Chapter 5 Basic Toolsweb.eng.fiu.edu/leet/TQM/chap5_2012.pdf · 5.6 Pareto Chart • The Pareto...

Chapter 5

Basic Tools

Introduction

• Chapters 2 and 3 discussed histograms, sample mean,

sample standard deviation, attribute/continuous data, and

special/common cause.

• This chapter continues the discussion of basic techniques

and offers a collection of data analysis, data presentation,

and improvement alternatives.

• A wide variety of tools are briefly described for the purpose

of aiding with the efficient building of strategies for

collecting and compiling information that leads to

knowledge. With this knowledge we can make better

decisions.

• Problem identification, defining, and solving.

Introduction

• Descriptive statistics help pull useful information from

data;

• Probability provides, a basis for inferential statistics

and sampling plans.

Introduction

7 management tools (7M tools) or 7 management and

planning tools (7 MP tools).

• affinity diagrams

• interrelationship digraphs

• tree diagrams

• prioritization matrices

• matrix diagrams

• process decision program charts (PDPC)

• activity network diagrams

Introduction

7 quality control tools:

• cause-and-effect diagram

• check sheet

• scatter diagram

• flowchart

• Pareto chart

• histogram

• control chart

Introduction

Tools working with ideas:

• Activity network diagram

• Affinity diagram

• Benchmarking

• Brainstorming

• Cause-and-effect

diagram

• Flowchart

• Force field

• Interrelationship digraph

(ID)

• Matrix diagram nominal

group technique (NGT)

• Prioritization matrices

• Process decision program

chart (PDPC)

• Tree diagram

• Why-why diagram

Introduction

Tools working with numbers:

• Check sheets

• Control chart

• Histogram

• Pareto chart

• Probability plot

• Run chart

• Scatter diagram

5.1 Descriptive Statistics

S4/IEEE Application Examples

• Random sample of last year’s invoices where the number

of days beyond the due date was measured and reported

(i.e., days sales outstanding [DSO])

• Random sample of parts manufactured over the last year,

where the diameters of the parts were measured and

reported


• Sample Mean: Arithmetic average of the data values (x1,

x2, x3 ,..., xi), which is mathematically expressed

𝑥 = 𝑥𝑖𝑛𝑖=1

𝑛

• Sample Median (𝑥_50): the number at (n+1)/2 rank

• TrMean (trimmed mean): Average of the values remaining

after both 5% of the largest and smallest values, rounded

to the nearest integer, are removed.

• Sample standard deviation (SD)

1

𝑛 − 1 (

𝑛

𝑖=1

𝑋𝑖 − 𝑋 )2 =

1

𝑛 − 1( 𝑋𝑖

2

𝑛

𝑖=1

−𝑛𝑋 2)


• SE mean (standard error of mean): 𝑆𝐷

𝑛

• Minimum: lowest number in data set

• Maximum: largest number in data set

• Q1: first quartile

• Q3: third quartile

5.2 Run Chart (Time Series Plot)

S4/IEE Application Examples

• One random paid invoice was selected each day from last

year’s invoices where the number of days beyond the due

date was measured and reported (i.e., days sales

outstanding [DSO]). The DSO for each sample was

plotted in sequence of occurrence on a run chart.

• One random sample of a manufactured part was selected

each day over the last yean where the diameters of the

parts were measured and reported. The diameter for each

sample was plotted in sequence of occurrence on a run

chart.


• A run chart permits the study of observed data for trends or

patterns over time.

• Can be used to compare a performance measurement

before and after a solution implementation to measure its

impact.

• 20~25 points are needed to establish pattern and

baselines.

• Problem exists with the interpretation of run chart: all

variation as important (over-reacting) control charts


www.pqsystems.com/.../chart_BasicRunChart.png

5.3 Control Chart


• One paid invoice was randomly selected each day from

last year’s invoices where the number of days beyond the

due date was measured and reported (i.e., days sales

outstanding [DSO]). The DSO for each sample was

plotted in sequence of occurrence on a control chart.

• One random sample of a manufactured part was selected

each day over the last year, where the diameters of the

parts were measured and reported. The diameter for each

sample was plotted in sequence of occurrence on a control

chart.

