The+Memory+Jogger+II

TheMemory

Jogger™ IIA Pocket Guide of Tools

for Continuous Improvement& Effective Planning

First Edition

Michael Brassard & Diane Ritter

GOAL/QPC

This handbook is designed to help you and everyperson in your organization to improve daily the

procedures, systems, quality, cost, and yields related toyour job. This continuous improvement process is thefocus of this book—sharing the philosophy and tools

that are fundamental to this effort.

MJII Frontmatter 06/14/01 10/7/03, 9:33 AM1

The Memory Jogger™ was first written, edited, and pro-duced in 1985 by Michael Brassard and Diane Ritter ofGOAL/QPC. They have, in the name of continuous im-provement, worked together again in substantially revis-ing the original tools and writing new material to create amore readable, user-friendly, and comprehensive refer-ence pocket manual. The Memory Jogger™ II now includesthe Seven Quality Control Tools, the Seven Managementand Planning Tools, and a problem-solving case studyexample, which are presented in a format that (we hope)will allow you to find relevant information so easily, you’llalways find space for it in your pocket or pocketbook.

© 1994 by GOAL/QPCAny reproduction of any part of this publication without

the written permission of GOAL/QPC is prohibited.

Edited by Francine OddoCover design by Lori Champney

Layout production by Michele KiersteadGraphics created by Deborah Crovo

GOAL/QPC12B Manor Parkway, Salem, NH 03079-2862

Toll free: 800-643-4316 or 603-893-1944Fax: 603-870-9122

E-mail: [email protected] site: www.goalqpc.com

GOAL/QPC is a nonprofit organization that helps compa-nies continually improve their Quality, Productivity, andCompetitiveness. (Hence, the “QPC” in our name.)

Printed in the United States of AmericaFirst Edition20 19 18

ISBN 1-879364-44-1


http://www.goalqpc.com

mailto:[email protected]

AcknowledgmentsOur sincerest appreciation and thanks go to our dedicated,knowledgeable, and tireless friends who have worked sohard to make this little book a BIG success.

Production & Administrative Staff: Lee Alphen,Steve Boudreau, Paul Brassard, Lori Champney,Deborah Crovo, Brian Hettrick, Michele Kierstead,Fran Kudzia, Larry LeFebre, Francine Oddo,Dorie Overhoff, & Debra Vartanian.

Concept & Content Reviewers: Del Bassett, Casey Collett,Ann McManus, Bill Montgomery, Bob Porter, & Paul Ritter.

Customers & Collaborators: To the customer focus groupsfor their insights and suggestions.

Companies Who Contributed Information for Illustrations:(listed in order of the company name)Bryce Colburne, AT&T TechnologiesRobert Brodeur, Bell CanadaKaren Tate, BGPRip Stauffer, BlueFire PartnersWeston F. Milliken, CUE ConsultingPaula McIntosh, Georgia State Dept. of Human Resources,

Div. of Rehabilitation ServicesDonald Botto, Goodyear Tire & Rubber CompanyTony Borgen & Paul Hearn, Hamilton StandardWalter Tomschin, Hewlett PackardDr. Verel R. Salmon, Millcreek Township School District,

Millcreek Township, PAJack Waddell, Novacor ChemicalsKirk Kochenberger, Parkview Episcopal Medical CenterMarie Sinioris, Rush-Presbyterian-St. Luke’s Medical CenterTim Barnes, SmithKline BeechamDennis Crowley, St. John Fisher CollegeBuzz Stapczynski, Town of Andover, MAKemper Watkins, MSgt., U.S. Air Force, Air Combat CommandCapt. T. C. Splitgerber, U.S. Navy, Naval Dental Center, San Diego


How to Use The Memory Jogger™ IIThe Memory Jogger™ II is designed for you to use as aconvenient and quick reference guide on the job. Put yourfinger on any tool within seconds!

Know the tool you need? Find it by using the:

Table of Contents. Tools, techniques, the case study—it’sin alphabetical order.Solid tab. This tab runs off the edge of the first page of eachnew section.

Not sure what tool you need? Get an idea by using the:

Tool Selector Chart. This chart organizes the tools bytypical improvement situations, such as working withnumbers, with ideas, or in teams.

What do the different positions of runners mean?

Getting Ready—An important first step is to selectthe right tool for the situation. When you see the“getting ready” position of the runner, expect a briefdescription of the tool’s purpose and a list of benefitsin using the tool.

Cruising—When you see this runner, expect to findconstruction guidelines and interpretation tips. Thisis the action phase that provides you with step-by-step instructions and helpful formulas.

Finishing the Course—When you see this runner,expect to see the tool in its final form. There areexamples from manufacturing, administration/service, and daily life to illustrate the widespreadapplications of each tool.

iv


ContentsAcknowledgments ........................................................ iiiHow to Use The Memory Jogger™ II ........................... ivTool Selector Chart ....................................................... viIntroduction ..................................................................... 1Activity Network Diagram (AND) .............................. 3

• Gantt Chart .......................................................... 9Affinity Diagram ........................................................... 12Brainstorming ................................................................ 19Cause & Effect/Fishbone Diagram ............................ 23Check Sheet .................................................................... 31Control Chart ................................................................. 36Data Points ..................................................................... 52Flowchart ........................................................................ 56Force Field Analysis ..................................................... 63Histogram ....................................................................... 66Interrelationship Digraph (ID) ................................... 76Matrix Diagram ............................................................. 85Nominal Group Technique (NGT) ............................. 91

• Multivoting ........................................................ 93Pareto Chart ................................................................... 95Prioritization Matrices ............................................... 105Problem-Solving/Process-Improvement Model:

Improvement Storyboard ................................. 115Process Capability ....................................................... 132Radar Chart .................................................................. 137Run Chart ..................................................................... 141Scatter Diagram ........................................................... 145Team Guidelines ......................................................... 150

• Team Stages ..................................................... 150• Conducting Effective Meetings .................... 154

Tree Diagram ............................................................... 156• Process Decision Program Chart (PDPC) .. 160

v


vi

Tool Selector ChartThis chart organizes the tools by typical improvementsituations, such as working with numbers, with ideas, orin teams.

Workingwith Ideas P

age

#

Gen

erat

ing/

Gro

upin

g

Dec

idin

g

Impl

emen

ting

AND

AffinityBrainstorming

C & E/Fishbone

NGT/Multivoting

FlowchartForce Field

Gantt

IDMatrix

PrioritizationPDPCRadar

Tree

312192356639768591105160137156


vii

Working withNumbers P

age

#

Cou

ntin

g

Mea

sure

s

Check SheetControl Charts

Data PointsHistogram

Pareto

Process CapabilityRun

Scatter

#

3136526695

132141145

Working in Teams Pag

e #

Impr

ovem

ent R

oadm

ap

Team

Roa

dmap

Storyboard Case Study

Starting TeamsMaintaining Teams

Ending Teams/ProjectsEffective Meetings

115150151153154


IntroductionIn classrooms to board rooms, on manufacturing

floors and in medical centers, organizations around theworld are using continuous quality improvement (CQI) astheir strategy to bring about dramatic changes in theiroperations. Their purpose is to stay competitive in aworld of instant communication and technologicaladvancement.

These organizations need to meet or exceedcustomer expectations while maintaining a cost-competitive position. Continuous quality improvement(CQI), a systematic, organization-wide approach forcontinually improving all processes that deliver qualityproducts and services, is the strategy many organiza-tions are adopting to meet today’s challenges and toprepare for those down the road.

In pursuing CQI, stick to these four basic principles:

1. Develop a strong customer focus2. Continually improve all processes3. Involve employees4. Mobilize both data and team knowledge to

improve decision making

1. Develop a Strong Customer FocusTotal customer focus includes the needs of bothexternal and internal customers. External custom-ers are the end users, internal customers are yourcoworkers and other departments in the organiza-tion.

2. Continually Improve All Processes• Identify them. A process is a sequence of repeat-

able steps that lead to some desired end oroutput: a typed document, a printed circuitboard, a “home-cooked” meal, arrival at work,and so on.

1


EquipmentPeople

MaterialsPolicies

Environment

Processstep

Product orservice

Desired endproduct or

service• • •

InputsValue Add Output/Input Final Output

• Improve them. Use the Plan, Do, Check, Act(PDCA) Cycle: PLAN what you want to accom-plish over a period of time and what you mightdo, or need to do to get there. DO what youplanned on doing. Start on a small scale!CHECK the results of what you did to see if theobjective was achieved. ACT on the informa-tion. If you were successful, standardize theplan, otherwise continue in the cycle to plan forfurther improvement.

3. Involve EmployeesEncourage teams—train them—support them—use their work—celebrate their accomplishments!

4. Mobilize Both Data and Team Knowledge toImprove Decision MakingUse the tools to get the most out of your data and theknowledge of your team members.• Graphically display number and word data;

team members can easily uncover patternswithin the data, and immediately focus on themost important targets for improvement.

• Develop team consensus on the root cause(s) ofa problem and on the plan for improvement.

• Provide a safe and efficient outlet for ideas at alllevels.

Use this book and the tools in it to focus on, improve,involve employees in, and direct your path towardcontinuous quality improvement.

2


©1994 GOAL/QPC Activity Network/Gantt 3

Activity NetworkDiagram (AND)Scheduling sequential& simultaneous tasks

Why use it?To allow a team to find both the most efficient path andrealistic schedule for the completion of any project bygraphically showing total completion time, the neces-sary sequence of tasks, those tasks that can be donesimultaneously, and the critical tasks to monitor.

What does it do?• All team members have a chance to give a realistic

picture of what their piece of the plan requires,based on real experience

• Everyone sees why he or she is critical to the overallsuccess of the project

• Unrealistic implementation timetables are dis-covered and adjusted in the planning stage

• The entire team can think creatively about how toshorten tasks that are bottlenecks

• The entire team can focus its attention and scarceresources on the truly critical tasks

How do I do it?1. Assemble the right team of people with firsthand

knowledge of the subtasks

2. Brainstorm or document all the tasks needed tocomplete a project. Record on Post-its™

Activity Network/Gantt 3

AND 3-11N 10/7/03, 9:17 AM3

©1994 GOAL/QPC4 Activity Network/Gantt

3. Find the first task that must be done, and place thecard on the extreme left of a large work surface

Determinetarget

audience fornew topic

Job/Task card Post-it™ Notes

4. Ask: “Are there any tasks that can be done simul-taneously with task #1?”• If there are simultaneous tasks, place the task card

above or below task #1. If not, go to the next step.

5. Ask, “What is the next task that must be done?Can others be done simultaneously?”• Repeat this questioning process until all the recorded

tasks are placed in sequence and in parallel.

Tip At each step always ask, “Have we forgottenany other needed tasks that could be done simul-taneously?”

Determinetarget

audiencefor

new topic Assesscom-

petitor’sofferings

Reviewfeedback

fromsimilar

courses

AND 3-11N 10/7/03, 9:17 AM4


6. Number each task, draw the connecting arrows,and agree on a realistic time for the completion ofeach task• Record the completion time on the bottom half of

each card.

1. Determine target

audience for new topic

3. Assess competitor's

offerings

2. Review feedback from

similar courses

T = 14 days

T = 21 days

T = 7 days

Tip Be sure to agree on the standard time unit foreach task, e.g., days, weeks. Elapsed time is easierthan “dedicated” time, e.g., 8 hours of dedicatedtime versus 8 hours over a two-week period(elapsed time).

7. Determine the project’s critical path• Any delay to a task on the critical path will be added

to the project’s completion time, unless anothertask is accelerated or eliminated. Likewise, theproject’s completion time can be reduced by accel-erating any task on the critical path.

• There are two options for calculating the totalcritical path and the tasks included within it.Longest cumulative path. Identify total projectcompletion time. Add up each path of connectedactivities. The longest cumulative path is the

AND 3-11N 10/7/03, 9:17 AM5


quickest possible implementation time. This isthe project’s critical path.

Calculated slack. Calculate the “slack” in the start-ing and completion times of each task. This iden-tifies which tasks must be completed exactly asscheduled (on the critical path) and those that havesome latitude.

1. Determine target

audience for new topic

3. Assess competitor's

offerings

2. Review feedback from

similar courses

T = 14 days

T = 21 days

T = 7 days

14 35

14 35

Earliest Start (ES)

Earliest Finish (EF)

LatestStart(LS)

LatestFinish(LF)

0 140 14

14 2128 35

Finding the critical path by calculating the slack

ES = The largest EF of any previous connected task

EF = ES + the time to complete that task

LS = LF - the time to complete that task

LF = The smallest LS of any connected following task

When ES = LS AND EF = LF, that task is on the critical path, and therefore there is no schedule flexibility in this task.

AND 3-11N 10/7/03, 9:17 AM6


Tip Determining the longest cumulative path issimpler than calculating the slack, but canquickly become confusing in larger ANDs.The calculated slack option determines the totalproject and slack times; and therefore the totalproject time and critical path are identified “auto-matically.”

Pro

ject

dur

atio

n: 1

61 d

ays

Dev

elop

ing

a N

ew C

ours

e

Pro

ject

S

tart

1. D

eter

min

e ta

rget

au

dien

ce fo

r ne

w to

pic

2. R

evie

w

feed

back

fr

om s

imila

r co

urse

s

4. D

evel

op

cour

se

obje

ctiv

es

6. D

evel

op

cour

se

form

at

3. A

sses

s co

mpe

titor

's

offe

rings

5. C

hoos

e ge

ogra

phic

lo

catio

n fo

r fin

al c

ours

e

T =

14

days

0 1

40

14

T =

7

days

14

2128

35

T =

21

days

14

3514

35

T =

2

days

35 3

754

56

T =

14

days

42 5

642

56

T =

7

days

35 4

235

42

1

See

ne

xt

page

AND 3-11N 10/7/03, 9:17 AM7


7. W

rite

draf

t co

nten

t9.

Dev

elop

vi

sual

aid

s11

. Con

duct

pi

lot o

fferin

gs

8. D

evel

op

mar

ketin

g st

rate

gy

T =

30

days

56

8656

86

T =

30

days

86 1

1686

116

T =

21

days

116

137

116

137

13. M

odify

dr

aft m

ater

ial

T =

21

days

137

158

137

158

T =

7da

ys56

63

71 7

8

10. S

ecur

e a

trai

ning

site

T =

14

days

63 7

778

92

12. D

evel

op

and

prin

t br

ochu

reT

= 2

1da

ys77

98

92 1

13

14. D

istr

ibut

e co

urse

br

ochu

reT

= 4

5da

ys 9

8 1

4311

3 15

8

15. C

ondu

ct

trai

ning

T =

3da

ys15

8 1

6115

8 1

61

Pro

ject

F

inis

h1

from

pr

evio

us

page

Dev

elop

ing

a N

ew C

ours

e

AND 3-11N 10/7/03, 9:17 AM8


VariationsThe constructed example shown in this section is in theActivity on Node (AON) format. For more information onother formats such as Activity on Arrow (AOA) and Prece-dence Diagram (PDM), see The Memory Jogger Plus+®.

Another widely used, schedule-monitoring method is theGantt chart. It is a simple tool that uses horizontal bars toshow which tasks can be done simultaneously over the lifeof the project. Its primary disadvantage is that it cannotshow which tasks are specifically dependent on each other.

Information provided courtesy of BGP

Qtr

4, 1

993

Qtr

1, 1

994

Qtr

2, 1

994

Qtr

3, 1

994

IDN

ame

1 2 3 4 5 6 7 8 9 10 11 12 13 14

ISO

orie

ntat

ion

Pla

n tra

inin

gA

udito

r tra

inin

gP

lan

pre-

audi

t

Pre

-aud

it

Pre

-aud

it pr

ep.

Pre

-aud

it re

port

Cor

rect

ive

actio

n

Gen

erat

e sc

hedu

leP

lan

audi

t

Aud

it pr

epar

atio

n

Aud

it re

port

Oct

. N

ov.

Dec

.Ja

n. F

eb.

Mar

.A

pr.

May

Jun

.Ju

l. A

ug.

Sep

.

1/3

1/21

1/28

ISO

900

0 A

udit

Sch

edul

e

Rel

ease

doc

umen

tatio

n re

ques

t fo

r pr

e-au

dit

All

docu

men

tatio

n fo

r pr

e-au

dit

colle

cted

Hol

d pr

e-au

dit o

rient

atio

n m

eetin

g

*Cor

Act

= C

orre

ctiv

e ac

tion

Key

Dat

es

Cor

Act* m

eetin

gs

Ful

l aud

it

Rel

ease

full

audi

t sch

edul

eC

omm

ence

full

audi

tF

inis

h fu

ll au

dit

Hol

d fu

ll au

dit o

rient

atio

n m

eetin

g

2/25

6/13

7/29

8/12

AND 3-11N 10/7/03, 9:17 AM9


Activity NetworkPhase I

ISO 9000 Certification Audit Schedule

Ear

liest

S

tart

(E

S)

Ear

liest

F

inis

h (E

F)

Late

stS

tart

(LS

)

Late

stF

inis

h(L

F)

Info

rmat

ion

prov

ided

cou

rtes

y of

BG

P

Pro

ject

S

tart

1. IS

O

orie

ntat

ion

T =

4

wee

ks0

4

0

4

2. P

lan

train

ing

T =

1

wee

k4

5

4

5

4. P

lan

pre-

audi

t

T =

2w

eeks

5

75

7

5. P

re-a

udit

prep

arat

ion

T =

3

wee

ks7

10

7 1

0

6. P

re-a

udit

T =

2

wee

ks10

12

10 1

2

3. A

udito

r tra

inin

g

T =

4

wee

ks4

8

6 1

0

1

See

ne

xt

page

AND 3-11N 10/7/03, 9:17 AM10


Activity NetworkPhase I

ISO 9000 Certification Audit Schedule (continued)

Info

rmat

ion

prov

ided

cou

rtes

y of

BG

P

Pro

ject

F

inis

h7.

Pre

-aud

it re

port

T =

1

wee

k12

13

12 1

3

9. G

ener

ate

audi

t sc

hedu

leT

= 1

w

eek

13 1

413

14

10. P

lan

audi

t

T =

16

wee

ks14

30

14 3

0

11. A

udit

prep

arat

ion

T =

2

wee

ks30

32

30 3

2

13. A

udit

repo

rt

T =

2

wee

ks39

41

39 4

1

8. C

orre

ctiv

e ac

tion

T =

24

wee

ks1

3

37

15

3

9

1

14. C

orre

ctiv

e ac

tion

mee

tings

T =

4

wee

ks41

45

41 4

5

from

pr

evio

us

page

12. F

ull

audi

t

T =

7

wee

ks32

39

32 3

9

Not

e: T

he A

ND

sho

ws

that

the

cert

ific

atio

n p

roce

ss w

ill t

ake

45 w

eeks

. The

bol

d a

rrow

,in

dic

atin

g th

e cr

itic

al p

ath,

cle

arly

sho

ws

thos

e ta

sks

that

mu

st b

e co

mp

lete

d a

ssc

hed

ule

d. T

he ta

sks

off t

he c

riti

cal p

ath

wil

l als

o re

quir

e ca

refu

l mon

itor

ing

sinc

e th

ere

is o

nly

two

wee

ks o

f sl

ack

tim

e in

the

sche

du

le.

AND 3-11N 10/7/03, 9:17 AM11

©1994 GOAL/QPC12 Affinity

What are the issues involved in planningfun family vacations?

Affinity DiagramGathering &grouping ideas

Why use it?To allow a team to creatively generate a large number ofideas/issues and then organize and summarize naturalgroupings among them to understand the essence of aproblem and breakthrough solutions.

What does it do?• Encourages creativity by everyone on the team at

all phases of the process• Breaks down longstanding communication barriers• Encourages non-traditional connections among

ideas/issues• Allows breakthroughs to emerge naturally, even

on long-standing issues• Encourages “ownership” of results that emerge

because the team creates both the detailed inputand general results

• Overcomes “team paralysis,” which is brought onby an overwhelming array of options and lack ofconsensus

How do I do it?1. Phrase the issue under discussion in a full sentence

12 Affinity

Affinity 12-18N 10/7/03, 9:16 AM12

©1994 GOAL/QPC Affinity 13

Tip From the start, reach consensus on the choiceof words you will use. Neutral statements workwell, but positive, negative, and solution-ori-ented questions also work.

2. Brainstorm at least 20 ideas or issuesa) Follow guidelines for brainstorming.b) Record each idea on a Post-it™ note in bold, large

print to make it visible 4–6 feet away. Use atminimum, a noun and a verb. Avoid using singlewords. Four to seven words work well.

What are the issues involved in planning fun family vacations?

