problems/ initial view
 Better result than with an ineffective group
 May stifle full development of the idea

Group Brainstorming:
 Full use of the creativity and experience of the group
 Ownership of the solution/ good teambuilding
 Must be facilitated well

©Loftus Consulting 58
Individual Brainstorming-Mind Maps

©Loftus Consulting
 Group Brainstorming
 Steps:
1. Assign a facilitator/ scribe
2. Write the process name on a flipchart
3. Give time to participants on their own to generate their ideas at the start
1. Each team member provides an idea (round table or open forum)
2. Scribe writes ideas on the flipchart (no critique)
4. Review list with the team for clarity and ranking
 Rules:
1. Everyone participates
2. NO critique of ideas
3. Keep the pace quick- stop when the pace slows
4. Build on each other’s ideas
5. Evaluate ideas only at the end
6. Have fun!

©Loftus Consulting 60
Role of Facilitator
Prior to Meeting
 Select the right group to address the specific problem:
 Experts from the ‘coalface’
 Technical Experts
 Independent Thinkers
 Give the participants sufficient time to gather their thoughts/ data
 Assess the best way to get all ideas/ inputs from all participants:
 Go-around
 Open discussion
 Data and Experience
 Select the right tool(s) to bring structure to the discussion:
 5 Why’s
 Cause and Effect
Role of Facilitator
During the Meeting
 Define the Problem
 Initial Problem Statement
 Scope: Is-Is Not
 Refine the Problem Definition based on initial discussions

 Gather Factual and Experienced Based Inputs

 Disband and regroup later if data needs to be gathered
 Be aware of Group Dynamics- Gather Everyone’s inputs- Be aware of Groupthink
 Ask facilitating questions to ensure full scope is explored
 Record all ideas- there are no bad ideas
 Keep the pace quick
 Rank Ideas
 Collate same ideas that are expressed in different ways
 Use ‘wisdom of the crowd’ to prioritise top ideas for further action. Carefully select voting
process
Role of Facilitator
End of Meeting/ Follow Up
 Establish Root Cause
 Can occur as part of the meeting or as a subsequent activity
 Assist in data analysis, 5Why’s, DoE’s, etc to investigate initial findings in detail

 Agree Actions
 Prioritise, Assign Responsibility, Set Timeframe

 Communicate
 Inform connected parties
 Seek required support where needed
 Integrate with other systems e.g. CAPA

 Celebrate Success!
 Communicate results
 Beware of Groupthink!
 The Road to Abilene story
 Members’ desire for unanimity may override realistic appraisal
 There are symptoms e.g.
 Illusion of invulnerability
 Stereotyping
 Self-censorship
 There are ways of managing it e.g.
 Leaders should be open to criticism
 The role of devil’s advocate
 Use of outside experts

©Loftus Consulting
 Cause & Effect Diagram
(Fishbone)
Causes:
Measurements Materials Man
Change of supplier New inspector
New Gauge Melt flow index New operator
New ingredient Different shift

EFFECT
Temperature change Change to MOD New equipment

Humidity Change to inspection plan Maintenance

Mother Nature Methods Machines

65
 Cause & Effect Diagram
 Use to add structure to brainstorming, if required
 Minimise risk of ‘blind spots’
 Use in conjunction with 5Why’s/ Brainstorming
techniques
 Lacks a strong prioritisation technique e.g
 Necessary cause (effect cannot occur without it) Vs
 Sufficient cause (effect will occur in its presence)

©Loftus Consulting
C& E (Fishbone) Process
Rules:
 Team brainstorms as described earlier
 Make sure all possible causes are brainstormed by
reviewing all headings on the fishbone
 Capture all possible Causes (not solutions) of a particular
effect under each heading
 Use the team’s knowledge to prioritise focus areas (multi-
voting techniques)

67
 Voting Techniques
1. Counting (e.g. Pareto)
2. Consensus (e.g. Brainstorming)
3. Open Voting (e.g. Red dots)
4. Scoring (e.g. FMEA)
5. Interactive Voting (e.g. Key pads)

©Loftus Consulting
 POKA-YOKE

POKA YOKERU
Inadvertent To Avoid
Mistake

Developed By Shigeo Shingo, Who Also

Developed The SMED Concept

©Loftus Consulting 69
Levels of Error Proofing

-No controls
- Contain/
- Instruction
100% inspect
Awareness - Visual Aids Detection Prevention - Avoidance
- Defect
- Visual
Detection
Controls

©Loftus Consulting 70
Poka-Yoke DVD

71
Key Learnings from DVD
Effective Poka-Yoke:
 Awareness Level: ‘To err is human’
 3 Elements to Detection Level:
 100% effective inspection
 Immediate feedback

 Low cost

 Prevention Level: Not always possible, practical or necessary

Best solutions come proactively from individuals closest to

the operation
72
Identify Error Proofing
 ...in the kitchen
 ...in the car
 ...in the plant

 Identify opportunities for error-proofing in your role

©Loftus Consulting
 Error Proofing in Practice

The window in the envelope is not only a labour saving device. It prevents
the contents of an envelope intended for one person being inserted
in an envelope addressed to another.