5.3 Control Chart

• Control limits are a function of data variability, not

specification limits.

• It gives not only process monitoring and control, but also

direction for improvements.

• It can separate special cause from common causes

• Early identification of special causes

• 94% of the troubles belong to system (common cause);

only 6% are special cause (Deming 1986)

• Monitoring should be of KPIVs (not only on KPOVs)

• If process is shown to be in-control, issues should be

looked collectively over some period of time Process

Improvement techniques

5.3 Control Chart

http://support.sas.com/rnd/app/qc/qc/shwclms.gif

5.4 Probability Plot

• Probability plots are most often associated with tests to

assess the validity of normality assumptions.

• When data are a straight line on a normal probability plot,

the data are presumed to be from a normal distribution.

• Probability plots also apply to other distributions, such as

the Weibull distribution.

• Probability plots can also be used to make percentage of

population statements (very useful in describing the

performance of business and other processes)

5.4 Probability Plot

Minitab:

Graph

Probability Plot

• Single

5.5 Check Sheets

• Check sheets contain the systematic recording and

compiling of data from historical or current observations.

• The information can indicate patterns and trends.

• After agreement is reached on the definition of events or

conditions, data are collected over a period of time and

presented in tabular form.

Problem Week

Total 1 2 3

A /// //// // 10

B / // // 5

C //// / / 6

5.5 Check Sheets

http://www.asq.org/learn-about-quality/data-

collection-analysis-tools/overview/check-sheet.html

http://www.asq.org/learn-about-quality/data-collection-analysis-tools/overview/check-sheet.html













5.6 Pareto Chart


• Transactional workflow metric (could similarly apply to

manufacturing; e.g., inventory or time to complete a

manufacturing process): Random sample of last year’s

invoices where the number of days beyond the due date

was measured and reported (i.e., days sales outstanding

[DSO] ). If an invoice was beyond 30 days late, it was

considered a failure or defective transaction. A Pareto

chart showed the frequencies of delinquencies by

company invoiced.

5.6 Pareto Chart


• Transactional quality metric: Random sample of last year’s

invoice, where the invoices were examined to determine if

there were any errors when filling out the invoice or within

any other step of the process. Multiple errors or defects

could occur when executing an invoice. The total number

of defects when the invoice was executed was divided by

the total number of opportunities for failure to estimate the

defect per million opportunity (DPMO) rate of the process.

A Pareto chart showed the frequencies of delinquencies by

type of failure.

5.6 Pareto Chart


• Manufacturing quality metric: Random sample of printed circuit

boards over the last year, where the boards were tested for failure.

The number of defective boards was divided by the sample size to

estimate the defective rate of the process. A Pareto chart showed

the frequencies of defective units by printed circuit board type.

• Manufacturing quality metric: Random sample of printed circuit

boards over the last year, where the boards were tested for failure.

Multiple failures could occur on one board. The total number of

defects on the boards was divided by the total number of

opportunities for failure (sum of the number of components and

solder joints from the samples) to estimate the defect per million

opportunity (DPMO) rate of the process. A Pareto chart showed

the frequencies of defects by failure type.

5.6 Pareto Chart

• The Pareto principle basically states that a vital few of the

manufacturing process characteristics cause most of the

quality problems on the line, while a trivial many of the

manufacturing process characteristics cause only a small

portion of the quality problems.

• Pareto chart may need to consider data from different

perspectives. [by machine, or by shift]

5.6 Pareto Chart

Coffee Type

Coffee Amount

Grind Time

Brew Time

Water Temp

Cup SizeCup Type

Count 320 280 176 174 144 112 104

Percentage 24.4% 21.4% 13.4% 13.3% 11.0% 8.5% 7.9%

Cumulative % 24.4% 45.8% 59.2% 72.5% 83.5% 92.1% 100.0%

0.0%

10.0%

20.0%

30.0%

40.0%

50.0%

60.0%

70.0%

80.0%

90.0%

100.0%

0

200

400

600

800

1000

1200

1400

RA

NK

Coffee Pareto Chart

5.6 Pareto Chart: Procedure

1. Define the problem and process characteristics to use in the

diagram.

2. Define the period of time for the diagram—for example,

weekly, daily, or shift.