Ask kids for ideas

Find a good range of price alternatives

Consider everyone's

hobbies

Determine total budget

Use a creative

travel agent

Find locations with activities for all ages

Recall good vacations of

the past

Look at family pictures

Combine vacation with business trip

Illustration Note: There are 10 to 40 more ideas in a typical Affinity Diagram

Tip A “typical” Affinity has 40–60 items; it is notunusual to have 100–200 ideas.

3. Without talking: sort ideas simultaneously into5–10 related groupingsa) Move Post-it™ notes where they fit best for you;

don’t ask, simply move any notes that you thinkbelong in another grouping.

Affinity 12-18N 10/7/03, 9:16 AM13


b) Sorting will slow down or stop when each personfeels sufficiently comfortable with the groupings.

Tip Sort in silence to focus on the meaning behindand connections among all ideas, instead of emo-tions and “history” that often arise in discussions.

Tip As an idea is moved back and forth, try to seethe logical connection that the other person ismaking. If this movement continues beyond a rea-sonable point, agree to create a duplicate Post-it™.

Tip It is okay for some notes to stand alone. These“loners” can be as important as others that fit intogroupings naturally.

4. For each grouping, create summary or headercards using consensusa) Gain a quick team consensus on a word or phrase

that captures the central idea/theme of each


Illustration Note: There are 5 to 10 more groupings of ideas in a typical Affinity Diagram

Ask kids for ideas


Consider everyone's

hobbies


Use a creative

travel agent



the past



Affinity 12-18N 10/7/03, 9:16 AM14


grouping; record it on a Post-it™ note and place itat the top of each grouping. These are draftheader cards.

b) For each grouping, agree on a concise sentencethat combines the grouping’s central idea andwhat all of the specific Post-it™ notes add to thatidea; record it and replace the draft version. Thisis a final header card.

c) Divide large groupings into subgroups as neededand create appropriate subheaders.

d)Draw the final Affinity Diagram connecting allfinalized header cards with their groupings.

Illustration Note: There are 5 to 10 groupings of ideas in a typical Affinity. This is a partial Affinity.


Ask kids for ideas


Consider everyone's

hobbies


Use a creative

travel agent



the past



Define an ideal vacation

through family consensus

Determine the most your

budget will buy

Use multiple sources for

vacation research

Affinity 12-18N 10/7/03, 9:16 AM15


Tip Spend the extra time needed to do solid headercards. Strive to capture the essence of all of the ideasin each grouping. Shortcuts here can greatly reducethe effectiveness of the final Affinity Diagram.

It is possible that a note within a groupingcould become a header card. However, don’tchoose the “closest one” because it’s convenient.The hard work of creating new header cards oftenleads to breakthrough ideas.

VariationsAnother popular form of this tool, called the KJMethod, was developed by the Japanese anthropolo-gist Jiro Kawakita while he was doing fieldwork in the1950s. The KJ Method, identified with Kawakita’s ini-tials, helped the anthropologist and his students gatherand analyze data. The KJ Method differs from theAffinity Diagram described above in that the cards arefact-based and go through a highly structured refine-ment process before the final diagram is created.

Affinity 12-18N 10/7/03, 9:16 AM16


AffinityIssues Surrounding

Implementation of the Business Plan

The

gro

up c

ould

func

tion

mor

e ef

fect

ivel

y

Fun

ctio

nal g

roup

s no

t tr

ustin

g ea

ch o

ther

Gro

up m

embe

rs n

ot

mak

ing

indi

vidu

al

com

mitm

ent t

o su

cces

s of

the

plan

Rew

ards

do

not

com

pens

ate

team

pl

ayin

g

No

stro

ng c

omm

itmen

t to

the

grou

p

Insu

ffici

ent t

eam

ap

proa

ch to

new

pr

oduc

t dev

elop

men

t an

d in

trod

uctio

n

The

m/u

s pe

rcep

tion

Com

mun

icat

ion

betw

een

func

tiona

l gr

oups

diff

icul

t

Gro

up is

not

foca

l po

int f

or c

onfli

ct

reso

lutio

n

Com

mun

icat

ion

issu

es

with

in th

e gr

oup

Info

rmat

ion

prov

ided

cour

tesy

of G

oody

ear

Rec

onci

liatio

n w

ith

corp

orat

e re

sour

ce

allo

catio

n

Ow

ners

hip

of p

lan

does

n’t c

ross

fu

nctio

nal l

ines

Pla

n no

t int

egra

ted

Poo

r de

finiti

on o

f pr

iorit

izat

ion

for

mar

ket

intr

oduc

tion

Fig

htin

g da

ily

prob

lem

s (a

lliga

tor/

swam

p)

Pla

nnin

g ap

proa

ch

not s

tand

ardi

zed

Mea

ns n

ot c

lear

ly

defin

ed

Unr

ealis

tic g

oals

cr

eate

Blu

e S

ky

attit

ude

Pla

n is

not

link

ed to

un

it fin

anci

al g

oals

Our

bus

ines

s pl

anni

ng

appr

oach

mus

t be

impr

oved

Mor

e gr

oupi

ngs

on n

ext p

age

Affinity 12-18N 10/7/03, 9:16 AM17


AffinityIssues Surrounding

Implementation of the Business Plan (cont.)

Note: The Affinity helped the team bring focus to the manyopinions on business planning. The headers that surfacedbecame the key issues in the ID example (shown in the ID toolsection).

Lim

ited

reso

urce

s ar

e a

chal

leng

e

Com

plex

ity d

riven

by

cust

omer

dem

ands

re

quire

s ad

ded

inve

stm

ent

Cap

ital a

vaila

bilit

y lim

its o

ppor

tuni

ties

Lack

of t

ime

and

reso

urce

s

Lack

inte

grat

ion

of

supp

ort g

roup

pla

ns

Fas

ter

pace

of

prod

uct

intr

oduc

tions

st

retc

hes

reso

urce

s

Fas

t new

pro

duct

in

trod

uctio

ns s

tret

ch

reso

urce

s

Cap

acity

may

not

m

eet n

eeds

Sal

es fo

reca

st is

no

t acc

urat

e

Pro

duct

ion

capa

city

Pro

duct

ion

capa

bilit

y to

su

ppor

t cha

ngin

g re

quire

men

ts

Ext

erna

l fa

ctor

s im

pact

im

plem

enta

tion

New

gov

ernm

ent

regu

latio

ns

Pos

sibi

lity

of

econ

omic

do

wnt

urn

Info

rmat

ion

prov

ided

cour

tesy

of G

oody

ear

Affinity 12-18N 10/7/03, 9:16 AM18

©1994 GOAL/QPC Brainstorming 19Brainstorming 19

BrainstormingCreating bigger& better ideas

Why use it?To establish a common method for a team to creativelyand efficiently generate a high volume of ideas on anytopic by creating a process that is free of criticism andjudgment.

What does it do?• Encourages open thinking when a team is stuck in

“same old way” thinking• Gets all team members involved and enthusiastic

so that a few people don’t dominate the wholegroup

• Allows team members to build on each other’screativity while staying focused on their jointmission

How do I do it?There are two major methods for brainstorming.

• Structured. A process in which each team mem-ber gives ideas in turn.

• Unstructured. A process in which team membersgive ideas as they come to mind.

Either method can be done silently or aloud.

Brainstorming 19-22N 10/7/03, 9:18 AM19

©1994 GOAL/QPC20 Brainstorming

Structured1. The central brainstorming question is stated,

agreed on, and written down for everyone to seeBe sure that everyone understands the question,issue, or problem. Check this by asking one or twomembers to paraphrase it before recording it on aflipchart or board.

2. Each team member, in turn, gives an idea. Noidea is criticized. Ever!With each rotation around the team, any membercan pass at any time. While this rotation processencourages full participation, it may alsoheighten anxiety for inexperienced or shy teammembers.

3. As ideas are generated, write each one in large,visible letters on a flipchart or other writingsurfaceMake sure every idea is recorded with the samewords of the speaker, don’t interpret or abbrevi-ate. To ensure this, the person writing shouldalways ask the speaker if the idea has beenworded accurately.

4. Ideas are generated in turn until each personpasses, indicating that the ideas (or members)are exhaustedKeep the process moving and relatively short—5 to 20 minutes works well, depending on howcomplex the topic is.

5. Review the written list of ideas for clarity and todiscard any duplicatesDiscard only ideas that are virtually identical. It isoften important to preserve subtle differencesthat are revealed in slightly different wordings.


©1994 GOAL/QPC Brainstorming 21

UnstructuredThe process is the same as in the structured methodexcept that ideas are given by everyone at any time.There is no need to “pass” since ideas are not solicitedin rotation.

VariationsThere are many ways to stimulate creative team think-ing. The common theme among all of them is thestimulation of creativity by taking advantage of thecombined brain power of a team. Here are three ex-amples:

• Visual brainstorming. Individuals (or the team)produce a picture of how they see a situation orproblem.

• Analogies/free-word association. Unusual con-nections are made by comparing the problem toseemingly unrelated objects, creatures, or words.For example: “If the problem was an animal, whatkind would it be?”

• 6-3-5 method. This powerful, silent method isproposed by Helmut Schlicksupp in his bookCreativity Workshop. It is done as follows:

a) Based on a single brainstorming issue, eachperson on the team (usually 6 people) has 5minutes to write down 3 ideas on a sheet of paper.

b) Each person then passes his or her sheet ofpaper to the next person, who has 5 moreminutes to add 3 more ideas that build on thefirst 3 ideas.


©1994 GOAL/QPC22 Brainstorming

c) This rotation is repeated as many times asthere are team members, e.g., 6 team mem-bers = 6 rotations, 6 sheets of paper, 18 ideasper sheet.

This interesting process forces team members toconsciously build on each other’s perspectivesand input.


©1994 GOAL/QPC C & E/Fishbone 23

Cause & Effect/Fishbone DiagramFind & cure causes,NOT symptoms

Why use it?To allow a team to identify, explore, and graphicallydisplay, in increasing detail, all of the possible causesrelated to a problem or condition to discover its rootcause(s).

What does it do?• Enables a team to focus on the content of the prob-

lem, not on the history of the problem or differingpersonal interests of team members

• Creates a snapshot of the collective knowledgeand consensus of a team around a problem. Thisbuilds support for the resulting solutions

• Focuses the team on causes, not symptoms

How do I do it?1. Select the most appropriate cause & effect format.

There are two major formats:• Dispersion Analysis Type is constructed by plac-

ing individual causes within each “major” causecategory and then asking of each individual cause“Why does this cause (dispersion) happen?” Thisquestion is repeated for the next level of detailuntil the team runs out of causes. The graphicexamples shown in Step 3 of this tool section arebased on this format.

C & E/Fishbone 23

C_E 23-30N 10/7/03, 9:21 AM23

©1994 GOAL/QPC24 C & E/Fishbone

• Process Classification Type uses the major stepsof the process in place of the major cause cat-egories. The root cause questioning process is thesame as the Dispersion Analysis Type.

2. Generate the causes needed to build a Cause &Effect Diagram. Choose one method:• Brainstorming without previous preparation• Check Sheets based on data collected by team

members before the meeting

3. Construct the Cause & Effect/Fishbone Diagrama) Place the problem statement in a box on the

righthand side of the writing surface.• Allow plenty of space. Use a flipchart sheet,

butcher paper, or a large white board. A papersurface is preferred since the final Cause & EffectDiagram can be moved.

Late pizza deliveries on

Fridays & Saturdays

Causes“Bones”

(Major cause categories)Effect

Tip Make sure everyone agrees on the problemstatement. Include as much information as possibleon the “what,” “where,” “when,” and “how much”of the problem. Use data to specify the problem.

C_E 23-30N 10/7/03, 9:21 AM24


b) Draw major cause categories or steps in the pro-duction or service process. Connect them to the“backbone” of the fishbone chart.

• Be flexible in the major cause “bones” that areused. In a Production Process the traditionalcategories are: Machines (equipment), Methods(how work is done), Materials (components orraw materials), and People (the human element).In a Service Process the traditional methods are:Policies (higher-level decision rules), Procedures(steps in a task), Plant (equipment and space),and People. In both types of processes, Environ-ment (buildings, logistics, and space), and Mea-surement (calibration and data collection) arealso frequently used. There is no perfect set ornumber of categories. Make them fit the problem.

Late pizza deliveries on

Fridays & Saturdays

Machinery/Equipment People

MaterialsMethods

Illustration Note: In a Process Classification Typeformat, replace the major “bone” categories with:“Order Taking,” “Preparation,” “Cooking,” and “Delivery.”

C_E 23-30N 10/7/03, 9:21 AM25


Late pizza deliveries

on Fridays & Saturdays

Machinery/Equipment People

MaterialsMethods

People don’tshow up

Unreliable cars

Ovenstoo small

Driversget lost

Run out of ingredients

Poor dispatching

Poor handling of

large orders

c) Place the brainstormed or data-based causes inthe appropriate category.

• In brainstorming, possible causes can be placed ina major cause category as each is generated, oronly after the entire list has been created. Eitherworks well but brainstorming the whole list firstmaintains the creative flow of ideas without beingconstrained by the major cause categories orwhere the ideas fit in each “bone.”

• Some causes seem to fit in more than one category.Ideally each cause should be in only one category, butsome of the “people” causes may legitimately belong intwo places. Place them in both categories and see howthey work out in the end.

Tip If ideas are slow in coming, use the major causecategories as catalysts, e.g., “What in ‘materials’ iscausing . . . ?”

d) Ask repeatedly of each cause listed on the“bones,” either:

• “Why does it happen?” For example, under “Runout of ingredients” this question would lead to

C_E 23-30N 10/7/03, 9:21 AM26


more basic causes such as “Inaccurate ordering,”“Poor use of space,” and so on.

Late

piz

za

deliv

erie

s on

F

riday

s &

S

atur

days

Peo

ple

don'

t sho

w u

p

Hig

h tu

rnov

erLo

w p

ay

No

team

wor

k No

trai

ning

Peo

ple

Don

't kn

ow to

wn

Driv

ers

get l

ost

Hig

h tu

rnov

er

Get

wro

ng

info

rmat

ion

Poo

r tr

aini

ng

Rus

hed

Mac

hine

ry/E

quip

men

t

Unr

elia

ble

cars

Low

pay

Kid

s ow

n ju

nks

No

mon

ey fo

r re

pairs

Ove

ns to

o sm

all

Poo

r us

e of

spa

cePoo

r tr

aini

ngH

igh

turn

over

No

capa

city

for

peak

per

iods

Met

hods

Poo

r ha

ndlin

g of

la

rge

orde

rs

Lack

of e

xper

ienc

eP

oor

disp

atch

ing

Man

y ne

w s

tree

ts

Don

't kn

ow

tow

n Hig

h tu

rnov

er

Run

out

of i

ngre

dien

ts

Inac

cura

te

orde

ring

Poo

r us

e of

spa

ce

Lack

of

trai

ning M

ater

ials

Hig

h tu

rnov

er

Hig

h tu

rnov

er

C_E 23-30N 10/7/03, 9:21 AM27


• “What could happen?” For example, under “Runout of ingredients” this question would lead to adeeper understanding of the problem such as“Boxes,” “Prepared dough,” “Toppings,” and so on.

Tip For each deeper cause, continue to push fordeeper understanding, but know when to stop. Arule of thumb is to stop questioning when a causeis controlled by more than one level of manage-ment removed from the group. Otherwise, theprocess could become an exercise in frustration.Use common sense.

e) Interpret or test for root cause(s) by one or more ofthe following:

• Look for causes that appear repeatedly within oracross major cause categories.

• Select through either an unstructured consensusprocess or one that is structured, such as NominalGroup Technique or Multivoting.

• Gather data through Check Sheets or other for-mats to determine the relative frequencies of thedifferent causes.

VariationsTraditionally, Cause & Effect Diagrams have been cre-ated in a meeting setting. The completed “fishbone” isoften reviewed by others and/or confirmed with datacollection. A very effective alternative is CEDAC®, inwhich a large, highly visible, blank fishbone chart isdisplayed prominently in a work area. Everyone postsboth potential causes and solutions on Post-it™ notes ineach of the categories. Causes and solutions are re-viewed, tested, and posted. This technique opens upthe process to the knowledge and creativity of everyperson in the operation.

C_E 23-30N 10/7/03, 9:21 AM28


Cause & Effect/FishboneBed Assignment Delay

Sys

tem

inco

rrec

t

Mac

hine

(P

CIS

)Ti

min

g

Hos

pita

l pro

cedu

res

Com

mun

icat

ion

Pat

ient

wai

ts

for

bed

Not

ent

ered

Not

use

d

No

trus

tN

eed

mor

e tr

aini

ng

Fun

ctio

ns n

ot u

sefu

l

Not

use

dpe

ndin

g di

scha

rge

Dis

char

ged

patie

nt

did

not l

eave

Wai

t for

res

ults

Wai

t for

lunc

hW

ait f

or r

ide

Cal

l hou

seke

epin

g to

o la

te

Wai

t for

MD Cal

l hou

seke

epin

g to

o ea

rly

Thi

nk it

will

take

m

ore

time

Pat

ient

arr

ives

to

o ea

rly

Tran

sfer

too

early

fr

om a

noth

er h

ospi

tal

Cal

l hou

seke

epin

g w

hen

clea

n

Nur

sing

sho

rtag

e

Uni

t cle

rk s

taffi

ng

Uni

t cle

rk tr

aini

ng

Res

ourc

es

Uni

t cle

rk u

naw

are

of d

isch

arge

or

tran

sfer

On

brea

kN

ot to

ldShi

ft ch

ange

Res

erva

tion

unaw

are

Not

ent

eredU

nit s

witc

h be

dA

dmitt

ing

unaw

are

bed

is c

lean

Del

ayed

en

try

San

dbag

Too

busy

Inap

prop

riate

E

R a

dmitt

ance

Man

y tr

ansf

ers

Spe

cial

ty b

eds

Car

diac

mon

itors

Dou

ble

room

sP

hysi

cian

did

no

t writ

e or

der

Med

icin

e ad

mit

quot

a

Phy

sici

an m

isus

e –

inpa

tient

MD

pro

cedu

res

Information provided courtesy ofRush-Presbyterian-St. Luke’s Medical Center

C_E 23-30N 10/7/03, 9:21 AM29


Cause & Effect/FishboneCauses for Bent Pins (Plug-In Side)

Information provided courtesy of AT&T

Ben

t pin

s pl

ug-in

sid

e

Han

dlin

gP

eopl

e

Des

ign

Tool

s

Met

hods

Gen

eral

han

dlin

g

Rep

air

Test

Sto

rage

In/O

ut o

f Box

es

Bac

kpla

ne m

ount

ing

Mou

nt to

fram

eH

eavy

/Aw

kwar

d

Gau

ging

Impr

oper

in

sert

ion

Dam

aged

con

nect

orS

tora

ge

Ret

aine

r in

sert

ion

Pla

cem

ent -

pi

ns o

n be

nch

No/

Don

't Li

ke

Fixt

ures

Myl

ar ta

pe

Scr

ibe

slip

sP

lace

men

t on

edg

esA

ssem

bly

diffi

culty

Larg

e pa

rt #

Com

plex

de

sign

004

prep

VA

Sho

ws

pins

on

benc

h

No

Fixt

ure

Car

eles

snes

s

Lack

of a

ttent

ion

Atti

tude O

n-th

e-jo

b tr

aini

ng

Usi

ng w

rong

tool

s

New

tool

s -

long

lead

tim

eIm

prop

er s

izesBits

Not

eno

ugh

Gau

ges

Scr

ews

com

e lo

ose

Dam

aged

con

nect

ors

Test

fixt

ures

Dam

aged

con

nect

ors

Sto

rage

Lack

of

fixtu

res

Not

Eno

ugh

Uni

vers

al

Cas

ting

ties/

plie

rs

No

Stif

fene

r P

late

sB

ad p

anel

al

ignm

ent –

bar

e

Des

igne

r no

t on

site

Des

igne

r ca

n't r

eact

to

pro

blem

s

Not

des

igne

d fo

r m

anuf

actu

ring

2 ba

ckpl

anes

Larg

e pa

rt #

Ass

embl

y di

fficu

lty

C_E 23-30N 10/7/03, 9:22 AM30

©1994 GOAL/QPC Check Sheet 31

1 2 TotalA |||| ||| 7

B ||| |||| 7

C || | 3

Check SheetCounting &accumulating data

Why use it?To allow a team to systematically record and compiledata from historical sources, or observations as theyhappen, so that patterns and trends can be clearlydetected and shown.

What does it do?• Creates easy-to-understand data that come from

a simple, efficient process that can be applied toany key performance areas

• Builds, with each observation, a clearer picture of “thefacts” as opposed to the opinions of each teammember

• Forces agreement on the definition of each condi-tion or event (every person has to be looking forand recording the same thing)

• Makes patterns in the data become obvious quickly

How do I do it?1. Agree on the definition of the events or conditions

being observed• If you are building a list of events or conditions as

the observations are made, agree on the overalldefinition of the project.Example: If you are looking for reasons for latepayments, agree on the definition of “late.”