©Loftus Consulting 74
 Error Proofing in Practice

Cabinets can fall over if too many drawers are pulled out. For some cabinets,
opening one drawer locks all the rest and so reducing the risk of tipping over.
75

©Loftus Consulting
 Error Proofing in Practice

Parking garages have low clearance. To insure that cars entering the garage will fit,
garages are fitted with a go/no-go gauge at the entrance. Hitting the swinging sign
or pipe will not damage the vehicle as much as driving into a concrete beam.

©Loftus Consulting 76
 Error Proofing in Practice

Even bathroom sinks have a mistake-proofing device. It is the little hole

near the top of the sink that helps prevent overflows.

©Loftus Consulting 77
 Error Proofing in Practice

This stall door is designed so you cannot unlatch the door without moving your coat.

©Loftus Consulting
Error Proofing in Practice
Engineering change form requires different
signatures depending on the nature of the change
being considered. Sometimes engineers would get
too many signatures, and sometimes not enough.

The revised form identifies the nature of the change

in the columns and indicates unnecessary signatures
in gray. Creating forms that help the user fill them
out correctly is a part of mistake-proofing.

©Loftus Consulting
Error Proofing in Practice
Wouldn’t it make more sense to have the
crosswalks line up?
Well, no. In this case, the offset crosswalks and
accompanying railings force pedestrians to turn
and walk toward oncoming traffic. If they look
up at all they will see oncoming vehicles. If not,
instructions are printed on the road:
“LOOK LEFT.”
This cross walk also alerts drivers of the cross
walk by the jagged lane markers.

©Loftus Consulting
 Marketing Errors!
Coors put its slogan, "Turn it loose," into Spanish, where it was read as "Suffer from
diarrhea."
Clairol introduced the "Mist Stick," a curling iron, into German only to find out that "mist"
is slang for manure. Not too many people had use for the "manure stick".
Scandinavian vacuum manufacturer Electrolux used the following in an American
campaign: Nothing sucks like an Electrolux.
When Gerber started selling baby food in Africa, they used the same packaging as in the US,
with the beautiful baby on the label. Later they learned that in Africa, companies routinely
put pictures on the label of what's inside, since most people can't read English.
An American T-shirt maker in Miami printed shirts for the Spanish market which promoted
the Pope's visit. Instead of "I saw the Pope" (el Papa), the shirts read "I saw the potato" (la
papa).
In Italy, a campaign for Schweppes Tonic Water translated the name into "Schweppes Toilet
Water."
Pepsi's "Come alive with the Pepsi Generation" translated into "Pepsi brings your ancestors
back from the grave," in Chinese.
When General Motors introduced the Chevy Nova in South America, it was apparently
unaware that "no va" means "it won't go." After the company figured out why it wasn't selling
any cars, it renamed the car in its Spanish markets to the Caribe.
 Applying the Principles
- Guinness Lite Vs Guinness Mid-Strength

©Loftus Consulting
Bias
Cognitive Dissonance & Hindsight Bias
 Cognitive dissonance refers to a situation involving
conflicting attitudes, beliefs or behaviours. This
produces a feeling of mental discomfort leading to an
alteration in one of the attitudes, beliefs or behaviours
to reduce the discomfort and restore balance
 Hindsight bias is a psychological phenomenon in
which one becomes convinced that one accurately
predicted an event before it occurred. It causes
overconfidence in one's ability to predict other future
events.
Cognitive Dissonance & Hindsight Bias
 ‘ I knew it all along’
 Can lead to overconfidence regarding ability to predict
the outcomes of future events
 Can make it hard to learn from mistakes
 Tendency to avoid information that would create
cognitive dissonance because it is incompatible with
our current beliefs

Less punishment/ more reward needed to break the cycle

Cognitive Dissonance & Hindsight Bias
- A Human Condition
 Fragile egos we want to protect- blame others
 We don’t like random events
 We see things only closest to us
 We default to the most obvious and may miss other
viable explanations

Can lead to dangerous predictions and wrong

conclusions
The Five Guiding Principles of Problem Solving

Customer-
Focused

Team-Based Data-Driven
Guiding
Principles

Documented Right the First Time

Team-Based Principle
Working as a team results in more effective and efficient problem solving

– Characteristics of a high-performance team

• Right people
• Sense of purpose
• Shared leadership
• Trust and mutual respect
• Flexibility and adaptability
• Open communication
• Building on differences
• Continuous learning
 Process Control
Capability Conformance
Analysis Analysis

Eliminate Investigate for

Assignable Cause Assignable Cause

Capability analysis
• What is the currently "inherent" capability of my process when it is "in control"?
• “Have I got a fighting chance to make good product?”

Conformance analysis
• Quality Management System controls daily activities
• Data monitoring identifies when control has likely been lost and special cause variation has occurred

Investigate for assignable cause

• Find “Root Cause(s)” of special variation
• Use one of our Problem Identification tools or Six Sigma

Eliminate assignable cause

• Implement corrective actions to improve process capability
 Thank You!

Contact Details:
[email protected]