3. Total the number of times each characteristic occurred.

4. Rank the characteristics according to the totals from step 3.

5. Plot the number of occurrences of each characteristic in

descending order in a bar graph form along with a

cumulative plot of the magnitudes from the bars.

6. Trivial columns can be lumped under one column

designation; however, care must be exercised not to forget

a small but important item.

5.7 Benchmarking

• Benchmarking involves the search of an organization for the

best practices, adaptation of the practices to its processes,

and improving with the focus of becoming the best in class.

• Benchmarking can involve comparisons of products,

processes, methods, and strategies.

• Sources of information for benchmarking include the

Internet, in-house published material, professional

associations, universities, advertising, and customer

feedback.

http://totalqualitymanagement.files.wordpress.com

/2008/11/picture24.png

5.7 Types of Benchmarking

• Internal benchmarking makes comparisons between similar

operations within an organization.

• Competitive benchmarking makes comparisons with the

best direct competitor.

• Functional benchmarking makes comparisons of similar

process methodologies.

• Generic benchmarking makes comparisons of processes

with exemplary and innovative processes of other

companies.

5.8 Brainstorming

• Very valuable means of generating new ideas and involving

a group.

• Many ways

• to conduct a brainstorming session

• to compile the information from the session

http://www.howtoplaza.com/wp-

content/uploads/2010/05/productive-brainstorming-session.png

5.8 Brainstorming

Formal process of brainstorming

• Setup:

• Table arranged in a manner to encourage discussion

• Problem or question is written down for everyone to see

• Basic rules:

1. Ask each member in rotation for one idea

2. Rule out all evaluations or critical judgments

3. Encourage wild ideas

4. Encourage good-natured laughter and informality

5. Target for quantity, not quality

6. Look for improvements and combinations of ideas

5.8 Brainstorming

Guideline for the leader:

1. The problem needs to be simply stated.

2. Two or more people should document the ideas in plain sight so

that the participants can see the proposed ideas and build on the

concepts.

3. The name of the participant who suggested the idea should be

placed next to it.

4. Ideas typically start slowly and build speed. Change in speed

often occurs after someone proposes an offbeat idea. This

change typically encourages others to try to surpass it.

5. A single session can produce over 100 ideas, but many will not

be practical.

6. Many innovative ideas can occur after a day or two has passed.

5.8 Brainstorming

• A follow-up session can be used to sort the ideas into

categories and rank them.

• When ranking ideas, members vote on each idea that they

think has value.

• For some idea considerations it is beneficial to have a

discussion of the pros and cons before the vote

• A circle is drawn around the ideas receive the most votes

• Through sorting and ranking, many ideas can be combined

while others are eliminated

5.8 Brainstorming Applications

• Problem definition

• Factors within a DOE

• Test strategy

• Inputs to the cause-and-effect diagram

5.9 Nominal Group Technique

(NGT)

Nominal group technique expedites team consensus on relative

importance of problems, issues, or solutions.

• An NGT is conducted by displaying a generated list of items,

perhaps from a brainstorming session, on a flipchart or board.

• Eliminating duplications and making clarifications, then creates a

final list.

• The new final list of statements is then prominently displayed,

each item is assigned a letter, A, B ,...

• On a sheet of paper, each person ranks the statements, assigning

the most important a number equal to the number of statements

and the least important the value of one.

• Results from the individual sheets are combined to create a total

overall prioritization number for each statement.

5.9 Nominal Group Technique

(NGT)

http://www.wsa-intl.com/Portals/70018/images/202.jpg

5.10 Force Field Analysis

Force field analysis can be used to analyze what forces in an

organization are supporting and which are restraining progress.

• After an issue or problem is identified, a brainstorming session is

conducted to create a list of driving forces and then a list of

restraining force.

• A prioritization is then conducted of the driving forces that could be

strengthened.

• There is then a prioritization of the restraining forces that could be

reduced to better achieve the desired result.