Check Sheet 31

Check Sheet 31-35N 10/7/03, 9:23 AM31

©1994 GOAL/QPC32 Check Sheet

• If you are working from a standard list of eventsor conditions, make sure that there is agreementon the meaning and application of each one.Example: If you are tracking sales calls fromvarious regions, make sure everyone knowswhich states are in each region.

2. Decide who will collect the data; over what period;and from what sources• Who collects the data obviously depends on the

project and resources. The data collection periodcan range from hours to months. The data cancome from either a sample or an entire popula-tion.

• Make sure the data collector(s) have both the timeand knowledge they need to collect accurate in-formation.

• Collect the data over a sufficient period to be surethe data represents “typical” results during a“typical” cycle for your business.

• Sometimes there may be important differenceswithin a population that should be reflected bysampling each different subgroup individually.This is called stratification.Example: Collect complaint data from businesstravelers separately from other types of travelers.Collect scrap data from each machine separately.

Tip It must be safe to record and report “badnews,” otherwise the data will be filtered.

Check Sheet 31-35N 10/7/03, 9:23 AM32


j

g

DateTotal

3/10 3/11 3/12 3/13 3/14 3/15 3/16

Lab delays 9 4 6 6 3 12 12 52

No beds available 2 7 2 4 5 8 3 31

Incompletepatient

information7 3 1 2 2 4 5 24

Total 33 28 36 30 25 47 38 237

Project: Admission Delays

Location: Emergency Room

Name: (if applicable)

Dates: 3/10 to 3/16

Shift: All

Reason:

cd

h j

e

g

abf i

3. Design a Check Sheet form that is clear, complete,and easy to use• A complete Check Sheet, illustrated below, in-

cludes the following:Source Information

a Name of project b Location of data collectionc Name of person recording data, if it appliesd Date(s)e Other important identifiers

Content Information f Column with defect/event name

Columns with collection days/datesh Totals for each column i Totals for each row

Grand total for both the columns and rows

Check Sheet 31-35N 10/7/03, 9:23 AM33

©1994 GOAL/QPC34 Check Sheet

4. Collect the data consistently and accurately• Make sure all entries are written clearly.Tip Managers and/or team members can do their

part to help the data collector(s) do their job wellby simply showing an interest in the project. Askthe collector(s) how the project is working out.Show your support—tell the data collector(s) youthink it is important to collect the information.Above all—act on the data as quickly as possible!

VariationsDefect LocationShows the concentration of defects by marking a draw-ing of the product each time a defect is found.

Generic computer monitor

xxx XXxxx

Front

xxxxxx

Back

xx

Project: Monitor casing damageLocation: Assembly

Name: (if applicable) Shift: FirstDate: 6/11

Task ChecklistTasks in producing a product or delivering a service arechecked off as they are done. In complex processes thisis a form of “mistake-proofing.”

Check Sheet 31-35N 10/7/03, 9:23 AM34


Check SheetKeyboard Errors

in Class Assignment

Information provided courtesy of Millcreek Township School District,Millcreek Township, Pennsylvania

Cen

terin

gS

pelli

ngP

unct

uatio

nM

isse

d pa

ragr

aph

Wro

ng n

umbe

rsW

rong

pag

e nu

mbe

rsTa

bles

|| ||||

|||||

| ||

|| |

||||| ||| | |||

|

Mis

take

sM

arch

12

3

||| ||||

||||

||||

| ||

||| |||

|| |||

|

||| ||||

||||

||||

||||

| ||| || ||||

Tota

l

8 23 40 4 10 4 13

3435

33To

tal

102

Check Sheet 31-35N 10/7/03, 9:23 AM35

©1994 GOAL/QPC36 Control Charts

Control ChartsRecognizingsources of variation

Why use it?To monitor, control, and improve process performanceover time by studying variation and its source.

What does it do?• Focuses attention on detecting and monitoring

process variation over time• Distinguishes special from common causes of

variation, as a guide to local or managementaction

• Serves as a tool for ongoing control of a process• Helps improve a process to perform consistently

and predictably for higher quality, lower cost,and higher effective capacity

• Provides a common language for discussing pro-cess performance

How do I do it?There are many types of Control Charts. The ControlChart(s) that your team decides to use will be deter-mined by the type of data you have. Use the TreeDiagram on the next page to determine which ControlChart(s) will best fit your situation. Other types ofControl Charts, which are beyond the scope of thisbook, include the Pre-Control Chart, the MovingAverage & Range Chart, the Cumulative Sum Chart,and Box Plots.

UCL

X

LCL

36 Control Charts

Contrl Chrts 36-51N 10/7/03, 9:24 AM36

©1994 GOAL/QPC Control Charts 37

Based on the type of data and sample size you have,choose the appropriate Control Chart.

Choose Appropriate Control Chart

Variable data:measured & plotted on

a continuous scale, e.g., time, temperature,

cost, figures

Attribute data:counted & plotted as discrete events, e.g., shipping errors, %

waste, absenteeism

Defective** data

Defect* data

Variable sample size, usually ≥ 50

Constant sample size, usually ≥ 50

p Chartnp Chartc Chartu Chart

Sample size is small,

usually 3 to 5

Sample size is large,

usually ≥ 10

Sample size = 1

Sample size is small,

median value

X and Rm

* Defect = Failure to meet one of the acceptance criteria. A defective unit might have multiple defects.

** Defective = An entire unit fails to meet acceptance criteria, regardless of the number of defects on the unit.

X and R–

X and s–

X and R~

Constant sample size, usually c > 5

Variable sample size

Contrl Chrts 36-51N 10/7/03, 9:24 AM37


Constructing Control Charts

1. Select the process to be charted

2. Determine sampling method and plan• How large a sample can be drawn? Balance the

time and cost to collect a sample with the amountof information you will gather. See the Tree Dia-gram on the previous page for suggested samplesizes.

• As much as possible, obtain the samples underthe same technical conditions: the same machine,operator, lot, and so on.

• Frequency of sampling will depend on whetheryou are able to discern patterns in the data. Con-sider hourly, daily, shifts, monthly, annually,lots, and so on. Once the process is “in control,”you might consider reducing the frequency withwhich you sample.

• Generally, collect 20–25 groups of samples beforecalculating the statistics and control limits.

• Consider using historical data to set a baseline.Tip Make sure samples are random. To establish

the inherent variation of a process, allow theprocess to run untouched, i.e., according to stan-dard procedures.

3. Initiate data collection• Run the process untouched, and gather sampled

data.• Record data on an appropriate Control Chart

sheet or other graph paper. Include any unusualevents that occur.

Contrl Chrts 36-51N 10/7/03, 9:24 AM38


4. Calculate the appropriate statisticsa) If you have attribute data, use the Attribute Data

Table, Central Line column.

Attribute Data Table

Type ControlChart

Samplesize Central Line Control Limits

Fractiondefective Variable,

usually≥50p Chart

Numberdefective Constant,

usually≥50np Chart

Numberof defects Constant

c Chart

Numberof defectsper unit Variable

u Chart

*LCLp = p – 3 p(1 – p) n

*UCLp = p + 3 p(1 - p) n

LCLnp = np – 3 np(1 – p)

UCLnp = np + 3 np(1 – p)

LCLc = c – 3 c

UCLc = c + 3 c

*UCLu = u + 3 u n

*LCLu = u – 3 u n

For each subgroup:p = np/n

For all subgroups:p = Σnp/Σn

For each subgroup:np = # defective unitsFor all subgroups:

np = Σnp/k

For each subgroup:c = # defects

For all subgroups:c = Σc/k

For each subgroup:u = c/n

For all subgroups:u = Σc/Σn

np =c =n =

k =

# defective units# of defectssample size withineach subgroup# of subgroups

* This formula creates changing control limits. To avoid this, use average sample sizes n for those samples that are within ±20% of the average sample size. Calculate individual limits for the samples exceeding ±20%.

Contrl Chrts 36-51N 10/7/03, 9:24 AM39


b) If you have variable data, use the Variable DataTable, Central Line column.

Variable Data Table

LCLX = X – A3s

Type Control Chart

Samplesize n Control Limits

Average & Range <10, but

usually 3 to 5

Central Line*

–R = (R1+R2+…Rk)

k

Average & Standard Deviation Usually

≥10 –s = (s1+s2+…sk)k

X = (X1+X2+…Xk)=

k

– – –

Median & Range <10, but

usually 3 or 5

–R = (R1+R2+…Rk)

k

X = (X1+X2 +…Xk)k

~ ~ ~~

Individuals& Moving

Range

X and Rm

1

Rm = (R1 + R2 +…Rk–1)k – 1

= (X1+X2 +…Xk)X–

k

–

X = (X1+X2+…Xk)=

k

– – –

Rm = |(Xi + 1 – Xi)|

–

–LCLx = X – A2R–

–LCLR = D3R

–UCLx = X + A2R–

–UCLR = D4R

–UCLs = B4s–LCLs = B3s

–UCLX = X + A2R

~–~

–UCLX = X + E2Rm

–LCLR = D3R

–UCLR = D4R

–LCLX = X – A2R

~~

–LCLX = X – E2Rm

––

–UCLRm = D4Rm

–LCLRm = D3Rm

UCLX = X + A3s––

––

–

––––

X and R–

X and R~

X and s–

~k = # of subgroups, X = median value within each subgroup

*X = ∑Xin

=

=

~

~

Contrl Chrts 36-51N 10/7/03, 9:24 AM40


5. Calculate the control limitsa) If you have attribute data, use the Attribute Data

Table, Control Limits column.b) If you have variable data, use the Variable Data

Table, Control Limits column for the correct for-mula to use.

• Use the Table of Constants to match the numericvalues to the constants in the formulas shown inthe Control Limits column of the Variable DataTable. The values you will need to look up willdepend on the type of Variable Control Chart youchoose and on the size of the sample you havedrawn.

Tip If the Lower Control Limit (LCL) of an At-tribute Data Control Chart is a negative number,set the LCL to zero.

Tip The p and u formulas create changing controllimits if the sample sizes vary subgroup to sub-group. To avoid this, use the average sample size,n, for those samples that are within ±20% of theaverage sample size. Calculate individual limitsfor the samples exceeding ±20%.

6. Construct the Control Chart(s)• For Attribute Data Control Charts, construct one

chart, plotting each subgroup’s proportion ornumber defective, number of defects, or defectsper unit.

• For Variable Data Control Charts, construct twocharts: on the top chart plot each subgroup’smean, median, or individuals, and on the bottomchart plot each subgroup’s range or standarddeviation.

Contrl Chrts 36-51N 10/7/03, 9:24 AM41


Table of Constants

Sample size

n

X and R Chart X and s Chart

A2 D3 D4 A3 B3 B4 c4*2 1.880 0 3.267 2.659 0 3.267 .7979

3 1.023 0 2.574 1.954 0 2.568 .8862

4 0.729 0 2.282 1.628 0 2.266 .9213

5 0.577 0 2.114 1.427 0 2.089 .9400

6 0.483 0 2.004 1.287 0.030 1.970 .9515

7 0.419 0.076 1.924 1.182 0.118 1.882 .9594

8 0.373 0.136 1.864 1.099 0.185 1.815 .9650

9 0.337 0.184 1.816 1.032 0.239 1.761 .9693

10 0.308 0.223 1.777 0.975 0.284 1.716 .9727

Sample Size

n

X and R Chart X and Rm Chart

A2 D3 D4 E2 D3 D4 d2*2 - - - - 0 3.267 2.659 0 3.267 1.128

3 1.187 0 2.574 1.772 0 2.574 1.693

4 - - - - 0 2.282 1.457 0 2.282 2.059

5 0.691 0 2.114 1.290 0 2.114 2.326

6 - - - - 0 2.004 1.184 0 2.004 2.534

7 0.509 0.076 1.924 1.109 0.076 1.924 2.704

8 - - - - 0.136 1.864 1.054 0.136 1.864 2.847

9 0.412 0.184 1.816 1.010 0.184 1.816 2.970

10 - - - - 0.223 1.777 0.975 0.223 1.777 3.078

~

~

* Useful in estimating the process standard deviation σ.Note: The minimum sample size shown in this chart is 2because variation in the form of a range can only be calculatedin samples greater than 1. The X and Rm Chart creates theseminimum samples by combining and then calculating thedifference between sequential, individual measurements.

Contrl Chrts 36-51N 10/7/03, 9:24 AM42


• Draw a solid horizontal line on each chart. Thisline corresponds to the process average.

• Draw dashed lines for the upper and lower con-trol limits.

Interpreting Control Charts• Attribute Data Control Charts are based on one

chart. The charts for fraction or number defective,number of defects, or number of defects per unit,measure variation between samples. Variable DataControl Charts are based on two charts: the oneon top, for averages, medians, and individuals,measures variation between subgroups over time;the chart below, for ranges and standard devia-tions, measures variation within subgroups overtime.

• Determine if the process mean (center line) iswhere it should be relative to your customerspecifications or your internal business needs orobjectives. If not, then it is an indication thatsomething has changed in the process, or thecustomer requirements or objectives havechanged.

• Analyze the data relative to the control limits;distinguishing between common causes and spe-cial causes. The fluctuation of the points withinthe limits results from variation inherent in theprocess. This variation results from commoncauses within the system, e.g., design, choice ofmachine, preventive maintenance, and can onlybe affected by changing that system. However,points outside of the limits or patterns within thelimits, come from a special cause, e.g., humanerrors, unplanned events, freak occurrences, thatis not part of the way the process normallyoperates, or is present because of an unlikelycombination of process steps. Special causes must

Contrl Chrts 36-51N 10/7/03, 9:24 AM43


be eliminated before the Control Chart can be usedas a monitoring tool. Once this is done, the pro-cess will be “in control” and samples can be takenat regular intervals to make sure that the processdoesn’t fundamentally change. See“Determining if Your Process is Out of Control.”

• Your process is in “statistical control” if the pro-cess is not being affected by special causes, theinfluence of an individual or machine. All thepoints must fall within the control limits and theymust be randomly dispersed about the averageline for an in-control system.

Tip “Control” doesn’t necessarily mean that theproduct or service will meet your needs. It onlymeans that the process is consistent. Don’t confusecontrol limits with specification limits—specifica-tion limits are related to customer requirements,not process variation.

Tip Any points outside the control limits, onceidentified with a cause (or causes), should beremoved and the calculations and charts redone.Points within the control limits, but showing in-dications of trends, shifts, or instability, are alsospecial causes.

Tip When a Control Chart has been initiated andall special causes removed, continue to plot newdata on a new chart, but DO NOT recalculate thecontrol limits. As long as the process does notchange, the limits should not be changed. Controllimits should be recalculated only when a perma-nent, desired change has occurred in the process,and only using data after the change occurred.

Contrl Chrts 36-51N 10/7/03, 9:24 AM44


Tip Nothing will change just because you chartedit! You need to do something. Form a team toinvestigate. See “Common Questions for Investi-gating an Out-of-Control Process.”

Determining if Your Process is“Out of Control”

A process is said to be “out of control” if either one ofthese is true:

1. One or more points fall outside of the control limits

2. When the Control Chart is divided into zones, asshown below, any of the following points are true:

Zone AZone BZone CZone CZone BZone A

Upper Control Limit(UCL)

Average

Lower Control Limit(LCL)

a) Two points, out of three consecutive points, are onthe same side of the average in Zone A or beyond.

b) Four points, out of five consecutive points, are onthe same side of the average in Zone B or beyond.

c) Nine consecutive points are on one side of theaverage.

d) There are six consecutive points, increasing ordecreasing.

e) There are fourteen consecutive points that alter-nate up and down.

f) There are fifteen consecutive points within ZoneC (above and below the average).

Contrl Chrts 36-51N 10/7/03, 9:24 AM45


A B C C B A

UC

L

LC

LX=

12a

2b2c

2d2e

2f

12a

2b2c

2d2e

2f

Test

s fo

r C

on

tro

l

Sou

rce:

Llo

yd S

. Nel

son,

Dir

ecto

r of

Sta

tist

ical

Met

hod

s, N

ashu

a C

orp

orat

ion,

New

Ham

psh

ire

Contrl Chrts 36-51N 10/7/03, 9:24 AM46


Common Questions for Investigatingan Out-of-Control Process

❐ Yes ❐ No Are there differences in the meas-urement accuracy of instruments/methods used?

❐ Yes ❐ No Are there differences in the methodsused by different personnel?

❐ Yes ❐ No Is the process affected by the environ-ment, e.g., temperature, humidity?

❐ Yes ❐ No Has there been a significant changein the environment?

❐ Yes ❐ No Is the process affected by predictableconditions? Example: tool wear.

❐ Yes ❐ No Were any untrained personnel in-volved in the process at the time?

❐ Yes ❐ No Has there been a change in the sourcefor input to the process? Example:raw materials, information.

❐ Yes ❐ No Is the process affected by employeefatigue?

❐ Yes ❐ No Has there been a change in policiesor procedures? Example: mainten-ance procedures.

❐ Yes ❐ No Is the process adjusted frequently?❐ Yes ❐ No Did the samples come from different

parts of the process? Shifts? Indi-viduals?

❐ Yes ❐ No Are employees afraid to report “badnews”?

A team should address each “Yes” answer as a potentialsource of a special cause.

Contrl Chrts 36-51N 10/7/03, 9:24 AM47


Individuals & MovingRange Chart

IV Lines Connection Time

Process/Operation: IV Lines Connection Open Heart Admissions

Characteristic:Time in seconds

Sample Size:One

Sample Frequency:Each patient

Individuals:Ranges:

Department:Intensive CareBy:EW

Date:6/10

720

630

540

450

360

270

180

90

600

480

540

240

420

450

480

690

240

360

450

300

480

120

240

210

210

180

240

300

300

130

120

300

180

210

— 120

60 300

180

30 30 210

450

120

90 150

180

360

120

30 0 30 60 60 0 120

60 80 120

30

EW

EW

EW

EW

EW

EW

EW

EW

MA

EW

EW

EW

EW

EW

EW

EW

MA

EW

EW

EW

EW

EW

EW

EW

EW

EW

4/9

11 12 13 25 30 5/2 3 4 5 9 13 14 14 16 17 19 20 22 23 23 24 30 30 6/6 6

X

Rm

Who?

When?

420360300 240 180 120 60 0

UCL

X

LCL

Rm

k = 26n = 2

∑X = 8470∑R = 2990

UCL = 645UCL = 392

LCL = 7LCL = 0

LCL

UCL

X = 325.77R = 119.6

––

0

Sec

onds

Information provided courtesy ofParkview Episcopal Medical Center

Note: Something in the process changed, and now it takes lesstime to make IV connections for patients being admitted foropen heart surgery.

Contrl Chrts 36-51N 10/7/03, 9:24 AM48


60

50

40

30

20

10

% Failed

Month

Year

UCL

LCL

Historical Statistics:p = 39 UCL = 47 LCL = 31

p

Regular Hours Flex Time

1992 1993

40 36 36 42 42 40 20 26 25 19 20 18 16 10 12 12

Jul

Aug

Sep Oct

Nov

Dec

Jan

Feb

Mar

Apr

May

Jun

Jul

Aug

Sep Oct

Information provided courtesy of U.S. Navy,Naval Dental Center, San Diego

Note: Providing flex time for patients resulted in fewer appoint-ments missed.

p ChartGeneral Dentistry: Percent of

Patients Who Failed to Keep Appointments

Contrl Chrts 36-51N 10/7/03, 9:24 AM49


u ChartShop Process Check

Solder Defects


10000 –

9000 –

8000 –

7000 –

6000 –

5000 –

4000 –

3000 –

2000 –

1000 –

0 –

Jan

2 –

Jan

3 –

Jan

4 –

Jan

5 –

Jan

6 –

Jan

7 –

Jan

8 –

Jan

9 –

Jan

10

–

Jan

11

–

UCL

UDef

ect

Rat

e (P

PM

)

LCL

Checks (3 per day)

Historical Ave.: 2974 ppmHistorical UCL: 8758 ppmHistorical LCL: 0 ppm

Contrl Chrts 36-51N 10/7/03, 9:24 AM50


X & R Chart

Note: Hours 1, 16 and 22 should be reviewed to understandwhy these sample averages are outside the control limts

3.0

n = 10 parts randomly sampled each hour

2.01.81.61.41.21.0.8.6.4.2

LCL

UCL

4.0

4.5

5.0

3.5

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Ave

.

