Improvement of quality awareness using six sigma methodology for achieving higher cmmi level

International Journal of Advanced Research in Management (IJARM), B.P. Mahesh, Dr. M.S.

Prabhuswamy & Mamatha. M

20

IMPROVEMENT OF QUALITY AWARENESS USING SIX SIGMA

METHODOLOGY FOR ACHIEVING HIGHER CMMI LEVEL

B.P. Mahesh

Assistant Professor, Department of Industrial Engineering and Management

M.S.Ramaiah Institute of Technology, Bangalore-560054, India

[email protected] (+91-9448739040)

Dr. M.S. Prabhuswamy

Professor, Department of Mechanical Engineering

S.J. College of Engineering, Mysore-570006, India


Mamatha. M

Project Manager, FINACLE

Infosys Technologies Limited, Electronics City, Bangalore- 560100, INDIA


ABSTRACT

Globalization and increased competition gives rise to new approaches to

managing Quality and Productivity. New approaches and frame works such as TQM,

Business Process Re-engineering (BPR), Capability Maturity Model (CMM), etc., have

been extensively deployed in organizations. Along with these approaches, in the face

of a complex dynamic environment, the organizational survival hinges on adaptation

and human competence also. Managing the creative and innovative ability of the

human capital would make a difference between success and failure of any

organization. Six Sigma methodologies provide a highly prescriptive cultural

infrastructure and an adaptive framework for obtaining sustainable results in

manufacturing as well as service organizations. In this article, the research scholar

presents the application of Six Sigma framework for achieving a higher CMMI level

through improvement of quality awareness among process users. The pilot

implementation of recommendations of the study showed improved awareness, better

involvement and enhanced commitment from the process users to follow the

standardized processes for achieving the organization’s goal of being a CMMI level 4

assessed organization.

KEYWORDS

Capability Maturity Model Integration; Six Sigma; Quality Function Deployment;

Failure Mode and Effect Analysis; Quality Management System; Critical to Quality.

I J ARM © IAEME

International Journal of Advanced Research in Management (IJARM), Volume 1, Issue 1, June 2010. pp. 20-41 http://www.iaeme.com/ijarm.html



21

1. INTRODUCTION

Six Sigma methodology has been effectively implemented in many

manufacturing and service sectors. But there is a lot of scope for implementing Six

Sigma methodology in the various areas of Information Technology sector. Software

Engineering Institute – Capability Maturity Model Integration (SEI – CMMI) provides

a road map for organizations to achieve excellence in the Information Technology

sector. The present study was undertaken at a multinational Research and Development

center located in Bangalore. The organization is currently SEI – CMM level 3 assessed

and is striving to achieve CMMI (Capability Maturity Model – Integration) level 4

assessment. To achieve CMMI level 4 assessments, all process users must follow

standardized processes as specified in the Quality Management System (QMS) of the

organization. The initial observation by the research scholar revealed that the process

users were not strictly adhering to specified standardized processes, thus causing a

hindrance for the organization to achieve CMMI level 4.

The objective of the study was to increase the awareness, understanding and

perceived importance of QMS amongst the process users. The Six Sigma - DMAIC

(Define, Measure, Analyze, Improve and Control) methodology was applied to meet

the set objective. The various TQM tools and techniques used in the study were

Structured Survey, Process Mapping, Quality Function Deployment (QFD), Pareto

Analysis, Failure Modes and Effects analysis (FMEA) and Regression Analysis.

2. LITERATURE REVIEW

Six Sigma is a statistical concept that measures a process in terms of defects.

Achieving Six Sigma means processes are delivering 3.4 defects per million

opportunities (DPMO). In other words, they are working almost perfectly.

Sigma is a term in statistics that measures standard deviation. In its business

use, it indicates defects in the outputs of a process, and helps us to understand how far

the process deviates from perfection. One sigma represents 691462.5 DPMO, which

translates to a percentage of non-defective outputs of only 30.854%. That’s obviously

really poor performance. If we have processes functioning at a three sigma level, this

means we are allowing 66807.2 errors per million opportunities, or delivering 93.319%

non-defective outputs. That is much better, but we are still wasting money and

disappointing our customers. The central idea of Six Sigma management is that if we

can measure the defects in a process, we can systematically figure out ways to

eliminate them, to approach a quality level of zero defects, which is the ultimate goal

of TQM.

DMAIC refers to a data-driven quality strategy for improving processes, and is

an integral part of the company's Six Sigma Quality Initiative. This methodology can

be applied to the product or process that is in existence. DMAIC is an acronym for five

interconnected phases: Define, Measure, Analyze, Improve, and Control. Each step in

the cyclical DMAIC Process is required to ensure the best possible results (Figure 1).