5.10 Force Field Analysis

Restraining Forces Action Items

Programs reinforce the old Create reward and recognition programs that reinforce

the new

People do not have skills sets needed

to succeed

Create training programs so that people have the

opportunity to learn new

Phrases still refer to the old Introduce new vocabulary so people know you are

speaking of the new

Current cultural supports staying the

same Change culture so that it supports the new

People concerned about making

mistakes

Reassure them that mistakes are learning

opportunities and failure will not be punished

People concerned how changes

affect their job

Provide information that clarifies how performance will

be evaluated

People isolated in the ways they are

going about the new Celebrate successes

People wonder how serious the

company is about this change

Provide consistent communications that clarify how

serious the company is about completing this change

5.11 Cause-and-Effect Diagram


• An S4/IEE project was created to improve the 30,000 foot-

level metric days sales outstanding (DSO). A process

flowchart was created to describe the existing process. A

team created a cause-and-effect diagram in a

brainstorming session to trigger a list of potential causal

inputs and improvement ideas.

• An S4/IEE project was created to improve the 30,000 foot-

level metric, the diameter of a manufactured part. A

process flow chart was created to describe the existing

process. A team created a cause-and-effect diagram in a


inputs and improvement ideas.


• also known as an Ishikawa diagram (after its originator

Karoru Ishikawa) or fishbone diagram.

• This technique is useful to trigger ideas and promote a

balanced approach in group brainstorming sessions in

which individuals list the perceived sources (causes) of a

problem (effect).

• A cause-and-effect diagram provides a means for teams to

focus on the creation of a list of process input variables

that could affect key process output variables.

• Addresses strata issues based on key characteristics (e.g.,

who, what, where, and when).


• When constructing a cause-and-effect diagram, it is often

appropriate to consider six areas of causes that can

contribute to an effect: materials, machine, method,

personnel, measurement, and environment.

• Each one of these characteristics is then investigated for

sub-causes. Sub-causes are specific items or difficulties

that are identified as a factual or potential cause to the

problem (effect).


Variations in creating a cause-and-effect diagram:

• A team may choose to emphasize the most likely causes

by circling them.

• it can also be beneficial to identify noise factors (n) (e.g.,

ambient room temperature and a raw material

characteristic that cannot be controlled), and factors that

can be controlled (c) (e.g., process temperature or speed)

by placing the letter n or c next to the named effect.

• Include score with an importance and ease-of-resolution

matrix.

Cause-Effect (CE) Analysis

www.syncfusion.com/.../img/Fishbone_larger.png


Impact

Impl High Low

Easy 1 2

Hard 3 4

5.12 Affinity Diagram

Affinity diagram can organize and summarize the natural

grouping from a large number of ideas and issues.

• Boldly record each brainstorming idea individually on a

post note, using at a minimum a noun and verb.

• Next, place the post note on a wall and ask everyone,

without talking, to move the notes to the place where they

think the issue best fits.

• Upon completion of this sorting, create a summary or

header sentence for each grouping. (Create subgroups for

large groupings as needed with a subhead description.)

• Connect all finalized headers with their groupings by

drawing lines around the groupings.

Affinity

Diagram

http://www.asq.org/learn-

about-quality/idea-

creation-

tools/overview/affinity.html

http://www.asq.org/learn-about-quality/idea-creation-tools/overview/affinity.html









5.12 Affinity Diagram

Infrastructure

• Align projects with business needs

• Create system to pull projects from business metrics needs

• Establish project accountability

• Plan steering committee meetings

• Select champions, sponsors, and team leaders

• Determine strategic projects and metrics

• Communication plans

• Incentive plans

• Schedule project report outs

• Champion/sponsor training

• Compile lessons learned from past projects

Project Execution

• Project scoping

• Project approval

• Phase report outs

• Project closure

• Project leveraging

Training

• Champion training

• Black belt training

• Green belt training

• Use training material

w/roadmap

Culture

• Create buy-in

• Evaluate obstacles

and facilitate change

• Integrate 6σ into

daily activities

• Communication plan

5.13 Interrelationship Digraph (ID)

Interrelationship digraph permits systematic identification,

analysis, and classification of cause-and-effect relationships,

enabling teams to focus on key drivers or outcomes to

determine effective solutions.

• Assemble a team of 4 to 6 members

• Arrange the 5-25 items from another tool (e.g., an affinity

diagram) in a circular pattern on a flipchart.