3.76

4.21

4.29

4.36

4.13

3.77

4.17

4.21

4.22

4.00

4.30

4.20

4.32

4.18

4.02

3.71

4.08

4.23

3.98

4.46

3.96

3.63

4.48

4.30

4.29

4.13

1.01

1.27

0.48

1.32

1.52

1.03

1.15

1.07

0.70

2.05

0.95

0.99

1.06

1.21

1.33

0.78

1.21

1.23

1.08

1.64

1.20

0.98

0.91

1.19

1.03

1.14

LCL

UCL

X=

Hou

rly

#X

R

Dia

met

er i

n M

illi

met

ers

R

Information provided courtesy of BlueFire Partners, Inc.and Hamilton Standard

Contrl Chrts 36-51N 10/7/03, 9:24 AM51

©1994 GOAL/QPC52 Data Points

Data PointsTurning datainto information

What Type of Data Do You Have?• Words?• Numbers?

– Attribute data? Attribute data can be countedand plotted as discrete events. It includes thecount of the numbers or percentages of good orbad, right or wrong, pass or fail, yes or no.Example: Number of correct answers on a test,number of mistakes per typed page, percentdefective product per shift.

– Variable data? Variable data can be measuredand plotted on a continuous scale.Example: Length, time, volume, weight.

Do You Need to Collect Data?• If you need to know the performance of an entire

population, the more economical and less timeconsuming method is to draw a sample from apopulation. With a sample, you can make infer-ences about, or predict, the performance of apopulation. Basic sampling methods:– Random. Each and every observation or data

measure has an equally likely chance of beingselected. Use a random number table or ran-dom number generator to select the samples.

– Sequential. Every nth sample is selected.– Stratified. A sample is taken from stratified

data groups.

36

78

51

307

43

93

6/8

?

52 Data Points

Data Points 52-55N 10/7/03, 9:26 AM52

©1994 GOAL/QPC Data Points 53

Can You Categorize Your Data Into Subgroups?• When you stratify data, you break it down into

meaningful subcategories or classifications, andfrom this point you can focus your problem solving.

Example: Data often comes from many sourcesbut is treated as if coming from one. Data on minorinjuries for a plant may be recorded as a singlefigure, but that number is actually the sum total ofinjuries by 1) type (cuts, burns, scrapes), 2) location(eyes, hands, feet), and 3) department (mainte-nance, shipping, production). Below is an exampleof how data has been stratified by plant department.

What Patterns are Important in Your Data?Predictable patterns or distributions can be describedwith statistics.

• Measures of location– Mean (or average). Represented by X (or X-bar),

the mean is the sum of the values of the sample(X1, X2, X3 . . . Xn) divided by the total number(n) of sampled data.Example: For the sample: (3, 5, 4, 7, 5)

(3 + 5 + 4 + 7 + 5)

5X = = 4.8

35-30-25-20-15-10-

5-0-

35-30-25-20-15-10-

5-0-

# In

jurie

s/M

onth

Dept. A and B Combined

∆ = Dept. Ao = Dept. B

J F M A M J

# In

jurie

s/M

onth

J F M A M J

∆∆

∆ ∆ ∆∆

o o oo o o

(Appears fairly stable over time.) (Dept. A is running higher and may be increasing over time.)

Data Points 52-55N 10/7/03, 9:26 AM53

©1994 GOAL/QPC54 Data Points

– Median. When sampled data are rank ordered,lowest to highest, the median is the middlenumber.Example: For the sample: (3, 5, 4, 7, 5)Median of (3, 4, 5, 5, 7) = 5When there are an even number of values, themedian is the average of the middle two values.Example: For the sample: (2, 5, 7, 4, 5, 3)Median of (2, 3, 4, 5, 5, 7) = 4.5

– Mode. The most frequently occurring value(s)in a sample.Example: For the sample: (3, 5, 4, 7, 5)Mode = 5

• Measures of variation– Range. Represented by R, the range is the dif-

ference between the highest data value (Xmax)and the lowest data value (Xmin).Example: For the sample: (3, 5, 4, 7, 5)R = 7 – 3 = 4

– Standard Deviation. Represented by s, the stan-dard deviation of a sample measures the varia–tion of the data around the mean. The lessvariation there is of the data values about themean, X, the closer s will be to zero (0).

Data Points 52-55N 10/7/03, 9:27 AM54

©1994 GOAL/QPC Data Points 55

Example: For the sample: (3, 5, 4, 7, 5) X = 4.8

s = [(3 – 4.8)2 + (5 – 4.8)2 + (4 – 4.8)2 + (7 – 4.8)2 + (5 – 4.8)2 ]

5 – 1

= [3.24 + .04 + .64 + 4.84 + .04]

4

= 8.8 4

= 2.2

= 1.48

The square of the standard deviation, s, is referred to asthe variance. Variance is not discussed in this book.

Data Points 52-55N 10/7/03, 9:27 AM55

©1994 GOAL/QPC56 Flowchart

FlowchartPicturingthe process

Why use it?To allow a team to identify the actual flow or sequenceof events in a process that any product or servicefollows. Flowcharts can be applied to anything from thetravels of an invoice or the flow of materials, to the stepsin making a sale or servicing a product.

What does it do?• Shows unexpected complexity, problem areas,

redundancy, unnecessary loops, and where sim-plification and standardization may be possible

• Compares and contrasts the actual versus theideal flow of a process to identify improvementopportunities

• Allows a team to come to agreement on the stepsof the process and to examine which activitiesmay impact the process performance

• Identifies locations where additional data can becollected and investigated

• Serves as a training aid to understand the com-plete process

How do I do it?1. Determine the frame or boundaries of the process

• Clearly define where the process under studystarts (input) and ends (final output).

• Team members should agree to the level of detailthey must show on the Flowchart to clearly un-derstand the process and identify problem areas.

56 Flowchart

Flow Chart 56-62N 10/7/03, 9:27 AM56

©1994 GOAL/QPC Flowchart 57

• The Flowchart can be a simple macro-flowchartshowing only sufficient information to under-stand the general process flow or it might bedetailed to show every finite action and decisionpoint. The team might start out with a macro-flowchart and then add in detail later or onlywhere it is needed.

2. Determine the steps in the process• Brainstorm a list of all major activities, inputs,

outputs, and decisions on a flipchart sheet fromthe beginning of the process to the end.

3. Sequence the steps• Arrange the steps in the order they are carried

out. Use Post-it™ notes so you can move themaround. Don’t draw in the arrows yet.

Tip Unless you are flowcharting a new process,sequence what is, not what should be or the ideal.This may be difficult at first but is necessary to seewhere the probable causes of the problems are inthe process.

4. Draw the Flowchart using the appropriate symbolsAn oval is used to show the materials, informa-tion or action (inputs) to start the process or toshow the results at the end (output) of theprocess.

A box or rectangle is used to show a task oractivity performed in the process. Although mul-tiple arrows may come into each box, usuallyonly one output or arrow leaves each activity box.

Flow Chart 56-62N 10/7/03, 9:27 AM57


A diamond shows those points in the processwhere a yes/no question is being asked or adecision is required.

A circle with either a letter or a number identifiesa break in the Flowchart and is continued else-where on the same page or another page.

Arrows show the direction or flow of the process.

• Keep the Flowchart simple using the basic sym-bols listed above. As your experience grows, useother, more graphic symbols to represent thesteps. Other symbols sometimes used include:– A half or torn sheet of paper for a report com-

pleted and/or filed– A can or computer tape wheel for data entry

into a computer database– A large “D” or half circle to identify places in the

process where there is a delay or wait for furtheraction

• Be consistent in the level of detail shown.– A macro-level flowchart will show key action

steps but no decision boxes.– An intermediate-level flowchart will show ac-

tion and decision points.– A micro-level flowchart will show minute detail.

• Label each process step using words that areunderstandable to everyone.

• Add arrows to show the direction of the flow ofsteps in the process. Although not a rule, if youshow all “yes” choices branching down and “no”choices branching to the left, it is easier to follow

➝

A

Flow Chart 56-62N 10/7/03, 9:27 AM58


the process. Preferences and space will later dic-tate direction.

• Don’t forget to identify your work. Include thetitle of your process, the date the diagram wasmade, and the names of the team members.

5. Test the Flowchart for completeness• Are the symbols used correctly?• Are the process steps (inputs, outputs, actions,

decisions, waits/delays) identified clearly?• Make sure every feedback loop is closed, i.e.,

every path takes you either back to or ahead toanother step.

• Check that every continuation point has a corre-sponding point elsewhere in the Flowchart or onanother page of the Flowchart.

• There is usually only one output arrow out of anactivity box. If there is more than one arrow, youmay need a decision diamond.

• Validate the Flowchart with people who are noton the team and who carry out the process ac-tions. Highlight additions or deletions they rec-ommend. Bring these back to the team to discussand incorporate into the final Flowchart.

6. Finalize the Flowchart• Is this process being run the way it should be?• Are people following the process as charted?• Are there obvious complexities or redundancies

that can be reduced or eliminated?• How different is the current process from an ideal

one? Draw an ideal Flowchart. Compare the two(current versus ideal) to identify discrepanciesand opportunities for improvements.

Flow Chart 56-62N 10/7/03, 9:27 AM59


1.0 Determine party size

2.0Find

location

3.0Invite guests

1.1 Decide on budget

1.2 Decide on guest list

2.1 Decide theme

2.2 Select location

3.1 Complete invitations

3.2 Send invitations

• • •

Planning a Party

VariationsThe type of Flowchart just described is sometimes re-ferred to as a “detailed” flowchart because it includes,in detail, the inputs, activities, decision points, andoutputs of any process. Four other forms, describedbelow, are also useful.

Macro FlowchartRefer to the third bulleted item in Step 1 of this sectionfor a description. For a graphic example, see Step 2 ofthe Improvement Storyboard in the Problem-Solving/Process Improvement Model section.

Top-down FlowchartThis chart is a picture of the major steps in a workprocess. It minimizes the detail to focus only on thosesteps essential to the process. It usually does not in-clude inspection, rework, and other steps that result inquality problems. Teams sometimes study the top-down flowchart to look for ways to simplify or reducethe number of steps to make the process more efficientand effective.

Flow Chart 56-62N 10/7/03, 9:27 AM60


Deployment FlowchartThis chart shows the people or departments respon-sible and the flow of the process steps or tasks they areassigned. It is useful to clarify roles and track account-ability as well as to indicate dependencies in the se-quence of events.

Workflow FlowchartThis type of chart is used to show the flow of people,materials, paperwork, etc., within a work setting. Whenredundancies, duplications, and unnecessary com-plexities are identified in a path, people can take actionto reduce or eliminate these problems.

Chris Karin Lauren

No

Yes

Is there time to do graphics?

Draws graphics

Writes ad

Plans ad

Sends adout

Ad completed

Flow Chart 56-62N 10/7/03, 9:27 AM61


FlowchartProposed Patient

Appointment Procedure

Appointment book marked “failure”

Preparation of appointment book

Opening of appointment book

Appoint-ment

shore or fleet?

Appointment issued (PT reminded to

confirm 24 hours prior to

appointment)

Refer to ships corpsman, inform PT they can call at 1500

to make their own appointments for the

next working day

Failure report submitted from front desk to fleet liaison

Did patient call in to

confirm 24 hours prior to

appoint-ment?

Appointment book marked

“patient showed”

Appointment canceled and slot refilled by

new patient

Standby patient placed in

appointment slot

Patient given confirmation number

Fleet liaison sends failure notices to

commands

Yes

Yes No

Fleet

Does patient show for appoint-

ment?

No

Shore

Information provided courtesy ofU.S. Navy, Naval Dental Center, San Diego

Flow Chart 56-62N 10/7/03, 9:27 AM62

©1994 GOAL/QPC Force Field 63

Force FieldAnalysisPositives &negatives of change

Why use it?To identify the forces and factors in place that supportor work against the solution of an issue or problem sothat the positives can be reinforced and/or the nega-tives eliminated or reduced.

What does it do?• Presents the “positives” and “negatives” of a situ-

ation so they are easily compared• Forces people to think together about all the aspects

of making the desired change a permanent one• Encourages people to agree about the relative

priority of factors on each side of the “balance sheet”• Encourages honest reflection on the real underly-

ing roots of a problem and its solution

How do I do it?1. Draw a large letter “T” on a flipchart

a) At the top of the T, write the issue or problem thatyou plan to analyze.

• To the far right of the top of the T, write a descrip-tion of the ideal situation you would like toachieve.

b) Brainstorm the forces that are driving you to-wards the ideal situation. These forces may beinternal or external. List them on the left side.

+

Force Field 63

Force Field 63-65N 10/7/03, 9:28 AM63

©1994 GOAL/QPC64 Force Field

c) Brainstorm the forces that are restraining move-ment toward the ideal state. List them on the rightside.

2. Prioritize the driving forces that can be strength-ened or identify restraining forces that wouldallow the most movement toward the ideal state ifthey were removed• Achieve consensus through discussion or by us-

ing ranking methods such as Nominal GroupTechnique and Multivoting.

Tip When choosing a target for change, rememberthat simply pushing the positive factors for achange can have the opposite effect. It is oftenmore helpful to remove barriers. This tends tobreak the “change bottleneck” rather than justpushing on all the good reasons to change.

Force Field 63-65N 10/7/03, 9:28 AM64

©1994 GOAL/QPC Force Field 65

Force FieldFear of Public Speaking

Not

e: T

he F

orce

Fie

ld c

an h

elp

ind

ivid

ual

s an

d te

ams

sele

ct ta

rget

s fo

r cha

nge.

Gen

eral

ly,

whe

n yo

u f

ocu

s on

res

trai

ning

for

ces,

not

dri

ving

for

ces,

thi

s w

orks

bes

t. Fo

r ex

amp

le,

usi

ng in

dex

car

ds

for

key

poi

nts

may

red

uce

the

fear

“M

ay f

orge

t wha

t to

say.

”

Incr

ease

s se

lf-es

teem

Hel

ps c

aree

r

Com

mun

icat

es id

eas

Con

trib

utes

to a

pla

n/so

lutio

n

Enc

oura

ges

othe

rs to

spe

ak

Hel

ps o

ther

s to

cha

nge

Incr

ease

s en

ergy

of g

roup

Hel

ps c

larif

y sp

eake

r’s

idea

s by

get

ting

feed

back

from

oth

ers

Ham

s ca

n be

ham

s (r

ecog

nitio

n fr

om o

ther

s)

Hel

ps o

ther

s to

see

new

per

spec

tive

Pas

t em

barr

assm

ents

Afr

aid

to m

ake

mis

take

s

Lack

of k

now

ledg

e on

the

topi

c

Afr

aid

peop

le w

ill b

e in

diffe

rent

Afr

aid

peop

le w

ill la

ugh

May

forg

et w

hat t

o sa

y

Too

reve

alin

g of

per

sona

l tho

ught

s

Afr

aid

of o

ffend

ing

grou

p

Fea

r th

at n

ervo

usne

ss w

ill s

how

Lack

of c

onfid

ence

in p

erso

nal a

ppea

ranc

e

+ D

rivi

ng

Fo

rces

Res

trai

nin

g F

orc

es –

Idea

l sta

te: T

o sp

eak

conf

iden

tly,

clea

rly, a

nd c

onci

sely

in a

ny s

ituat

ion.

Force Field 63-65N 10/7/03, 9:28 AM65

©1994 GOAL/QPC66 Histogram

HistogramProcess centering,spread, and shape

Why use it?To summarize data from a process that has been col-lected over a period of time, and graphically present itsfrequency distribution in bar form.

What does it do?• Displays large amounts of data that are difficult

to interpret in tabular form• Shows the relative frequency of occurrence of the

various data values• Reveals the centering, variation, and shape of the

data• Illustrates quickly the underlying distribution of

the data• Provides useful information for predicting future

performance of the process• Helps to indicate if there has been a change in the

process• Helps answer the question “Is the process capable

of meeting my customer requirements?”

How do I do it?1. Decide on the process measure

• The data should be variable data, i.e., measuredon a continuous scale. For example: temperature,time, dimensions, weight, speed.

66 Histogram

Histogram66-75N 10/7/03, 9:29 AM66

©1994 GOAL/QPC Histogram 67

2. Gather data• Collect at least 50 to 100 data points if you plan on

looking for patterns and calculating thedistribution’s centering (mean), spread (varia-tion), and shape. You might also consider collect-ing data for a specified period of time: hour, shift,day, week, etc.

• Use historical data to find patterns or to use as abaseline measure of past performance.

3. Prepare a frequency table from the dataa) Count the number of data points, n, in the sample.

9.99.89.7

10.29.99.39.0

10.09.59.6

10.39.59.99.9

9.39.89.4

10.110.710.2

9.59.79.79.49.69.5

10.410.2

10.210.1

9.69.89.39.29.79.4

10.610.1

9.79.89.39.8

9.49.9

10.010.110.3

9.99.79.89.59.59.79.99.69.3

10.19.79.8

10.39.99.79.89.4

10.110.110.1

9.210.2

9.6

9.69.89.9

10.09.89.99.89.6

10.010.2

9.810.0

9.79.5

9.99.9

10.110.210.3

9.89.3

10.09.89.89.7

10.09.79.6

10.110.010.4

9.89.59.59.6

10.310.1

9.510.0

9.79.7

10.7

9.89.6

10.010.7

9.99.49.79.89.69.3

10.09.7

10.7

In this example, there are 125 data points, n = 125.b) Determine the range, R, for the entire sample.

The range is the smallest value in the set of datasubtracted from the largest value. For ourexample:

R = Xmax – Xmin = 10.7 – 9.0 = 1.7

c) Determine the number of class intervals, k, needed.

Histogram66-75N 10/7/03, 9:29 AM67


For our example, 125 data points would be di-vided into 7–12 class intervals.

Tip These two methods are general rules of thumbfor determining class intervals. In both methods,consider using k = 10 class intervals for ease of“mental” calculation.

Tip The number of intervals can influence the pat-tern of the sample. Too few intervals will pro-duce a tight, high pattern. Too many intervalswill produce a spread out, flat pattern.

d)Determine the class width, H.• The formula for this is:

Number of Data Points

Under 5050 – 100

100 – 250Over 250

Number of Classes (k)

5 – 76 – 107 – 12

10 – 20

H = R = 1.7 = .17 k 10

• Round your number to the nearest value with thesame decimal numbers as the original sample. Inour example, we would round up to .20. It is usefulto have intervals defined to one more decimalplace than the data collected.

e) Determine the class boundaries, or end points.

• Method 1: Take the square root of the total number ofdata points and round to the nearest whole number.

k = 125 = 11.18 = 11 intervals

• Method 2: Use the table below to provide a guide-line for dividing your sample into a reasonablenumber of classes.

Histogram66-75N 10/7/03, 9:29 AM68


• Use the smallest individual measurement in thesample, or round to the next appropriate lowestround number. This will be the lower end pointfor the first class interval. In our example thiswould be 9.0.

• Add the class width, H, to the lower end point.This will be the lower end point for the next classinterval. For our example:

9.0 + H = 9.0 + .20 = 9.20

Thus, the first class interval would be 9.00 andeverything up to, but not including 9.20, that is,9.00 through 9.19. The second class intervalwould begin at 9.20 and be everything up to, butnot including 9.40.

Tip Each class interval must be mutually exclusive,that is, every data point will fit into one, and onlyone class interval.

• Consecutively add the class width to the lowestclass boundary until the k class intervals and/orthe range of all the numbers are obtained.

f) Construct the frequency table based on the valuesyou computed in item “e.”A frequency table based on the data from ourexample is shown below.

1 9.00-9.19 9.1 l 1 2 9.20-9.39 9.3 l l l l l l l l 9 3 9.40-9.59 9.5 l l l l l l l l l l l l l 16 4 9.60-9.79 9.7 l l l l l l l l l l l l l l l l l l l l l l 27 5 9.80-9.99 9.9 l l l l l l l l l l l l l l l l l l l l l l l l l 31 6 10.00-10.19 10.1 l l l l l l l l l l l l l l l l l l 22 7 10.20-10.39 10.3 l l l l l l l l l l 12 8 10.40-10.59 10.5 l l 2 9 10.60-10.79 10.7 l l l l 5 10 10.80-10.99 10.9 0

Class#

Class Boundaries

Mid-Point Frequency Total

Histogram66-75N 10/7/03, 9:29 AM69


4. Draw a Histogram from the frequency table• On the vertical line, (y axis), draw the frequency

(count) scale to cover class interval with the high-est frequency count.

• On the horizontal line, (x axis), draw the scalerelated to the variable you are measuring.

• For each class interval, draw a bar with the heightequal to the frequency tally of that class.

40-

30-

20-

10-

0-

Thickness

Specifications: 9 ± 1.5 USL

Spec.Target

9.2 9.4 9.6 9.8 10.0 10.2 10.4 10.6 10.89.0

Fre

quen

cy

Histogram66-75N 10/7/03, 9:29 AM70


5. Interpret the Histograma) Centering. Where is the distribution centered? Is

the process running too high? Too low?