22

Figure 1 Six Sigma – DMAIC Methodology

The DMAIC Methodology is explained in simple terms as follows.

� Define the Customer, their critical to quality (CTQ) issues, and the core business

process involved.

� Measure the performance of the Core Business Process involved.

� Analyze the data collected and process map to determine root causes of defects and

opportunities for improvement.

� Improve the target process by designing creative solutions to fix and prevent

problems.

� Control the improvements to keep the process on the new course.

Doug Sanders and Cheryl R Hild [1] have stated that process knowledge is very

important in obtaining Six Sigma solutions. Also, the metrics associated need not

always be number of people trained in Six Sigma, or savings in cost, but defects per

unit, sigma level and rolled-throughput yield.

Cherly Hild, Doug Sanders and Tony Copper [2] have opined that to achieve

optimal outcomes in continuous process, non linear and complex relationships among

process factors must be managed. The data from continuous processes are often

plentiful in terms of processing variables and limited with regard to product

characteristics. With continuous processes, the variation in the main product stream

does not necessarily reflect the true level of variation exhibited by the process.

Goh T.N [3] has brought out an intuitive perspective on the fundamental

mechanics of design of experiments (DOE) in a way that would help enlighten a non-

statistician during the course of deployment of DOE related methodologies, regardless

of the context used. He has stated that in most of the experiments involving multifactor

processes, interactions of 3rd

order and higher, often turn out to be insignificant and are

immaterial to subsequent process characterization and optimization.

Piere Bayle et al, [4] designed and optimized the braking subsystem for a new

product. They also stated that focus is placed on the factors that have the strongest

effect on the response, but there is as much information and insight provided about

direction of future work by considering the implications of factors with little or no

effect.

Spencer Graves [5] has used the tool of forecasted Pareto, which combined

Rolled Throughput Yield (RTY) and sales forecast. RTY estimates the probability

whether a product passes through a process defect free or not as recommended by Six

Sigma proponents, because it seems to be a highly correlated scrap rework, warranty

etc. It is relatively easy to compute from data obtainable from many processes.

DEFINE MEASURE ANALYZE IMPROVE

CONTROL



23

Goh T.N [6] has explained, in a non mathematical language, the rationale and

mechanics of DOE as seen in its deployment in Six Sigma. He has stated the

advantages of DOE over process monitoring techniques. He has described about the

shifting emphasis in the deployment of DOE.

Dana Rasis et al [7] distinguished between black belt and green belt Six Sigma

projects on the basis of five criteria. A case study has been discussed presenting the

definition and measure phases of DMAIC method. The authors identified the CTQ and

performed gauge Repeatability and Reproducibility study on each CTQ.

Charles Ribardo and Theodore T Allen [8] have stated that desirability function

do not explicitly account for the combined effect of the mean and dispersion of quality.

The authors have proposed a desirability function that addresses these limitations and

estimates the effective yield. They have used an Arc welding application to illustrate

how the proposed desirability function can yield a substantially higher level of quality.

The proposed desirability function is based on the estimates of yield that is the fraction

of confirming units.

Goh T.N and M Sie [9] have described some alternative techniques for the

monitoring and control of a process that has been successfully implemented. The

techniques are particularly useful to Six Sigma black belts in dealing with high quality

processes. The methodology ensures a smooth transition from a low sigma process

management to maintenance of high sigma performance in the closing phase of a Six

Sigma project.

Rick L. Edgeman and David Bigio [10] have stated that the future Six Sigma

will be integrated with other tools, used in nontraditional sectors, more adapted and

strengthened. One can expect new concepts like lean Six Sigma, best Six Sigma, lean

best Six Sigma, Six Sigma in health care, lean design and macro Six Sigma to be

applied in manufacturing and service industries.

Mohammed Ramzan and Goyal [11] have stated that Six Sigma provides a

systematic, disciplined and quantitative approach to continuous improvement. Through

the application of statistical thinking, it uncovers the relationship between variation and

its effect on waste, operating cost, cycle time, profitability and customer satisfaction.

The scope of Six Sigma encompasses all aspects of the organization that is from

marketing to product and process designing to accounting to after sale service.

3. OBJECTIVE OF THE STUDY

The objective of the study is to measure the current process user’s awareness

about the organization’s QMS and to improve upon the average awareness level from

the existing 55% to around 70%. The increased awareness, understanding and

perceived importance of QMS enable to have more commitment from the process users

to follow the standardized processes and prepare the necessary documents for

achieving the organization’s goal of being a CMMI level 4 assessed organization.