• Draw relationship between the items by choosing any one

of the items as a starting point, with a stronger cause or

influence indicated by the origination of an arrow.


• Upon the completion of a chart, get additional input from

others and then tally the number of outgoing and input

arrows for each item.

• A high number of outgoing arrows indicates that the item is

a root cause or driver that should be addressed initially.

• A high number of incoming arrows indicates a key outcome

item.

• A summary ID shows the total number of incoming and

outgoing arrows next to each item.

• Driver and outcome items can be highlighted using a

double box or bold box.


http://www4.asq.org/blogs/statistics/Images/Interrelationship.jpg

5.14 Tree Diagram

• Tree diagrams can help people uncover, describe, and

communicate a logical relationship that is hierarchical

between important events or goals.

• Similarly, a tree can describe the hierarchy that leads to a

desirable or undesirable event. (fault tree or FT).

• With this approach a big idea or problem is partitioned into

smaller components.

• Logical operators such as AND or OR gates can connect

lower elements to higher elements in the hierarchy.

5.14 Tree Diagram

5.14 Tree Diagram

http://www.conceptdraw.com/resources/images/solutions-

screens/business/Root_Cause_Tree_Diagram.png

5.15 Why-Why Diagram

• A variation of the cause-and-effect diagram and tree

diagram (Higgins 1994; Majaro 1988).

• These final responses can be used as wisdom of the

organization inputs for further investigation.

Why-

Why

Diagram

http://www.syque.com/improvement/images/image343.gif

5.16 Matrix Diagram and

Prioritization Matrices

• A matrix diagram is useful to discover relationships

between two groups of ideas.

• A prioritization matrix quantifies and prioritizes items within

a matrix diagram: activities, goals, or characteristics

• Applications: cause-and-effect matrix and quality function

deployment (QFD)

5.16 Matrix Diagram and

Prioritization Matrices

• Within a prioritization matrix, one can assign relative

importance weights.

• Simply assigned by the organization or team.

• Other techniques: The analytical hierarchy process

(AHP)

• Within the AHP approach, a number of decision-

makers can integrate their priorities into a single

priority matrix using a pairwise fashion.

• This result of this matrix is a prioritization of the

factors.

5.16 Analytical Hierarchy Process

(AHP)

A B C D E

A. Fire in the belly B2 A2 A2 A3

B. Soft skill B1 B2 B3

C. Project management C1 C3

D. Analytical skill D2

E. Statistical knowledge

A B C D E

A. Fire in the belly 1 1/2 2/1 2/1 3/1

B. Soft skill 1 1/1 2/1 3/1

C. Project management 1 1/1 3/1

D. Analytical skill 1 2/1

E. Statistical knowledge 1


(AHP)

A B C D E

A. Fire in the belly 1 1/2 2/1 2/1 3/1

B. Soft skill 2/1 1 1/1 2/1 3/1

C. Project management 1/2 1/1 1 1/1 3/1

D. Analytical skill 1/2 1/2 1/1 1 2/1

E. Statistical knowledge 1/3 1/3 1/3 1/2 1

A B C D E

A. Fire in the belly 1.0000 0.5000 2.0000 2.0000 3.0000

B. Soft skill 2.0000 1.0000 1.0000 2.0000 3.0000

C. Project management 0.5000 1.0000 1.0000 1.0000 3.0000

D. Analytical skill 0.5000 0.5000 1.0000 1.0000 2.0000

E. Statistical knowledge 0.3333 0.3333 0.3333 0.5000 1.0000


(AHP)

A B C D E Sum Weights

A. Fire 5.0000 5.0000 7.5000 8.5000 17.5000 43.5000 0.2652

B. Soft 6.5000 5.0000 9.0000 10.5000 19.0000 50.0000 0.3049

C. Project 4.5000 3.7500 5.0000 6.5000 12.5000 32.2500 0.1966

D. Ana skill 3.1667 2.9167 4.1667 5.0000 10.0000 25.2500 0.1540

E. Stat 1.7500 1.4167 2.1667 2.6667 5.0000 13.0000 0.0793

164.0000

5.17 Process Decision Program Chart

(PDPC)

• A process decision program chart (PDPC) helps with the

organization and evaluation of processes and the creation

of contingency plans.