Processcentered

CustomerRequirement

Processtoo high

Processtoo low

b) Variation. What is the variation or spread of thedata? Is it too variable?

Processwithin

requirements

CustomerRequirements

Processtoo variable

Histogram66-75N 10/7/03, 9:29 AM71


c) Shape. What is the shape? Does it look like anormal, bell-shaped distribution? Is it positivelyor negatively skewed, that is, more data values tothe left or to the right? Are there twin (bi-modal)or multiple peaks?

Tip Some processes are naturally skewed; don’texpect every distribution to follow a bell-shapedcurve.

Tip Always look for twin or multiple peaks indi-cating that the data is coming from two or moredifferent sources, e.g., shifts, machines, people,suppliers. If this is evident, stratify the data.

d)Process Capability. Compare the results of yourHistogram to your customer requirements orspecifications. Is your process capable of meetingthe requirements, i.e., is the Histogram centeredon the target and within the specification limits?

Normal Distribution

Positively Skewed Negatively Skewed

Bi-Modal Distribution Multi-Modal Distribution

Histogram66-75N 10/7/03, 9:29 AM72


Tip Get suspicious of the accuracy of the data if theHistogram suddenly stops at one point (such as aspecification limit) without some previous de-cline in the data. It could indicate that defectiveproduct is being sorted out and is not included inthe sample.

(a) Centered and well withincustomer limits.

Action: Maintain present state.

LowerSpecification

Limit

UpperSpecification

Limit

Centering and Spread Compared toCustomer Target and Limits

Target

(b) No margin for error.Action: Reduce variation.

(c) Process running low.Defective product/service.

Action: Bring averagecloser to target.

(d) Process too variable.Defective product/service.Action: Reduce variation.

(e) Process off center andtoo variable. Defective

product/service.Action: Center better and

reduce variation.

Histogram66-75N 10/7/03, 9:29 AM73


Tip The Histogram is related to the Control Chart.Like a Control Chart, a normally distributed His-togram will have almost all its values within ±3standard deviations of the mean. See Process Capa-bility for an illustration of this.

VariationsStem & Leaf PlotThis plot is a cross between a frequency distributionand Histogram. It exhibits the shape of a Histogram,but preserves the original data values—one of its keybenefits! Data is easily recorded by writing the trailingdigits in the appropriate row of leading digits.

In this example, the smallest value is .057 and thelargest value is .164. Using such a plot, it is easy to findthe median and range of the data.

• Median = middle data value (or average of thetwo middle values) when the data is ranked fromsmallest to largest.

For this example, there are 52 data points. Therefore,the average of the 26th and 27th value will give themedian value.

Median = (.113 + .116)/2 = .1145

• Range = Highest value – lowest value = .164 – .057 = .107

.05 7

.06 4

.07 5 7

.08 1 9 3 9

.09 7 4 8 2 6 9 4

.10 7 2 0 4 3 5 9

.1 1 3 1 9 3 7 3 8 6 6

.12 2 4 8 0 8 9 0 5

.13 2 5 2 7 7 6

.14 0 3 6 9

.15 4 7

.16 4

.05 7

.06 4

.07 5 7

.08 1 3 9 9

.09 2 4 4 6 7 8 9

.10 0 2 3 4 5 7 9

.1 1 1 3 3 3 6 6 7 8 9

.12 0 0 2 4 5 8 8 9

.13 2 2 5 6 7 7

.14 0 3 6 9

.15 4 7

.16 4

Data ascollected

Data rank-ordered

Histogram66-75N 10/7/03, 9:29 AM74


HistogramTime Distribution of Calls

0

10

20

30

40

50

60

8:00

8:30

9:00

9:30

10:0

0

10:3

0

11:0

0

11:3

0

12:0

0

12:3

0

13:0

0

13:3

0

14:0

0

14:3

0

15:0

0

15:3

0

16:0

0

16:3

0

17:0

0

>17

:00

HOTrep data May 22 to August 4

Information provided courtesy ofSmithKline Beecham

Note: The Histogram identified three peak calling periods atthe beginning of the workday and before and after the tradi-tional lunch hour. This can help the HOTreps synchronizestaffing with their customer needs.

Histogram66-75N 10/7/03, 9:29 AM75

©1994 GOAL/QPC76 Interrelationship

What are the issues relatedto reducing litter?

• If using an original statement, (it didn’t comefrom a previous tool or discussion), create a com-

InterrelationshipDigraph (ID)Looking for drivers& outcomes

Why use it?To allow a team to systematically identify, analyze, andclassify the cause and effect relationships that existamong all critical issues so that key drivers or outcomescan become the heart of an effective solution.

What does it do?• Encourages team members to think in multiple

directions rather than linearly• Explores the cause and effect relationships

among all the issues, including the most contro-versial

• Allows the key issues to emerge naturally ratherthan allowing the issues to be forced by a domi-nant or powerful team member

• Systematically surfaces the basic assumptions andreasons for disagreements among team members

• Allows a team to identify root cause(s) even whencredible data doesn’t exist

How do I do it?1. Agree on the issue/problem statement

76 Interrelationship

ID 76-84N 10/7/03, 9:30 AM76

©1994 GOAL/QPC Interrelationship 77

plete sentence that is clearly understood andagreed on by team members.

• If using input from other tools, such as an AffinityDiagram, make sure that the goal under discussionis still the same and clearly understood.

2. Assemble the right team• The ID requires more intimate knowledge of the

subject under discussion than is needed for theAffinity. This is important if the final cause andeffect patterns are to be credible.

• The ideal team size is generally 4–6 people. How-ever, this number can be increased as long as theissues are still visible and the meeting is well facili-tated to encourage participation and maintain focus.

3. Lay out all of the ideas/issue cards that have eitherbeen brought from other tools or brainstormed• Arrange 5–25 cards or notes in a large circular

pattern, leaving as much space as possible fordrawing arrows. Use large, bold printing, includ-ing a large number or letter on each idea for quickreference later in the process.

Lack of respect for

othersLack of

awareness of impact

Inadequate penalties

Notenough

receptacles

Lack of parental

examples

Unneces-sary

packaging

A

D

E

F

C

B

ID 76-84N 10/7/03, 9:30 AM77


4. Look for cause/influence relationships betweenall of the ideas and draw relationship arrows• Choose any of the ideas as a starting point. If all of the

ideas are numbered or lettered, work through themin sequence.

• An outgoing arrow from an idea indicates that itis the stronger cause or influence.

Continued next page

1

3

2

4

A

B

C

D

E

F

Decision: “B” causes or influences “A”

Or?

A

B

C

D

E

F

Decision:No relationship

Or?

A

B

C

D

E

F

Decision:No relationship

Or?

A

B

C

D

E

F

Decision: “E” causes or influences “A”

Or?

Ask of each combination:1) Is there a cause/influence relationship?

2) If yes, which direction of cause/influence is stronger?

ID 76-84N 10/7/03, 9:30 AM78


5 6A

B

C

D

E

F

Decision:No relationship.

"A" is completed.

Or?A

B

C

D

E

F

Decision: “B” causes or influences “C.”

Now begin with “B” and repeat thequestions for all remaining

combinations.

Or?

Tip Draw only one-way relationship arrows in thedirection of the stronger cause or influence. Makea decision on the stronger direction. Do not drawtwo-headed arrows.

5. Review and revise the first round ID• Get additional input from people who are not on

the team to confirm or modify the team’s work.Either bring the paper version to others or repro-duce it using available software. Use a differentsize print or a color marker to make additions ordeletions.

6. Tally the number of outgoing and incoming arrowsand select key items for further planning• Record and clearly mark next to each issue the

number of arrows going in and out of it.• Find the item(s) with the highest number of outgo-

ing arrows and the item(s) with the highest num-ber of incoming arrows.

• Outgoing arrows. A high number of outgoing arrowsindicates an item that is a root cause or driver. This isgenerally the issue that teams tackle first.

ID 76-84N 10/7/03, 9:30 AM79


Lack of respect for

others

Lack of awareness of impact

Inadequate penalties

Not enough receptacles

Lack of parental

examples

Unnecessary packaging

A

B

C

D

E

F

In=1 Out=1

In=4 Out=1

In=1 Out=1.5

In=1.5 Out=1

In=0 Out=5

In=2 Out=0

Driver

Outcome

• Incoming arrows. A high number of incoming ar-rows indicates an item that is a key outcome. Thiscan become a focus for planning either as a mean-ingful measure of overall success or as a redefini-tion of the original issue under discussion.

Tip Use common sense when you select the mostcritical issues to focus on. Issues with very closetallies must be reviewed carefully but in the end,it is a judgment call, not science.

7. Draw the final ID• Identify visually both the key drivers (greatest

number of outgoing arrows) and the key outcomes(greatest number of incoming arrows). Typicalmethods are double boxes or bold boxes.

What are the isues relatedto reducing litter?

Informationprovided courtesy

of CUE Consulting

ID 76-84N 10/7/03, 9:30 AM80


VariationsWhen it is necessary to create a more orderly display ofall of the relationships, a matrix format is very effective.The vertical (up) arrow is a driving cause and thehorizontal (side) arrow is an effect. The example belowhas added symbols indicating the strength of the rela-tionships.The ”total“ column is the sum of all of the ”relationshipstrengths“ in each row. This shows that you are workingon those items that have the strongest effect on thegreatest number of issues.

Cau

se/

Dri

ver

Res

ult

/R

ider

Tota

l3 1 16

0 4 24

2 2 18

1 3 22

4 0 24

LogisticSupport

CustomerSatisfaction

Education& Training

PersonnelIncentives

Leadership

Lo

gis

tic

Su

pp

ort

Cu

sto

mer

Sat

isfa

ctio

n

Ed

uca

tio

n&

Tra

inin

g

Per

son

nel

Ince

nti

ves

Lea

der

ship

ID – Matrix Format

= 9 Significant= 3 Medium= 1 Weak

Relationship Strength

Information provided courtesy of U.S. Air Force, Air Combat Command

ID 76-84N 10/7/03, 9:30 AM81


Interrelationship DigraphIssues Surrounding

Implementation of the Business Plan

Communica-tion issues within the

group

External factors impact

implemen-tation

Means not clearly defined

Plan not integrated

Fast new product

introductions stretch

resources

Lack of time and resources

No strong commitment to the group

In = 3 Out = 2

In = 2 Out = 4

In = 1 Out = 2

In = 0 Out = 2

In = 2 Out = 0

In = 1 Out = 3

Driver

Driver

Planning approach not standardized

In = 0 Out = 5

Outcome

Outcome

In = 5 Out = 0

Capacity may not

meet needs

In = 5 Out = 1

Information providedcourtesy of Goodyear

Note: “The drivers” from this ID will be used as the goal inthe Tree example shown at the end of the Tree Diagram/PDPCsection.

ID 76-84N 10/7/03, 9:30 AM82


Interrelationship DigraphA Vision of Andoverin the 21st Century

Information provided courtesyof Town of Andover, MA

See next page for close up

Controlgrowth

In=5 Out=2

Preserve andpromote thenatural stateof Andover

In=9 Out=1

Encourage avariety ofmeans of

transportation

Total communitycommitment to

the task oflearning

Preservelegacy ofAndover

In=4 Out=7

Citizens treatedas valued

activecustomers

In=0 Out=11

All people’slives valued

and enhanced

In=6 Out=4

Thoughtful andresponsiblefinancing of

valued services

In=4 Out=7

The physicalsystems in town

workingsmoothly

In=4 Out=5

Maintain adiversity in

housing

In=3 Out=5

People shophere for their

needs bychoice

In=6 Out=2

Sense ofcommunity

through townactivities and

facilities

In=4 Out=3

In=4 Out=3

In=5 Out=3

ID 76-84N 10/7/03, 9:30 AM83


Interrelationship DigraphA Vision of Andoverin the 21st Century

Close up

1 This is the driver. If the focus on the citizen as acustomer becomes the core of the town’s vision theneverything else will be advanced.

2 This is the primary outcome. It puts the preserva-tion of nature in the town as a key indicator of thevision working.

Information provided courtesyof Town of Andover, MA

Preserve andpromote thenatural stateof Andover

In=9 Out=1

Preservelegacy ofAndover

In=4 Out=7

Citizens treatedas valued

activecustomers

In=0 Out=11

1

2

Driver

Outcome

ID 76-84N 10/7/03, 9:30 AM84

©1994 GOAL/QPC Matrix 85Matrix 85

Matrix DiagramFinding relationships

Why use it?To allow a team or individual to systematically iden-tify, analyze, and rate the presence and strength ofrelationships between two or more sets of information.

What does it do?• Makes patterns of responsibilities visible and

clear so that there is an even and appropriatedistribution of tasks

• Helps a team get consensus on small decisions, en-hancing the quality and support for the final decision

• Improves a team’s discipline in systematicallytaking a hard look at a large number of importantdecision factors

Types of MatricesMost Common

• L-shaped matrix. Two sets of items directly comparedto each other or a single set compared to itself.

Tourfacility

Reviewpersonnel& safety policies

Review business values

Introduceto team

members

HumanresourcesDivisionmanager

Supervisor

Associates

Tasks

Resources

Orienting New Employees

Conclusion: Supervisors and associates have taken on theorientation role rather than the traditional human resource function.

Primaryresponsibility

Teammembers

Resources

a b c123

Matrix 85-90N 10/7/03, 9:31 AM85

©1994 GOAL/QPC86 Matrix

• T-shaped matrix. Two sets of items compared to acommon third set.

Uncommon• Y-shaped matrix. Three sets of items compared to each

other. It “bends” a T-shaped matrix to allow com-parisons between items that are on the vertical axes.

Rarely Used• X-shaped matrix. Four sets of items compared to

each other. It is essentially two T-shaped matri-ces placed back to back.

Communicate organization

spirit

Communicate purpose of

organization

Resolve practical concerns

Reduce anxiety

GoalsTour

facility

Reviewpersonnel& safetypolicies

Review business values

Introduceto team

members

Human resources

Division manager

Supervisor

Associates

ResourcesTasks

Conclusion: The most important purpose of orientation is to reduce anxiety, and the most effective tasks focus on the personal issues.

Orienting New Employees

ResponsibilityPrimaryTeam membersResources

ImpactHighMediumLow

Matrix 85-90N 10/7/03, 9:31 AM86

©1994 GOAL/QPC Matrix 87

• C-shaped matrix. Shows the intersection of threesets of data simultaneously. It is a three-dimen-sional graphic.

• You can find more complete information on the Y-,X-, and C-shaped matrix in The Memory JoggerPlus+®.

How do I do it?1. Select the key factors affecting successful imple-

mentation• The most important step is to choose the issues or

factors to be compared. The format is secondary.Begin with the right issues and the best formatwill define itself. The most common use is thedistribution of responsibilities within an L-shaped or T-shaped matrix.

2. Assemble the right team• Select individuals that have the influence/power

to realistically assess the chosen factors.Tip When distributing responsibilities, include

those people who will likely be involved in theassigned tasks or who can at least be part of areview team to confirm small group results.

3. Select an appropriate matrix format• Base your choice of format on the number of sets of

items and types of comparisons you need to make.

4. Choose and define relationship symbols• The most common symbols in matrix analysis are

, , . Generally they are used to indicate:= High = 9= Medium = 3= Low = 1

Matrix 85-90N 10/7/03, 9:31 AM87


• The possible meanings of the symbols are almostendless. The only requirement is that the teamcomes to a clear understanding and creates anequally clear legend with the matrix.

5. Complete the matrix• If distributing responsibilities, use only one “pri-

mary responsibility” symbol to show ultimateaccountability. All other core team members canbe given secondary responsibilities.

Tip Focus the quality of the decision in each matrixcell. Do not try to “stack the deck” by consciouslybuilding a pattern of decisions. Let these patternsemerge naturally.

Tip Interpret the matrix using total numerical val-ues only when it adds value. Often the visualpattern is sufficient to interpret the overall results.

VariationsThe matrix is one of the most versatile tools available.The important skill to master is “matrix thinking.” Thisapproach allows a team to focus its discussion on re-lated factors that are explored thoroughly. The separateconclusions are then brought together to create high-quality decisions. Use your creativity in determiningwhich factors affect each other, and in choosing thematrix format that will help focus the discussion to-ward the ultimate decision.

Matrix 85-90N 10/7/03, 9:31 AM88

©1994 GOAL/QPC Matrix 89

MatrixLogistics Annual Plan

LQC

Obj

ectiv

es(M

atrix

)

Mea

sure

s

Sch

edul

es(A

ND

)

Res

earc

hcu

stom

er n

eeds

via

QF

D

Cap

ture

cus

tom

erco

mm

ents

Reducecustomer cost

Sur

vey

cust

omer

satis

fact

ion

Continueimplementation of total quality

Continue upgradingtech., prof., &

managerial skills ofemployeesPromote

environmentalresponsibility inour operations

Del

ight

our

cust

omer

s

%sa

tisfa

ctio

nvi

a su

rvey

List

of

cust

omer

need

s by

key

proc

esse

s#

ofco

mm

ents

or

# of

com

plai

nts

2003

2

004

TQ

Impl

emen

tatio

n(T

ree)

1 See

next

page

Con

tinue

toim

plem

ent

tota

l qua

lity

Goa

l

=

9 S

tron

g in

fluen

ce/r

elat

ions

hip

=

3 S

ome

influ

ence

/rel

atio

nshi

p

= 1

Wea

k in

fluen

ce/r

elat

ions

hip

Bla

nk =

No

influ

ence

/rel

atio

nshi

p

Qua

rter

Qua

rter

1

2

3

4

1

2

3

4

Info

rmat

ion

prov

ided

cou

rtes

y of

Bel

l Can

ada

Matrix 85-90N 10/7/03, 9:31 AM89


MatrixLogistics Annual Plan

Goa

ls(A

ND

)C

o-R

espo

nsib

ility

(Mat

rix)

Cos

t/Ben

efit

Ana

lysi

s

Status*

Sur

vey

cust

omer

satis

fact

ion

Res

earc

hcu

stom

er n

eeds

via

QF

D

Cap

ture

cus

tom

erco

mm

ents

2003

2

004

LQC

Departments

Boards ofmanagement

Logiqual

Otherstakeholders

Resourcesrequired ($)

Tangiblebenefits ($)

75%

cust

omer

satis

fact

ion

80%

cust

omer

satis

fact

ion

Fie

ld tr

ial

100%

cust

omer

need

sga

ther

ed

1%tra

nsac

tions

1.9%

03

0

4

03

0

4

25K

25K

TQ

Impl

emen

tatio

n(T

ree)

1fr

ompr

evio

uspa

ge

= 9

Prim

e re

spon

sibi

lity

= 3

Sec

onda

ry r

espo

nsib

ility

= 1

Kep

t inf

orm

ed

25K

25K

Cau

tion

Sto

pped

On

targ

et*S

tatu

s:

Del

ight

our

cust

omer

s

Con

tinue

toim

plem

ent

tota

l qua

lity

Goa

l

Info

rmat

ion

prov

ided

cou

rtes

y of

Bel

l Can

ada

Matrix 85-90N 10/7/03, 9:31 AM90

©1994 GOAL/QPC NGT/Multivoting 91

Nominal GroupTechnique (NGT)Ranking for consensus

Why use it?Allows a team to quickly come to a consensus on therelative importance of issues, problems, or solutions bycompleting individual importance rankings into ateam’s final priorities.

What does it do?• Builds commitment to the team’s choice through

equal participation in the process• Allows every team member to rank issues with-

out being pressured by others• Puts quiet team members on an equal footing

with more dominant members• Makes a team’s consensus (or lack of it) visible;

the major causes of disagreement can be dis-cussed

How do I do it?1. Generate the list of issues, problems, or solutions

to be prioritized• In a new team with members who are not accus-

tomed to team participation, it may feel safer todo written, silent brainstorming, especially whendealing with sensitive topics.

2. Write statements on a flipchart or board

A.B.C.D.

2+21+13+44+3

A.B.C.D.

2+21+13+44+3

NGT/Multivoting 91

Nominal Group 91-94N 10/7/03, 9:34 AM91

©1994 GOAL/QPC92 NGT/Multivoting

3. Eliminate duplicates and/or clarify meanings ofany of the statements• As a leader, always ask for the team’s permission

and guidance when changing statements.

4. Record the final list of statements on a flipchartor boardExample: Why does the department have inconsis-tent output?

A Lack of trainingB No documented processC Unclear quality standardsD Lack of cooperation with other

departmentsE High turnover

• Use letters rather than numbers to identify eachstatement so that team members do not get con-fused by the ranking process that follows.

5. Each team member records the corresponding letterson a piece of paper and rank orders the statementsExample: Larry’s sheet of paper looks like this:

A 4B 5C 3D 1E 2

• This example uses “5” as the most importantranking and “1” as the least important. Sinceindividual rankings will later be combined, this“reverse order” minimizes the effect of team


©1994 GOAL/QPC NGT/Multivoting 93

members leaving some statements blank. There-fore, a blank (value = 0) would not, in effect,increase its importance.

6. Combine the rankings of all team members

Larry Nina Norm Paige Si TotalA 4 5 2 2 1 = 14B 5 4 5 3 5 = 22C 3 1 3 4 4 = 15D 1 2 1 5 2 = 11E 2 3 4 1 3 = 13

“No documented process,” B, would be the highestpriority. The team would work on this first and thenmove through the rest of the list as needed.

VariationsOne Half Plus OneWhen dealing with a large number of choices it may benecessary to limit the number of items ranked. The “onehalf plus one” approach would rank only a portion ofthe total. For example, if 20 ideas were generated, thenteam members would rank only the top 11 choices. Ifneeded, this process could be repeated with the remain-ing 9 items, ranking the top 5 or 6 items, (half of 9 = 4.5+ 1 = 5.5), until a manageable number are identified.

Weighted MultivotingEach team member rates, not ranks, the relative impor-tance of choices by distributing a value, e.g., 100 points,across the options. Each team member can distributethis value among as many or as few choices as desired.


©1994 GOAL/QPC94 NGT/Multivoting

Example:

Larry Nina Norm Paige Si TotalA 20 10 = 30B 40 80 50 100 45 = 315C 30 5 10 25 = 70D 5 10 20 = 35E 10 10 20 10 = 50

With large numbers of choices, or when the voting forthe top choices is very close, this process can be re-peated for an agreed upon number of items. Stop whenthe choice is clear.


©1994 GOAL/QPC Pareto 95

Pareto ChartFocus on key problems

Why use it?To focus efforts on the problems that offer the greatestpotential for improvement by showing their relativefrequency or size in a descending bar graph.

What does it do?• Helps a team to focus on those causes that will

have the greatest impact if solved• Based on the proven Pareto principle: 20% of the

sources cause 80% of any problem• Displays the relative importance of problems in a

simple, quickly interpreted, visual format• Helps prevent “shifting the problem” where the “so-

lution” removes some causes but worsens others• Progress is measured in a highly visible format

that provides incentive to push on for more im-provement

How do I do it?1. Decide which problem you want to know more

aboutExample: Consider the case of HOTrep, an inter-nal computer network help line: Why do peoplecall the HOTrep help line; what problems arepeople having?

2. Choose the causes or problems that will be moni-tored, compared, and rank ordered by brainstorm-ing or with existing dataa) Brainstorming

Pareto 95

Pareto 95-104N 10/7/03, 9:34 AM95

©1994 GOAL/QPC96 Pareto

Example: What are typical problems that usersask about on the HOTrep help line?

b) Based on existing dataExample: What problems in the last month haveusers called in to the HOTrep help line?

3. Choose the most meaningful unit of measure-ment such as frequency or cost• Sometimes you don’t know before the study

which unit of measurement is best. Be preparedto do both frequency and cost.Example: For the HOTrep data the most impor-tant measure is frequency because the projectteam can use the information to simplify soft-ware, improve documentation or training, orsolve bigger system problems.

4. Choose the time period for the study• Choose a time period that is long enough to

represent the situation. Longer studies don’t al-ways translate to better information. Look first atvolume and variety within the data.

• Make sure the scheduled time is typical in orderto take into account seasonality or even differentpatterns within a given day or week.Example: Review HOTrep help line calls for 10weeks (May 22–August 4).

5. Gather the necessary data on each problem categoryeither by “real time” or reviewing historical data• Whether data is gathered in “real time” or his-

torically, check sheets are the easiest method forcollecting data.Example: Gathered HOTrep help line calls databased on the review of incident reports (historical).

Pareto 95-104N 10/7/03, 9:34 AM96


7. List the problem categories on the horizontal lineand frequencies on the vertical line• List the categories in descending order from left

to right on the horizontal line with bars aboveeach problem category to indicate its frequency orcost. List the unit of measure on the vertical line.

8. (Optional) Draw the cumulative percentage lineshowing the portion of the total that each problemcategory representsa) On the vertical line, (opposite the raw data, #, $,

etc.), record 100% opposite the total number and50% at the halfway point. Fill in the remainingpercentages drawn to scale.

Tip Always include with the source data and thefinal chart, the identifiers that indicate the source,location, and time period covered.

6. Compare the relative frequency or cost of eachproblem category

Example:

Problem Category Frequency Percent (%)Bad configuration 3 1Boot problems 68 33File problems 8 4Lat. connection 20 10Print problems 16 8Reflection hang 24 12Reflection sys. integrity 11 5Reflections misc. 6 3System configuration 16 8System integrity 19 9Others 15 7

Total 206

Pareto 95-104N 10/7/03, 9:34 AM97


b) Starting with the highest problem category, drawa dot or mark an x at the upper righthand cornerof the bar.

• Add the total of the next problem category to thefirst and draw a dot above that bar showing boththe cumulative number and percentage. Connectthe dots and record the remaining cumulativetotals until 100% is reached.

Information provided courtesy of SmithKline Beecham

HOTrep Problem Data220

200

180

160

140

120

100

80

60

40

20

Boo

t pro

blem

s (6

8)

Ref

lect

ion

hang

(24

)

Lat.

conn

ectio

n (2

0)

Sys

tem

inte

grity

(19

)

Sys

tem

con

fig. (

16)

Prin

t pro

blem

s (1

6)

Ref

lect

ion

syst

. int

egrit

y (1

1)

File

pro

blem

s (8

)

Ref

lect

ions

mis

c. (

6)

Bad

con

figur

atio

n (3

)

Oth

ers

(15)

100%

90

80

70

60

50

40

30

20

10

33%

12%10% 9% 8% 8% 5% 4% 3% 1% 7%

45%

55%

64%

72%

80%

85%

89%92%

93%

# of

Rep

orte

d O

ccur

renc

es

Pareto 95-104N 10/7/03, 9:34 AM98


9. Interpret the results• Generally, the tallest bars indicate the biggest con-

tributors to the overall problem. Dealing withthese problem categories first therefore makescommon sense. But, the most frequent or expen-sive is not always the most important. Alwaysask: What has the most impact on the goals of ourbusiness and customers?

VariationsThe Pareto Chart is one of the most widely and cre-atively used improvement tools. The variations usedmost frequently are:

A. Major Cause Breakdowns in which the “tallestbar” is broken into subcauses in a second, linkedPareto.

B. Before and After in which the “new Pareto” barsare drawn side by side with the original Pareto,showing the effect of a change. It can be drawn asone chart or two separate charts.

C. Change the Source of Data in which data iscollected on the same problem but from differentdepartments, locations, equipment, and so on,and shown in side-by-side Pareto Charts.

D. Change Measurement Scale in which the samecategories are used but measured differently.Typically “cost” and “frequency” are alternated.

Pareto 95-104N 10/7/03, 9:34 AM99


Information provided courtesy of Goodyear

50

40

30

20

10

0

Reduced Payment Freight BillsContract Rate Disputes

120110100

908070605040302010

0

Cla

ssReduced Payment Freight Bills

Total Bills (329)

Adv

ance

d ca

rrie

r

Load

ed to

full

visi

ble

capa

city

Del

iver

y to

min

e si

tes

Min

. chr

g.

NY

C a

rb.

Mis

c.

SW

C

NO

A

% o

f con

trac

t dis

pute

s#

of b

ills

Con

trac

t rat

edi

sput

es

Can

ada

Orig

inal

dest

inat

ion

Mis

c.

Wei

ght

Rec

onci

led

Deb

t

Byn

d.

ParetoA. Major Cause Breakdowns

Pareto 95-104N 10/7/03, 9:34 AM100


120

110

100 90 80 70 60 50 40 30 20 10 0

Janu

ary

bills

(32

9)

Red

uced

Pay

men

t Fre

ight

Bill

s–

Afte

r S

tand

ardi

zatio

n –

June

bill

s (5

6)

# of bills

Contract rate disputes

Class

Canada

Original destination

Misc.

Weight

Reconciled

Debt

Bynd.

Info

rmat

ion

prov

ided

cou

rtes

y of

Goo

dyea

r

ParetoB. Before and After

Pareto 95-104N 10/7/03, 9:34 AM101



0

5

10

15

20

25

30

35

For

got

Wor

kloa

d

Per

sona

lB

usin

ess

Leav

e

Mis

c.

Tran

sfer

red

Vehi

cle

% 31 25 21 8 8 4 2

% o

f fai

led

appo

intm

ents

Reason for Failed AppointmentsSource of Data is: Shore Commands

ParetoC. Change the Source of Data

Pareto 95-104N 10/7/03, 9:34 AM102


Reason for Failed AppointmentsSource of Data is: Fleet Commands

ParetoC. Change the Source of Data

0

5

10

15

20

25

30

Cor

psm

an

Und

erw

ay

For

got

Per

sona

lB

usin

ess

Vehi

cle

TAD

/Tr

ansf

er

Sch

ool

% 27 25 23 9 7 5 4

% o

f fai

led

appo

intm

ents


Pareto 95-104N 10/7/03, 9:34 AM103


ParetoD. Change Measurement Scale

25

20

15

10

5

0

Shi

ppin

g

Inst

alla

tion

Del

iver

y

Cle

rical

Mis

c.

42% of all complaints

Field Service Customer Complaints

# of

com

plai

nts

20,000

15,000

10,000

5,000

0

Inst

alla

tion

Cle

rica

l

Ship

pin

g

Deliv

ery

Mis

c.

13% of total cost

Cost to Rectify Field Service Complaints

Dolla

rs (

$)

Pareto 95-104N 10/7/03, 9:34 AM104

©1994 GOAL/QPC Prioritization 105

PrioritizationMatricesWeighing your options

Why use it?To narrow down options through a systematic ap-proach of comparing choices by selecting, weighting,and applying criteria.

What does it do?• Quickly surfaces basic disagreements so they

may be resolved up front• Forces a team to focus on the best thing(s) to do,

and not everything they could do, dramaticallyincreasing the chances for implementation success

• Limits “hidden agendas” by surfacing the criteriaas a necessary part of the process

• Increases the chance of follow-through becauseconsensus is sought at each step in the process(from criteria to conclusions)

• Reduces the chances of selecting someone’s “petproject”

How do I do it?There are three methods for constructing PrioritizationMatrices. The outline that follows indicates typicalsituations for using each method. Only the “Full Ana-lytical Criteria Method” is discussed here. The othersare covered fully in The Memory Jogger Plus+®.

a b c

ParetoPrioritization 105

Prioritization 105-114N 10/7/03, 9:35 AM105

©1994 GOAL/QPC106 Prioritization

Full Analytical Criteria Method

Typically use when:• Smaller teams are involved (3–8 people)• Options are few (5–10 choices)• There are relatively few criteria (3–6 items)• Complete consensus is needed• The stakes are high if the plan fails

Consensus Criteria MethodThis method follows the same steps as in the FullAnalytical Criteria Method except the ConsensusCriteria Method uses a combination of weightedvoting, and ranking is used instead of paired com-parisons.

Typically use when:• Larger teams are involved (8 or more people)• Options are many (10–20 choices)• There are a significant number of criteria (6–15

items)• Quick consensus is needed to proceed

Combination ID/Matrix MethodThis method is different from the other two methodsbecause it is based on cause and effect, rather thancriteria.

Typically use when:• Interrelationships among options are high and

finding the option with the greatest impact iscritical



Full Analytical Criteria Method

1. Agree on the ultimate goal to be achieved in aclear, concise sentence• If no other tools are used as input, produce a clear

goal statement through consensus. This state-ment strongly affects which criteria are used.

Choose the most enjoyable vacation for the whole family

2. Create the list of criteria• Brainstorm the list of criteria or review previous

documents or guidelines that are available, e.g.,corporate goals, budget-related guidelines.

• Cost• Educational value• Diverse activity• Escape reality

Tip The team must reach consensus on the finalcriteria and their meanings or the process is likelyto fail!

3. Using an L-shaped matrix, weight each criterionagainst each other• Reading across from the vertical axis, compare

each criterion to those on the horizontal axis.• Each time a weight (e.g., 1, 5, 10) is recorded in a

row cell, its reciprocal value (e.g., 1/5, 1/10) must berecorded in the corresponding column cell.

• Total each horizontal row and convert to a relativedecimal value known as the “criteria weighting.”



4. Compare ALL options relative to each weightedcriterion• For each criterion, create an L-shaped matrix

with all of the options on both the vertical andhorizontal axis and the criteria listed in thelefthand corner of the matrix. There will be asmany options matrices as there are criteria tobe applied.

• Use the same rating scale (1, 5, 10) as in Step 3,BUT customize the wording for each criterion.

• The relative decimal value is the “option rating.”

CostEduc.value

Diverseactivity

Escapereality

RowTotal

RelativeDecimal

Value

Cost 5

5

10 5

5

5

5.3

10.2

20

.60

.15

.28

.55

.02

Criteria

Criteria

Grand Total 36.1

Row TotalRating scores addedGrand TotalRow totals addedRelative Decimal ValueEach row total ÷ by the grand total

Educ.value

Diverseactivity

Escapereality

110

15

15

15

15

15

= Equally important= More important= Much more important= Less Important= Much less important

15

101/5

1/10

Criterion vs. Criterion



Tip The whole number (1, 5, 10) must always rep-resent a desirable rating. In some cases this maymean “less,” e.g., cost, in others this may mean“more,” e.g., tasty.

110

Disney World

Gettys-burg

New York City

Uncle Henry's

Row Total

RelativeDecimal

Value

5

10 5

5.3

15.2

.40

25

.12

.33

.01

.54

Grand Total 45.9

Cost

Disney World

Gettys-burg

New York City

Uncle Henry's

5

10

10

Continue Step 4 through three more Options/Criterion matrices, like this:

15

110

110

15

15

Options vs. Each Criterion (Cost Criterion)

= Equal cost= Less expensive= Much less expensive= More expensive= Much more expensive

15

101/5

1/10

Escape reality

Options

Opt

ions

Crt .

Diverse activity

Options

Opt

ions

Crt .

Educational value

Options

Opt

ions

Crt .



5. Using an L-shaped summary matrix, compareeach option based on all criteria combined• List all criteria on the horizontal axis and all

options on the vertical axis.• In each matrix cell multiply the “criteria weight-

ing” of each criterion (decimal value from Step 3)by the “option rating” (decimal value from Step4). This creates an “option score.”

• Add each option score across all criteria for a rowtotal. Divide each row total by the grand total andconvert to the final decimal value. Compare thesedecimal values to help you decide which optionto pursue.

.54 x .15(.08)

Cost(.15)

Educa-tionalvalue(.28)

Diverse activity(.55)

Escape reality(.02)

Row Total

RelativeDecimal

Value(RT ÷ GT)

Criteria

Disney World

Gettys-burg

New York City

Uncle Henry's

Optns.

.12 x .15(.02)

.33 x .15(.05)

.01 x .15(0)

.24 x .28(.07)

.37 x .28(.10)

.01 x .28(0)

.37 x .28(.10)

.40 x .55(.22)

.10 x .55(.06)

.49 x .55(.27)

.01 x .55(.01)

.65 x .02(.01)

.22 x .02(0)

.12 x .02(0)

.01 x .02(0)

.32

.22

.37

.09

.32

.22

.38

.09

.54 x .15(from Step 4 matrix) (from Step 3 matrix)

(.08)Option score

1.00Grand Total

Summary MatrixOptions vs. All Criteria



6. Choose the best option(s) across all criteriaTip While this is more systematic than traditional

decision making, it is not a science. Use commonsense and judgment when options are rated veryclosely, but be open to non-traditional conclusions.

VariationsSee The Memory Jogger Plus+® for full explanations ofboth the Consensus Criteria Method and the Combina-tion ID/Matrix Method. The Full Analytical CriteriaMethod, illustrated in this book, is recommended be-cause it encourages full discussion and consensus oncritical issues. The Full Analytical Criteria Method is asimplified adaptation of an even more rigorous modelknown as the Analytical Hierarchy Process. It is basedon the work of Thomas Saaty, which he describes in hisbook Decision Making for Leaders. In any case, use com-mon sense to know when a situation is importantenough to warrant such thorough processes.



PrioritizationChoosing a Standard

Corporate Spreadsheet ProgramB

est u

seof

har

dwar

e

Eas

e of

use

Max

imum

func

tiona

lity

Bes

tpe

rfor

man

ce

Tota

l(1

4 cr

iteria

)

Rel

ativ

eD

ecim

al V

alue

Best useof hardware .20 .10 3.7 .01

Ease of use 5.0 35.4 .08

Maximumfunctionality 10.0 69.0 .17

Bestperformance 45.2 .11

Grand Total(14 criteria) 418.1

5.0

5.05.0

5.0

.20

.20

.20

.20

Criteria

Information provided courtesy of Novacor Chemicals

Note: This constructed example, illustrated on three pages,represents only a portion of the prioritization process and onlya portion of Novacor’s spreadsheet evaluation process.Novacor Chemicals assembled a 16-person team, comprisedmainly of system users and some information systems staff.The team developed and weighted 14 standard criteria andthen applied them to choices in word processing, spreadsheet,and presentation graphics programs.

1 Weighting criteria (described in Step 3)This is a portion of a full matrix with 14 criteria in total.

This example continued next page




Corporate Spreadsheet Program (cont.)

Best integration–internal

Total

1.00 2.00 .33

1.00

1.00

6.00

1.00

1.00

1.00 2.00

2.00

.33

.33

Relative Decimal

Value

ProgramA

ProgramB

Program

C

ProgramA

ProgramB

ProgramC

Grand Total

This example continued next page

Lowest ongoing

costTotal

.10 .20 .30 .02

10.00 5.00 15.00 .73

5.00 .20 5.20 .25

20.50

ProgramA

ProgramB

ProgramC

ProgramA

ProgramB

ProgramC

Relative Decimal

Value

Grand Total


2 Comparing options (described in Step 4)These are just 2 of 14 matrices.




Corporate Spreadsheet Program (cont.)

3 Summarize Option Ratings Across All Criteria (described in Step 5)

This is a portion of a full matrix with 14 criteria in total.

Criteria Total(across 14

criteria)

.16 .18

.48 (.04) .30 .33

.49

.90

ProgramA

ProgramB

ProgramC

Easy to use (.08)

Best integration int. (.09)

Lowest ongoing

cost (.08)

.44

Relative Decimal

Value

.03 (.01)

.48 (.04)

.33 (.03)

.33 (.03)

.33 (.03)

.02 (0)

.73 (.06)

.25 (.02)

Grand Total

Options

Result: Program C was chosen. Even though 14 out of the 16team members were not currently using this program, theprioritization process changed their minds, and preventedthem from biasing the final decision.



©1994 GOAL/QPC PS/PI Model 115

Problem-Solving/Process-Improvement ModelImprovementStoryboard

There are many standard models for making improve-ments. They all attempt to provide a repeatable set ofsteps that a team or individual can learn and follow. TheImprovement Storyboard is only one of many modelsthat include typical steps using typical tools. Followthis model or any other model that creates a commonlanguage for continuous improvement within yourorganization.

Plan

1. Select the problem/process that will be ad-dressed first (or next) and describe the improve-ment opportunity.

2. Describe the current process surrounding theimprovement opportunity.

3. Describe all of the possible causes of the prob-lem and agree on the root cause(s).

4. Develop an effective and workable solution andaction plan, including targets for improvement.

Do

5. Implement the solution or process change.

Check

6. Review and evaluate the result of the change.

Act

7. Reflect and act on learnings.

P

D

C

A

PS/PI Model 115

PS/PI 115-131N 10/7/03, 9:37 AM115

©1994 GOAL/QPC116 PS/PI Model

Depending on your formal process structure, Step 1may be done by a steering committee, managementteam, or improvement team. If you are an improvementteam leader or member, be prepared to start with Step 1 orStep 2.

1. Select the problem/process that will be addressedfirst (or next) and describe the improvement op-portunity.• Look for changes in important business indicators• Assemble and support the right team• Review customer data• Narrow down project focus. Develop project pur-

pose statement

Typical toolsBrainstorming, Affinity Diagram, Check Sheet, ControlChart, Histogram, Interrelationship Digraph, ParetoChart, Prioritization Matrices, Process Capability, Ra-dar Chart, Run Chart

SituationStop ’N Go Pizza* is a small but recently growing pizzadelivery business with six shops. After a period of rapidgrowth, Stop ’N Go Pizza experienced a six-month declinein volume. Customers were leaving. Top managementformed a mixed team of store managers, kitchen staff,and delivery personnel to find out why, and to generatean implementation plan to correct the situation. Theteam used both the Run Chart and Pareto Chart.