• PDPC can help anticipate deviations from expected events

and provide insight to the creation of effective contingency

plans.

• PDPC can help determine the impact of problems or

failures on project schedules. Specific actions can be

undertaken for problem prevention or mitigation of impact

when they do occur.

• Subjective probabilities of occurrence can be assigned and

then used for the assignment of priorities.


(PDPC)

Berger et al. (2002) highlight the following steps, which are

common to all PDPC formats:

• Identify the process purpose.

• Identify the basic activities and related events associated

with the process.

• Annotate the basic activities and related events.

• Superimpose the possible (conceivable) deviations.

• Annotate the possible deviations.

• Identify and annotate contingency activities.

• Weight the possible contingencies.


(PDPC)


(PDPC)

http://www.syque.com/quality_tools/tools/TOOL

S12_files/image002.gif

5.18 Activity Network Diagram or

Arrow Diagram

• Activity network diagrams (arrow diagrams) help with the

definition, organization, and management of activities with

respect to time.

• The arrow diagram is used in program evaluation and

review technique (PERT) and critical path method (CPM)

methodologies.

Orientation

1

2

4

6

7 3 5 9

8

Manual

Testing

Dummy

System

Training

Dummy

System

Testing

Position

recruiting

System

development

Equipment

installation

Equipment testing

and modification

Final

debugging

System

changeover

Job

training

5.18 Activity Network Diagram or

Arrow Diagram

5.19 Scatter Diagram (Plot of Two Variables)


• An S4/IEE project was created to improve the 30,000-foot-

level metric, days sales outstanding (DSO). A process

flowchart was created to describe the existing process. A

team created a cause-and-effect diagram in a


inputs and improvement ideas. One input that the team

thought could be an important cause was the size of the

invoice. A scatter diagram of DSO versus size of invoice

was created.



• An S4/IEE project was created to improve the 30,000-foot-

level metric, the diameter of a manufactured part. A

process flowchart was created to describe the existing

process. A team created a cause-and-effect diagram in a


inputs and improvement ideas. One input that the team

thought could be an important cause was the temperature

of the manufacturing process. A scatter diagram of part

diameter versus process temperature was created.


• A scatter diagram (plot) assesses the relationship between

two variables (follow-up procedure to validate the

consensus relationship from a cause-and-effect diagram)

• 50 to 100 pairs of samples should be plotted

• the independent variable is on the x-axis while the

dependent variable is on the y-axis.

• A scatter diagram relationship does not predict a true

cause-and-effect relationship.

• The plot only shows the strength of the relationship

between two variables

• The correlation and regression techniques can be used to

test the statistical significance of relationships.

5.20 Example 5.1:

Improving a Process that has Defects

5.20 Example 5.1:


• From p-chart, the process is in control; no special causes

are noted.

• Defect rate is too high

5.20 Example 5.1:


From 3200

manufactured

PCBs

5.20 Example 5.1:


• A brainstorming session with experts in the field could then

be conducted to create a cause-and-effect diagram for the

purpose of identifying the most likely sources of the

defects.

• Regression analysis followed by a DOE might then be

most appropriate to determine which of the factors has the

most impact on the defect rate.

• After Changes are made to the process

• New Pareto chart to identify the next major cause

• Control charts to monitor the “insufficient solder”

problem

5.21 Example 5.2:

Reducing Total Cycle Time of A Process

• Development cycle time: 2~3 years

• Survey indicates the biggest deterrent is “procurement of

standard equipment and components”.

• Loss time and revenue due to delay is 22 times higher than

the equipment itself.

5.22 Example 5.3:

Improving A Service Process

• cycle time in a claim processing center, accounts payable

organization, or typing center.

• Current process flow

• Cause-and-effect diagram for improvement opportunities

• Pareto chart: simple solution

• Benefit/Cost analysis (B/C96)

Chapter 5 Basic Toolsweb.eng.fiu.edu/leet/TQM/chap5_2012.pdf · 5.6 Pareto Chart • The Pareto...

Documents

Transcript of Chapter 5 Basic Toolsweb.eng.fiu.edu/leet/TQM/chap5_2012.pdf · 5.6 Pareto Chart • The Pareto...