* The name Stop ’N Go Pizza, and the data associated with thiscase study are fictional. Any similiarity to an actual companyby this name is purely coincidental.

Plan

PS/PI 115-131N 10/7/03, 9:37 AM116


Run ChartAverage monthly volume of deliveries

(per shop)2400

2100

1800

1500

1200

900

600

300

0Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Un

it V

olu

me

1,951deliveries

Pareto ChartTypes of customer complaintsTotal=2520 October-December

(across 6 shops)2500

2000

1500

1000

500

0

Late

del

iver

ies

Wro

ng o

rder

Col

d fo

od

Tast

e

Oth

er

100%

75

50

25

(1890)

(220) (206) (117) (87)Tota

l # o

f Cus

tom

er C

ompl

aint

s

Illustration note: Delivery time was defined by the total time from when the order was placed to when the customer received it.

PS/PI 115-131N 10/7/03, 9:37 AM117


Pareto ChartLate delivery complaints

Total=1890 October-December(across 6 shops)

# of

Lat

e D

eliv

ery

Com

plai

nts

400350300250200150100500

C A B F D E

Shops

= Friday

= Saturday

= Other

(391)(358)

(313)(295)

(275) (258)

DecisionLate deliveries, (late from the time of order), were by farthe most frequent customer complaint across all loca-tions, especially on Fridays and Saturdays.

Team purpose statementReduce late deliveries on Fridays and Saturdays.

2. Describe the current process surrounding the im-provement opportunity.• Select the relevant process or process segment to

define the scope of the project• Describe the process under study

Typical toolsBrainstorming, Macro, Top-down, and DeploymentFlowcharts, Tree Diagram

SituationThe team began to understand the overall process forproducing and delivering their product and problems

PS/PI 115-131N 10/7/03, 9:37 AM118


that contributed to the project focus. The team used aMacro Flowchart.

Receive order

Prepareingredients

Bakepizza

Assembleorder

Deliver orderReceivepaymentReturn to shop

Process for producing and deliveringStop ’N Go Pizza

DecisionIt became obvious that the “late deliveries” went farbeyond the physical delivery process. Everything inthe Macro Flowchart affected the “order-to-eating”time. This total process improvement became theteam’s focus.

3. Describe all of the possible causes of the problemand agree on the root cause(s).• Identify and gather helpful facts and opinions on

the cause(s) of the problem• Confirm opinions on root cause(s) with data

whenever possible

Typical toolsAffinity Diagram, Brainstorming, C & E/Fishbone Dia-gram, Check Sheet, Force Field Analysis, Interrelation-ship Digraph, Multivoting, Nominal Group Tech-nique, Pareto Chart, Run Chart, Scatter Diagram

PS/PI 115-131N 10/7/03, 9:37 AM119


SituationThe team brainstormed all of the possible causes for“late deliveries” and then continued to ask “Why?” sothat possible cause patterns could emerge. The teamused a C & E/Fishbone Diagram and Run Charts.

C &

E/F

ish

bo

ne

Dia

gra

mR

easo

ns fo

r la

te p

izza

del

iver

ies

Late

piz

za

deliv

erie

s on

F

riday

s &

S

atur

days

Peo

ple

don'

t sho

w u

p

Hig

h tu

rnov

erLo

w p

ay

No

team

wor

k No

trai

ning

Peo

ple

Don

't kn

ow to

wn

Driv

ers

get l

ost

Hig

h tu

rnov

er

Get

wro

ng

info

rmat

ion

Poo

r tr

aini

ng

Rus

hed

Mac

hine

ry/E

quip

men

t

Unr

elia

ble

cars

Low

pay

Kid

s ow

n ju

nks

No

mon

ey fo

r re

pairs

Ove

ns to

o sm

all

Poo

r us

e of

spa

cePoo

r tr

aini

ngH

igh

turn

over

No

capa

city

for

peak

per

iods

Met

hods

Poo

r ha

ndlin

g of

la

rge

orde

rs

Lack

of e

xper

ienc

eP

oor

disp

atch

ing

Man

y ne

w s

tree

ts

Don

't kn

ow

tow

n Hig

h tu

rnov

er

Run

out

of i

ngre

dien

ts

Inac

cura

te

orde

ring

Poo

r us

e of

spa

ce

Lack

of

trai

ning M

ater

ials

Hig

h tu

rnov

er

Hig

h tu

rnov

er

PS/PI 115-131N 10/7/03, 9:37 AM120


Run Charts

Average turnover rate of employees(company-wide)

70

60

50

40

30

20

10

0

43.25%

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

%

Average training hours of new employees

14

12

10

8

6

4

2

0

8 hours

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

Ave

rage

# o

f Hou

rs

PS/PI 115-131N 10/7/03, 9:37 AM121


DecisionThe C & E/Fishbone Diagram repeatedly pointed to“turnover” and “lack of training” as root causes. Thisapplied to ALL areas of the operation, NOT just in theactual delivery portion of the process. Subsequent datashowed that as the business rapidly grew, less time wasput into training all employees. With this lack of adequatetraining, many employees felt a great deal of pressure atthe busiest times. They also felt that they were unable to dotheir jobs well. This combination of work pressure andlack of self-confidence often caused employees to quit.

4. Develop an effective and workable solution andaction plan, including targets for improvement.• Define and rank solutions• Plan the change process: What? Who? When?• Do contingency planning when dealing with new

and risky plans• Set targets for improvement and establish

monitoring methods

Typical toolsActivity Network Diagram, Brainstorming, Flowchart,Gantt Chart, Multivoting, Nominal Group Technique,PDPC, Prioritization Matrices, Matrix Diagram, TreeDiagram

SituationThe team used the combination of the Tree Diagram,Prioritization Matrices, Responsibility Matrix, andGantt Chart to create a workable plan that attacked theheart of the problem.

DecisionThe team focused on the most effective, efficient way todeliver the new training. They developed a four-month

PS/PI 115-131N 10/7/03, 9:37 AM122


Tree DiagramReduce turnover rate of employees

(company-wide)

implementation plan that featured the creative use ofvideotapes, role plays, peer ratings, and so on. The teamset the following targets based on past performance:

• Reduce turnover rate from 62 to 30 percent• Reduce average time of order-to-delivery from 40

to 25 minutes• Reduce customer complaints of late deliveries by

50 percent, without increasing other complaintcategories

• Increase average monthly volume to 2400 unitsper shop from the current 1891 units

Reduce turnover through

improved training in all facets of

operation

Delivertraining

conveniently

Delivertraining

effectively

At convenient

times

At convenientlocations

Before shift

During shift

After shift

At shop

At home

Hold trainee interest

Improve performance

Develop videos

Use texts

Use role play

Evaluate each other

Pre/post tests

Set clear performance

standards

PS/PI 115-131N 10/7/03, 9:37 AM123


Prioritization MatrixSelecting the best training program components

The total = the sum of [rating values x criteria weighting]For example, to find the total of the “Train before shift” row,do the following:[ (9) x .60] + [ (9) x .19] + [ (3) x .19] + [ (3) x .01] = 7.70Note: Weighting values of each criterion come from a criteria

matrix not shown.Task options come from the most detailed level of theTree Diagram on the previous page.

Effe

ctiv

enes

s

Fea

sibi

lity

Tim

e

Cos

t

Tota

l

Train before shift 7.70

Train during shift 3.78

Train after shift 1.77

Train at the shop

Train at home 5.29

Develop videos 8.83

Use texts 4.16

Use role play 8.89

Evaluate each other

Pre/post test 8.83

Set clear perfor-mance standards 8.89

Criteria & Weighting

Tasks & Options

7.70

7.70

(.60) (.19) (.19) (.01)

= 9 Excellent = 3 Fair = 1 Poor

PS/PI 115-131N 10/7/03, 9:37 AM124


Matrix & Gantt Chart CombinedNew training program timeline

Managers

Employees

Human resources

President

Janu

ary

Feb

ruar

yM

arch

Apr

il

Trai

n at

the

shop

be

fore

the

shift

Dev

elop

vid

eos

Use

rol

e pl

ay

Eva

luat

e ea

ch o

ther

Use

pre

/pos

t tes

t

Set

cle

ar p

erfo

rman

ce

stan

dard

s

Res

pons

ibili

ty

Task

s* = P

rimar

y re

spon

sibi

lity

= S

econ

dary

/team

mem

ber

= N

eed

info

rmat

ion

to/fr

om

*The

se w

ere

the

high

est r

ated

task

s fr

om th

e

Prio

ritiz

atio

n M

atrix

on

the

prev

ious

pag

e.

PS/PI 115-131N 10/7/03, 9:37 AM125


5. Implement the solution or process change.• It is often recommended to try the solution on a

small scale first• Follow the plan and monitor the milestones and

measures

Typical toolsActivity Network Diagram, Flowchart, Gantt Chart,Matrix Diagram, and other project management meth-ods, as well as gathering ongoing data with Run Charts,Check Sheets, Histograms, Process Capability, andControl Charts

SituationThe team used the Responsibility Matrix and GanttChart to guide the training rollout. The original planwas to start in only two shops. There were some initialproblems with employees not being paid for the train-ing time as well as some managers not getting coverageduring training.

DecisionThe solution was to pay the employees for one-half thetraining time and to set up overlapping shifts for bettercoverage for managers. Once these adjustments weremade, the training went so well and had so manyunintended benefits that it was expanded to all sixshops.

Do

PS/PI 115-131N 10/7/03, 9:37 AM126


6. Review and evaluate the result of the change.• Confirm or establish the means of monitoring the

solution. Are the measures valid?• Is the solution having the intended effect? Any

unintended consequences?

Typical toolsCheck Sheet, Control Chart, Flowchart, Pareto Chart,Run Chart

SituationThe training plan was developed in January and Febru-ary and rolled out in March and April. The team wentback to the original Pareto Chart data to compare it tothe current customer complaint data. In addition, theycollected current turnover and delivery time measuresto test the original connection with improved training.

Check

The Pareto Chart on the next page shows the types ofcustomer complaints after the training plan was rolled outin March and April.

Pareto ChartBefore employee training

0

500

1000

1500

2000

2500

La

ted

eliv

-e

rie

s

Tast

e

Wro

ng

o

rde

r

Co

ld

foo

d

Oth

er

75%

9% 8% 5% 3%

Tota

l # o

f Cust

om

er C

om

pla

ints

Total=2510October–December

(across 6 shops)

PS/PI 115-131N 10/7/03, 9:37 AM127


Pareto ChartAfter employee training

DecisionThe team reviewed the original targets, which were setin Step 4.

Turnover: While not at the 30 percent average, ithad decreased for six consecutive months from ahigh of 62 percent to 44 percent last month.Delivery Time: Steadily declined from a high of 40minutes to its most recent level of 28 minutes.Customer Complaints: Overall, complaints werereduced by 52 percent and within “late deliveries”by 63 percentSales Volume: The average volume last month wasat 2250 units, up for the third straight month.

Unexpected resultFor the first time, customers complained about the lackof a good selection on the menu. While Stop ’N Go Pizzawas working toward increasing its speed of pizza deliv-eries and standardizing its processes, the variety of themenu was perhaps too simplified.

Total=1310May–July

(across 6 shops)

Tota

l # o

f Cus

tom

er C

ompl

aint

s

53%

20%14%

5% 5% 3%

0

200

400

600

800

1000

Late

deliv

-er

ies

Sel

ec-

tion

Tast

e

Wro

ng

orde

r

Col

d fo

od

Oth

er

PS/PI 115-131N 10/7/03, 9:37 AM128


Run Charts

Average turnover rate of employees(company-wide)

40

35

30

25

20

15

10

5

0Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul

Ave

rage

Tim

e (in

min

utes

)

28minutes

Average delivery time(company-wide)

80

70

60

50

40

30

20

10

0Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul

% 44%

PS/PI 115-131N 10/7/03, 9:37 AM129


7. Reflect and act on learnings.• Assess the results and problem-solving process

and recommend changes• Continue the improvement process where

needed; standardization where possible• Celebrate success

Typical toolsAffinity Diagram, Brainstorming, ImprovementStoryboard, Radar Chart

SituationSix months after the new training was started, the team metto evaluate its results and process. Team members used aRadar Chart to illustrate their assessment of the team.

DecisionThe Radar Chart showed strong agreement amongteam members on the performance of “Results,” “Useof tools,” and “Impact on customers.” The performanceand consensus among team members were both lowerin “Standardization” and “Teamwork.” When the teampresented its storyboard to top management, the majorresult was a complete overhaul of basic training contentand delivery, as well as the new performance measuresthat would continue to be monitored. The final celebra-tion was . . . what else . . . an all-expense-paid Stop ’N GoPizza party!

Future PossibilitiesMore efficient mapping, routing, and dispatching ofpizza deliveries, as well as more staff cross-training.

Act

PS/PI 115-131N 10/7/03, 9:37 AM130


Radar ChartTeam evaluation of itself after new training

Note: The “x” mark indicates the team’s average performancerating while the shaded area indicates the range of ratingswithin the team.

5

4

3

2

1

Use of Tools

Teamwork

Results

Standardization

Impact on Customers

X

X

X

X

X

PS/PI 115-131N 10/7/03, 9:37 AM131

©1994 GOAL/QPC132 Process Capability

LSL USLProcess CapabilityMeasuring conformanceto customer requirements

Why use it?To determine whether a process, given its natural varia-tion, is capable of meeting established customer re-quirements or specifications.

What does it do?• Helps a team answer the question, “Is the process

capable?”• Helps to determine if there has been a change in

the process• Helps to determine percent of product or service

not meeting customer requirements

How do I do it?1. Determine the process grand average, X, and the

average range, R• Use a stable Control Chart, which means the

process is stable and normally distributed.

2. Determine the Upper Specification Limit (USL)and the Lower Specification Limit (LSL)• The USL and LSL are based on customer require-

ments. Recognize that these specification limitsare based solely on customer requirements and donot reflect the capacity of the process.

3. Calculate the process standard deviation• Process cabability is based on individual points

from a process under study. Information from aControl Chart can be used to estimate the process’

132 Process Capability

Proc Capability 132-136N 10/7/03, 9:36 AM132

©1994 GOAL/QPC Process Capability 133

where R and s are the averages of the subgroupranges and standard deviation, and d2 and c4 arethe associated constant values based on the sub-group sample sizes. See the Table of Constants inControl Charts.

• The process average is estimated simply by X, X,X.

4. Calculate the process capability• To measure the degree to which a process is or is

not capable of meeting customer requirements,capability indices have been developed to com-pare the distribution of your process in relation tothe specification limits.

• A stable process can be represented by a measureof its variation—six standard deviations. Com-paring six standard deviations of the processvariation to the customer specifications providesa measure of capability. Some measures of capabil-ity include Cp and its inverse Cr , Cpl , Cpu, and Cpk.

average and variation (standard deviation, s).• σ is a measure of the process (population) stan-

dard deviation and can be estimated from infor-mation on the Control Chart by

R sd2 c4

σ = or σ =

USL – LSL6 σ

Cp =

Cp (simple process capability)

Tip While Cp relates the spread of the process rela-tive to the specification width, it DOES NOT lookat how well the process average is centered to thetarget value.

˜



The process variation exceeds specification. Defectives are being made.

Cp<1

Cp=1

Cp>1

USLLSL

The process is just meeting specification. A minimum of .3% defectives will be made, more if the process is not centered.

The process variation is less than specification, however, defectives might be made if the process is not centered on the target value.

Cpl , Cpu, and Cpk (process capability indices)

The indices Cpl and Cpu (for single-sidedspecificiation limits) and Cpk (for two-sided speci-fication limits) measure not only the processvariation with respect to the allowable specifica-tion, they also take into account the location of theprocess average. Cpk is considered a measure ofthe process capability and is taken as the smallerof either Cpl or Cpu

X – LSL USL – X3 σ 3 σ

Cpl = Cpu = Cpk = min {Cpl, Cpu}

Tip If the process is near normal and in statisticalcontrol, Cpk can be used to estimate the expectedpercent of defective material. Estimating the per-centages of defective material is beyond the scopeof this book and can be found in statistical books.


©1994 GOAL/QPC Process Capability 135

LSL USLTarget

XCpl Cpu

X – LSL USL – X

Actual Spread, 3 σ Actual Spread, 3 σ

• If the process is not capable, form a team toidentify and correct the common causes of thevariation in the process.– Process capability, based on individual data of

the process population, is used to determine ifa process is capable of meeting customer re-quirements or specifications. It represents a“snapshot” of the process for some specificperiod of time.

– Control Charts use small sample sizes overtime and look at the averages. The controllimits are natural limits of variation of theaverages within the sample. These limits are not tobe confused with specification limits, whichare for individual data points in the population.

VariationsThe construction steps described in this section are basedon process capability of a Variable Data Control Chart.The process capability of an Attribute Data Control Chartis represented by the process averages p, np, c, and u.



Process CapabilityDie Cutting Process

207 208 209 210 211 212 213 214 215

USLLSL

X

Cp = USL – LSL = 213 – 207 = 6 = 1.938 6σ 6(.516) 3.096

Cpl = X – LSL = 212.5 – 207 = 5.5 = 3.553 3σ 3(.516) 1.548

Cpu = USL – X = 213 – 212.5 = 0.5 = 0.323 3σ 3(.516) 1.548

Cpk = min {Cpl, Cpu} = 0.323

Since Cpk<1, defective material is being made.

R 1.2d2 2.326

σ = = = .516

A Control Chart was maintained, producing the following statistics:X = 212.5 R = 1.2 n = 5Spec. = 210 ± 3 USL = 213 LSL = 207


©1994 GOAL/QPC Radar 137

Radar ChartRating organizationperformance

Why use it?To visually show in one graphic the size of the gapsamong a number of both current organization perfor-mance areas and ideal performance areas.

What does it do?• Makes concentrations of strengths and weak-

nesses visible• Clearly displays the important categories of per-

formance• If done well, clearly defines full performance in

each category• Captures the different perceptions of all the team

members about organization performance

How do I do it?1. Assemble the right team/raters

Tip It is critical to get varied perspectives to avoidorganization “blind spots.”

2. Select and define the rating categories• The chart can handle a wide number of categories,

with 5–10 categories as an average.• Brainstorm or bring headers from an Affinity

Diagram to create the categories.• Define both non-performance and full perfor-

mance within each category so ratings are doneconsistently.

Radar 137

Radar Chts 137-140N 10/7/03, 9:38 AM137

©1994 GOAL/QPC138 Radar

3. Construct the chart• Draw a large wheel on a flipchart with as many

spokes as there are rating categories.• Write down each rating category at the end of

each spoke around the perimeter of the wheel.• Mark each spoke on a zero to “n” scale with “0”

at the center equal to “no performance” and thehighest number on the scale at the outer ringequal to “full performance.” Performance can bemeasured either objectively or subjectively.

= Individual rating= Team consensus rating

A

B

C

D

E

F

G

H

4. Rate all performance categoriesa) Individual: Each person rates in silence, using

multicolored markers or adhesive labels directlyon the flipchart.

b) Team: Through consensus or an average of individualscores, get a team rating. Take into account both theclustering and the spread of the individual ratings.

A

B

C

D

E

F

G

H

5

45

646

72

Radar Chts 137-140N 10/7/03, 9:38 AM138

©1994 GOAL/QPC Radar 139

Tip Make the team rating highly visible on thechart. Be sure to differentiate the team ratingsfrom individual ratings on the chart by color ortype of mark.

5. Connect the team ratings for each category andhighlight as needed

Tip A gap score can be added to each category bysubtracting the team rating score from the highestnumber on the rating scale, e.g., on a scale of “10,”a team rating of “4” produces a gap score of “6” incategories B and E.

6. Interpret and use the results• The overall ratings identify gaps within each cat-

egory but not the relative importance of the cat-egories themselves. Work on the biggest gap inthe most critical category.

• Post the resulting Radar Chart in a prominentplace, review progress regularly, and update thechart accordingly. It is a great visual “reportcard.”

A (5)

B (6)

C (5)

D (4)

E (6)

(4) F

(3) G

(8) H5

4

5

646

72

Gap scores are in parentheses.

Radar Chts 137-140N 10/7/03, 9:38 AM139

©1994 GOAL/QPC140 Radar

RadarTQC Review Scores

4

3

2

1 4

3

2

1

4

3

2

1

4

3

2

1

4

3

2

1

Goal

World leader

Improvement Cycle

CustomerFocus

Planning Process

Total Participation

Process Management

Company's goal: to have 80% of all entities (34) achieve an overall score of >3.5.

To compute overall score:Sum of average

scores from each category# of categories

12.525

= 2.5 (maximum is 5)=

Range of ratings within the team

Average

Information provided courtesy of Hewlett Packard

Radar Chts 137-140N 10/7/03, 9:38 AM140

©1994 GOAL/QPC Run 141

Run ChartTracking trends

Why use it?To allow a team to study observed data (a performancemeasure of a process) for trends or patterns over aspecified period of time.

What does it do?• Monitors the performance of one or more pro-

cesses over time to detect trends, shifts, or cycles• Allows a team to compare a performance mea-

sure before and after implementation of a solu-tion to measure its impact

• Focuses attention on truly vital changes in the process• Tracks useful information for predicting trends

How do I do it?1. Decide on the process performance measure

2. Gather data• Generally, collect 20-25 data points to detect

meaningful patterns.

3. Create a graph with a vertical line (y axis) and ahorizontal line (x axis)• On the vertical line (y axis), draw the scale related

to the variable you are measuring.– Arrange the y axis to cover the full range of the

measurements and then some, e.g., 11/2 timesthe range of data.

Run 141

Run Chts 141-144N 10/7/03, 9:38 AM141

©1994 GOAL/QPC142 Run

• On the horizontal line (x axis), draw the time orsequence scale.

4. Plot the data• Look at the data collected. If there are no obvious

trends, calculate the average or arithmetic mean.The average is the sum of the measured valuesdivided by the number of data points. The me-dian value can also be used but the mean is themost frequently used measure of the “centering”of the sample. (See Data Points for more informa-tion on averages.) Draw a horizontal line at theaverage value.

Tip Do not redraw this average line every timenew data is added. Only when there has been asignificant change in the process or prevailingconditions should the average be recalculatedand redrawn, and then only using the data pointsafter the verified change.

(x axis)Time or sequence

(y a

xis)

Mea

sure

men

t

Average

Run Chts 141-144N 10/7/03, 9:38 AM142

©1994 GOAL/QPC Run 143

5. Interpret the Chart• Note the position of the average line. Is it where it

should be relative to a customer need or specifica-tion? Is it where you want it relative to yourbusiness objective?

Tip A danger in using a Run Chart is the tendencyto see every variation in data as being important.The Run Chart should be used to focus on trulyvital changes in the process. Simple tests can beused to look for meaningful trends and patterns.These tests are found in Control Charts in the“Determining if Your Process is Out of Control”section. Remember that for more sophisticateduses, a Control Chart is invaluable since it issimply a Run Chart with statistically-based limits.

Run Chts 141-144N 10/7/03, 9:38 AM143

©1994 GOAL/QPC144 Run

RunAverage Number of Days

for Determining Eligibility for Services

Ave . #

400

350

300

250

200

150

100 50 0

Dec '92

Jan

'93

Feb

Mar

Apr

May

Jun

Jul

Aug

Sep

Oct

Nov

Dec

Jan

'94

Pro

cess

Ave

rage

Info

rmat

ion

prov

ided

cou

rtes

y of

Geo

rgia

Sta

te D

epar

tmen

tof

Hum

an R

esou

rces

, Div

isio

n of

Reh

abili

tatio

n S

ervi

ces

Note: Eligibility requirements changed in May, making itmuch simpler for the department staff to make determina-tions. The trend is statistically significant because there are sixor more consecutive points declining.

Run Chts 141-144N 10/7/03, 9:38 AM144

©1994 GOAL/QPC Scatter 145Scatter 145

Scatter DiagramMeasuring relationshipsbetween variables

Why use it?To study and identify the possible relationship betweenthe changes observed in two different sets of variables.

What does it do?• Supplies the data to confirm a hypothesis that two

variables are related• Provides both a visual and statistical means to test

the strength of a potential relationship• Provides a good follow-up to a Cause & Effect

Diagram to find out if there is more than just aconsensus connection between causes and the effect

How do I do it?1. Collect 50–100 paired samples of data that you

think may be related and construct a data sheet

Course

123•••

40

AverageSession Rating(on a 1–5 scale)

4.23.74.3•••

3.9

AverageExperience of

Training Team (days)

220270270

•••

625

Theory: There is a possible relationship between the number of days of experience the training team has received and the ratings of course sessions.

Scatter 145-149N 10/7/03, 9:39 AM145

©1994 GOAL/QPC146 Scatter

2. Draw the horizontal (x axis) and vertical (y axis)lines of the diagram• The measurement scales generally increase as

you move up the vertical axis and to the right onthe horizontal axis.

4.54.44.34.24.14.03.93.83.73.63.5

150 200 250 300 350 400 450 500 550 600 650

Dependent variable (“effect”)

Independent variable (“cause”)

Looking for relationships,not cause & effect

4.54.44.34.24.14.03.93.83.73.63.5

150 200 250 300 350 400 450 500 550 600 650

Ove

rall

Rat

ing

of th

e S

essi

on

Average Experience of Training Team (in days)

4.0 y axis430 x axis

3. Plot the data on the diagram• If values are repeated, circle that point as many

times as appropriate.

Information provided courtesy of Hamilton Standard

Scatter 145-149N 10/7/03, 9:39 AM146

©1994 GOAL/QPC Scatter 147

4. Interpret the data• There are many levels of analysis that can be

applied to Scatter Diagram data. Any basic statis-tical process control text, like Kaoru Ishikawa’sGuide to Quality Control, describes additional cor-relation tests. It is important to note that all of theexamples in this chapter are based on straight-line correlations. There are a number of non-linear patterns that can be routinely encountered,e.g., y = ex, y = x2). These types of analyses arebeyond the scope of this book.

• The following five illustrations show the variouspatterns and meanings that Scatter Diagrams canhave. The example used is the training sessionassessment previously shown. The patterns havebeen altered for illustrative purposes. Pattern #3is the actual sample.

Tip The Scatter Diagram does not predict cause andeffect relationships. It only shows the strength ofthe relationship between two variables. Thestronger the relationship, the greater the likeli-hood that change in one variable will affectchange in another variable.

Scatter 145-149N 10/7/03, 9:39 AM147

©1994 GOAL/QPC148 Scatter

1. Positive Correlation. An in-crease in y may depend on anincrease in x. Session ratingsare likely to increase as trainerexperience increases.

2. Possible Positive Correla-tion. If x is increased, y mayincrease somewhat. Other vari-ables may be involved in thelevel of rating in addition totrainer experience.

Trainer Experience

4.5

4.0

3.5

150 400 650

Rat

ing

y

x

Trainer Experience

4.5

4.0

3.5

150 400 650

Rat

ing

y

x

Trainer Experience

4.5

4.0

3.5

150 400 650

Rat

ing

y

x

Trainer Experience

4.5

4.0

3.5

150 400 650

Rat

ing

y

x

Trainer Experience

4.5

4.0

3.5

150 400 650

Rat

ing

y

x

3. No Correlation. There is nodemonstrated connection be-tween trainer experience andsession ratings.

4. Possible Negative Correla-tion. As x is increased, y maydecrease somewhat. Othervariables, besides trainer ex-perience, may also be affect-ing ratings.

5. Negative Correlation. A de-crease in y may depend on anincrease in x. Session ratingsare likely to fall as trainer ex-perience increases.

Scatter 145-149N 10/7/03, 9:39 AM148

©1994 GOAL/QPC Scatter 149

ScatterCapacitance vs. Line Width

2.1

2.0

1.9

1.8

1.7

1.6

40 45 50 55 60 65 70

Cap

acita

nce

(pic

ofar

ads)

Line Width/Spacing (in microns)

Note: This Scatter Diagram shows that there is a strong positiverelationship between these two variables in producing micro-electronic circuits. Since capacitance measures a critical perfor-mance of a circuit, anything that affects it positively or nega-tively is also critical. The diagram shows that line width/spacing is something to watch closely, perhaps using a ControlChart or another type of statistical process control (SPC) tool.


Scatter 145-149N 10/7/03, 9:39 AM149

©1994 GOAL/QPC150 Team Guidelines

TeamGuidelinesFrom “me” to “we”

Starting TeamsThe most critical task for any new team is to establish itspurpose, process, and measures of team progress. Oncethe team has developed the following guidelines andcharters specific to its purpose, they should be recordedon a flipchart and posted at each team meeting forreference.

• Develop a Team Behavior Charter– Groundrules. Develop consensus groundrules of

acceptable and unacceptable individual andteam behavior.

– Decision making. Determine whether decisionswill be made by consensus, majority rule, oranarchy! Discuss whether there are, or shouldbe, exceptions to when the team should notfollow its usual process.

– Communication. Recognize the value of listen-ing and constructive feedback, and make theeffort, every day, to communicate constructively!

– Roles and participation. Discuss how the teamwill choose a leader, and generally how theteam process will be led. The individuals andteam must take responsibility to encourageequal participation.

– Values. Acknowledge and accept the uniqueinsight of each member of the team.

• Develop a Purpose Charter– Establish the answer to why the team exists.

150 Team Guidelines

Team Guides 150-155N 10/7/03, 9:40 AM150

©1994 GOAL/QPC Team Guidelines 151

– Bring together the individuals who wouldwork well together as a team. Determinewhether each person has the knowledge, skills,and influence required to participate effec-tively on the team.

– The team should discuss who its customers are.If the team has multiple customers, decidewhich customers have the highest priority, orat least how their needs will be balanced.

• Develop Measures of Team Progress– Discuss and agree on the desired signals, which

the team can assess both objectively and subjec-tively, that will indicate the team is making progress.

– Discuss and agree on the types of measures andoutcomes that will indicate the team hasreached success or failure.

– Estimate the date when the project should becompleted.

Maintaining MomentumMany teams enjoy terrific starts and then soon fizzle.The real challenge is to keep a team focused on itspurpose and not the histories of its members and theirrelationships to one another.

• Agree on the Improvement Model to Use– Standard steps. Use your organization’s stan-

dard step-by-step improvement process orchoose from the many published options. (Seethe Improvement Storyboard in the ProblemSolving/Process Improvement Model sectionfor one such standard process.)

Team Guides 150-155N 10/7/03, 9:40 AM151


– Data. Gather relevant data to analyze the cur-rent situation. Define what you know, andwhat you need to know, but know when tostop. Learn, as a team, to say when your workis good enough to proceed to the next step inthe process.

– Develop a plan. Use your organization’s stan-dard improvement model to provide the over-all structure of a project plan. Estimate timesfor each step and for the overall project. Moni-tor and revise the plans as needed.

• Use Proven Methods Based on Both Data andKnowledge– Data-based methods. Use tools in this booklet,

e.g., Run Chart, Pareto Chart, that reveal pat-terns within data. These tools often take theemotion out of discussions and keep the pro-cess moving.

– Knowledge-based methods. Many of the methodsin this booklet, e.g., Affinity Diagram, Interre-lationship Digraph, help to generate and ana-lyze ideas to reveal the important informationwithin. They help create consensus, which isthe ideal energy source for a team.

• Manage Team Dynamics– Use facilitators. A facilitator is someone who

monitors and helps team members to keeptheir interactions positive and productive. Thisis the stage when a facilitator can help the teamstay focused on its purpose while improving itsworking relationship.

Team Guides 150-155N 10/7/03, 9:40 AM152


– Manage conflict. As teams grow, so do conflicts.This is a natural process as communicationbecomes more open. The entire team can learntechniques for conflict resolution and use thefacilitator as a resource.

– Recognize agreement. Managing agreement isoften as much of an effort as managing dis-agreement. Test for agreement often and writedown the points of agreement as they occur.

– Encourage fair participation. Each team membermust eventually take responsibility for partici-pating consistently in all discussions. Like-wise, the entire team should be constantlyworking to “pull back” the dominant membersand draw out the quieter members.

Ending Teams/ProjectsMost teams and all projects must eventually end. Bothoften end in unsatisfactory ways or don’t “officiallyend” at all. Before ending, the team should review thefollowing checklist:

❐ We checked our results against our originalgoals and customer needs.

❐ We identified any remaining tasks to be done.

❐ We established responsibility for monitoringthe change over time.

❐ We documented and trained people, when nec-essary, in the new process.

Team Guides 150-155N 10/7/03, 9:40 AM153


❐ We communicated the changes to everyoneaffected by them.

❐ We reviewed our own team’s accomplish-ments for areas of improvement.

❐ We celebrated the efforts of the team with alunch, newsletter article, special presentation tothe company, or other expression of celebration.

❐ We feel proud of our contribution and ac-complishments, our new capabilities, and ournewly defined relationships with coworkers.

Conducting Effective MeetingsPreparation:

• Decide on the purpose of the meeting• Develop a meeting plan (who, what, where,

when, how many)• Identify the meeting leader• Prepare and distribute the agenda• Set up the meeting area

Beginning:• Start on time• Introduce the meeting leader• Allow team members to introduce themselves• Ask for a volunteer timekeeper• Ask for a volunteer recorder• Review, change, order the agenda• Establish time limits• Review prior meeting action items

Team Guides 150-155N 10/7/03, 9:40 AM154


Meeting Etiquette:• Raise your hand and be recognized before

speaking• Be brief and to the point• Make your point calmly• Keep an open mind• Listen without bias• Understand what is said• Avoid side conversations• Respect other opinions• Avoid personal agendas• Come prepared to do what’s good for

the company• Have fun

Ending:• Develop action items (who, what, when, how)• Summarize the meeting with the group• Establish the date and time for a follow-up meeting• Evaluate the meeting• End on time• Clean the meeting area

Next Steps:• Prepare and distribute the meeting activity report• Follow up on action items• Go to “Preparation”

Team Guides 150-155N 10/7/03, 9:40 AM155

©1994 GOAL/QPC156 Tree

Tree DiagramMapping the tasksfor implementation

Why use it?To break any broad goal, graphically, into increasinglevels of detailed actions that must or could be done toachieve the stated goals.

What does it do?• Encourages team members to expand their think-

ing when creating solutions. Simultaneously, thistool keeps everyone linked to the overall goalsand subgoals of a task

• Allows all participants, (and reviewers outsidethe team), to check all of the logical links andcompleteness at every level of plan detail

• Moves the planning team from theory to the realworld

• Reveals the real level of complexity involved inthe achievement of any goal, making potentiallyoverwhelming projects manageable, as well asuncovering unknown complexity

How do I do it?1. Choose the Tree Diagram goal statement

Goal: Increase workplace suggestions

156 Tree

Tree 156-164N 10/7/03, 9:40 AM156

©1994 GOAL/QPC Tree 157

• Typical sources:– The root cause/driver identified in an Interre-

lationship Digraph (ID)– An Affinity Diagram with the headers as major

subgoals– Any assignment given to an individual or team

• When used in conjunction with other manage-ment and planning tools, the most typical sourceis the root cause/driver identified in the ID.

Tip Regardless of the source, work hard to cre-ate—through consensus—a clear, action-ori-ented statement.

2. Assemble the right team• The team should consist of action planners with

detailed knowledge of the goal topic. The teamshould take the Tree only to the level of detail thatthe team’s knowledge will allow. Be prepared tohand further details to others.

• 4-6 people is the ideal group size but the TreeDiagram is appropriate for larger groups as longas the ideas are visible and the session is wellfacilitated.

3. Generate the major Tree headings, which are themajor subgoals to pursue• The simplest method for creating the highest, or

first level of detail, is to brainstorm the major taskareas. These are the major “means” by which thegoal statement will be achieved.

Tree 156-164N 10/7/03, 9:40 AM157


• To encourage creativity, it is often helpful to do an“Action Affinity” on the goal statement. Brain-storm action statements and sort into groupings,but spend less time than usual refining the headercards. Use the header cards as the Tree’s first-level subgoals.

Increase workplace suggestions

Goal Means

Create a workable process

Create capability

Measure results

Provide recognition

Tip Use Post-it™ notes to create the levels of detail.Draw lines only when the Tree is finished. Thisallows it to stay flexible until the process is fin-ished. The Tree can be oriented from left to right,right to left, or top down.

Tip Keep the first level of detail broad, and avoidjumping to the lowest level of task. Remember: “Ifyou start with what you already know, you’llend up where you’ve already been.”

Tree 156-164N 10/7/03, 9:40 AM158


4. Break each major heading into greater detail• Working from the goal statement and first-level

detail, placed either to the extreme left, right ortop of the work surface, ask of each first-levelitem:“What needs to be addressed to achieve the goalstatement?”Repeat this question for each successive level ofdetail.

• Stop the breakdown of each level when there areassignable tasks or the team reaches the limit to itsown expertise. Most Trees are broken out to thethird level of detail (not counting the overall goalstatement as a level). However, some subgoals arejust simpler than others and don’t require asmuch breakdown.

Increase workplace

suggestions

Goal

MeansCreate a workable process

Create capability

Create simple input system

Streamline evaluation

Create quick implementation

Provide information

Supply cost data

Provide single-page

documentation

Go online

Make evaluation a key manager responsibility

Do local evaluation

Monitor approved suggestions monthly

Provide a projected implementation date at

time of approval

Why?

How?

Tree 156-164N 10/7/03, 9:40 AM159


5. Review the completed Tree Diagram for logicalflow and completeness• At each level of detail, ask “Is there something

obvious that we have forgotten?”• As the Tree breaks down into greater detail (from

general to specific) ask, “If I want to accomplishthese results, do I really need to do these tasks?”

• As the Tree builds into broader goals (from thespecific to the general) ask, “Will these actionsactually lead to these results?”

• Draw the lines connecting the tasks.Tip The Tree Diagram is a great communication

tool. It can be used to get input from those outsidethe team. The team’s final task is to considerproposed changes, additions or deletions, and tomodify the Tree as appropriate.

VariationsThe Process Decision Program Chart (PDPC) is a valu-able tool for improving implementation through con-tingency planning. The PDPC, based on the Tree Dia-gram, involves a few simple steps.

1. Assemble a team closest to the implementation

2. Determine proposed implementation steps• List 4–10 broad steps and place them in sequence

in the first Tree level.

3. Branch likely problems off each step• Ask “What could go wrong?”

4. Branch possible and reasonable responses offeach likely problem

Tree 156-164N 10/7/03, 9:40 AM160


5. Choose the most effective countermeasures andbuild them into a revised plan

Goal

Step #1 Step #2 Step #3

Likely problems

Reasonable countermeasures

X = Difficult/impossibleO = Selected

X O X O O X

Tree 156-164N 10/7/03, 9:40 AM161


Awarding Unrestricted Financial Aid

Establish total financial aid budget

Decide who receives financial aid†

Decide how financial aid is awarded†

Determine amount for returning

students

Determine amount for new students†

Not sure about retention

Not sure about tuition rate of

increase

Not sure about outside sources

of aid

• Increase enrollment research – existing data• Additional quantitative data• Additional qualitative data

• Seek quicker decision on tuition rate increase• Plan for variety on rates of increase

• Survey students – expected resources• Participate in state & federal financial aid organizations

† Further information exists but is not shown

Information provided courtesy of St. John Fisher College

PDPC (Tree Variation)Awarding Unrestricted Financial Aid

Note: The PDPC surfaced a lack of accurate information as amajor problem. By anticipating this and filling the most criticalinformation gaps, the budget can be more accurate.

Tree 156-164N 10/7/03, 9:41 AM162


TreeImprove

Business Planning Interaction


Improve interactionamong functional

areas represented inthe group, in the

creation andimplementation of an

effective business plan

Developprocedures toassure teameffectiveness

Understandgoals &roles of

other areas

Interactiontechniques

Make groupmeetings more

effective

Provide system tocommunicate

progress

Write & circulatemeeting minutes

Interaction mustoccur with

regularfrequency

Publish & adhereto agenda, with

team input

Distributetracking charts

of teamperformance

Establish methodof communicating

progress orproblems

Require eachfunction toperiodicallyreport status

Show functionalinterdependencies

in plandevelopment

Use consensus-building techniquesin plan development& implementation

Identifyrelationships anddependencies in

project plans

Each functionshows its plan tofulfill overall plan

Participate in jointtraining of planning

methods

Use facilitatorapproach at

meetings

Use appropriatetools to makedecisions andsolve problems

1Seenextpage

Tree 156-164N 10/7/03, 9:41 AM163


TreeImprove

Business Planning Interaction(continued)

Improve interaction

. . . effectivebusiness plan

Develop procedures toassure team effectiveness

Understand goals & roles of

other areas

Understand regional & functional objectives

Improve understanding

of other functional

areas

Management decisions

affecting group must be

communicated ASAP

Define objectives &

role of the group

Each functionalarea to explain/discuss division

objectives

All functional areas must be represented &

contribute

Develop matrix of evaluation & rewards by

functional areas

Have “field trips” to different

functional areas

Interaction techniques

Rotate associates among functions

1from

previous page


Tree 156-164N 10/7/03, 9:41 AM164

IMPROVING THE WAY ORGANIZATIONS RUN

The+Memory+Jogger+II

Documents

Transcript of The+Memory+Jogger+II