24

4. DMAIC METHODOLOGY ADOPTED IN THE PRESENT STUDY

4.1 DEFINE PHASE The process users of the organization are only 55% aware of the uses/benefits of

the organization's QMS. This lack of awareness among the process users can lead to be

a hurdle for the organization in achieving CMMI Level 4 Assessment as per the set

deadlines. The process users who are well aware about the QMS & its benefits could

commit themselves to follow the standardized processes and prepare the relevant

documents which would result in having instances necessary for achieving the CMMI

Level 4 Assessment for the organization.

The Define Phase consists of Preparation of Project Charter, Collecting the Voice

of Customers (VOC), Identifying the Critical to Quality (CTQs) and Process Mapping.

• Preparation of Project Charter

The study starts with preparation of a document called Project Charter. This

document clarifies what is expected out of the research team. The major elements of

this document deals with the questions like,

� What is the problem for which the study is being carried out?

� What is the goal of the study?

� Why the study is worth doing?

� How the study's goal can be achieved?

� When the study's goal is supposed to be met?

� Who all are involved in the study?

� What are the challenges/risks that are foreseen in the study?

� Problem Statement

Process users are only 55% aware of the uses / benefits of QMS / QI Page as at

the starting of the study and are not fully following the standardized processes (as

available in the organization's QMS) in their projects.

All other issues have been dealt in the project charter in Figure 2.

• Collection of the VOC

The VOC was collected using a survey questionnaire. The customers for this

study are the process users who are the potential users of the organization's QMS. The

questions used for the purpose of collecting what the customers wanted were open

ended. Some of the questions included in the survey were like

� What would you like to have added on the QMS?

� How do you think Quality can be improved in the organization?

These questions were included in the questionnaire as well as were asked

verbally in the form of interviews. A standard template was used to collect all the

requirements and suggestions of the customers.

• Identification of the CTQs

The VOC, which was collected in the Define Phase with the help of the survey,

is used to identify the CTQs related to the process. These CTQs are used to carry out a



25

QFD. The outcome of this application can be used as the suggestions for improving the

process to make the process users at least 70% aware about the organization's QMS.

Goal

To achieve SEI - CMMI level 4 assessment

from the existing SEI - CMM level 3.

Risks

Getting time from the process users for the

survey.

New resources joining the organization, if

surveyed, can give inaccurate results.

Objective

To increase the average awareness level of

Quality / QMS among the process users

from the existing 55% to at least 70%.

Statement of Work

Modifying the process by which the

Process users are made aware of QMS at

the organization.

Value of the study It will ensure increased awareness level

about organization's QMS among the

process users and enable obtaining more

commitment from them to follow the

standardized processes that would result in

having instances necessary for achieving

the CMMI Level 4 Assessment for the

organization.

Methodology

The methodology used for the project is Six

Sigma DMAIC methodology.

Background Knowledge

The training used for making process users

aware of QMS in the organization.

Figure 2 Project Charter

• Process Mapping

The existing process for any process user / employee to be made aware about

the organization's QMS or the Quality related activities is mapped by studying the

system of induction trainings in the organization. This process is clearly depicted in

Figure 3. The shaded boxes on the process flow chart indicate where the improvements

in the process may take place.

4.2 MEASURE PHASE

The measure phase consists of Selecting CTQ characteristics using TQM tools

like QFD, FMEA & Process Mapping, Defining the performance standards and

Measurement system analysis.

• Selecting CTQ characteristics using Quality Function Deployment (QFD)

QFD may be defined as a systematic process used to integrate the customer

requirements with design, development, engineering, manufacturing and service

functions. The CTQs identified in the previous step are used to prepare the first House

of Quality. Figure 4 shows the VOC on the Y-axis and the requirements of the process

for quality awareness on the X-axis.

The Second House of Quality, as shown in the Figure 5 provides us with the

“HOWS” that tells us how the process can be more effective and efficient in making

the process users aware about the organization’s QMS.



26

New Employee

joins the organization

Employee work on his / her respective No

project until the batch size reaches 6

Yes

Project Manager (PM) /Project Leader

(PL) fills up the Templates or just educate

the employee in filling template.

Software Quality Analyst (SQA)/ Project

Quality Analyst (PQA) reviews the

documents, checks whether the processes

are being followed once a week / fortnight

(mostly with PM / PL)

QMS Awareness

among the employees

The "Hows" obtained as the suggestions from the Houses of Quality are as

follows.

a) Training to be more frequent.

b) Instructor to be trained for training.

c) Conducting regular quality quiz to evaluate the process users' quality awareness.

d) Employee scoring below 70% in the quality quiz to be helped by SQA/PQA.

e) Search functionality to be added on the QI page.

f) QTM and QR of each dept. to come up with dept. specific examples.

g) Project knowledge sharing for best practices related to quality to be initiated.

h) Training invitee list to be compared with the Training attendee list.

From the Pareto Charts as shown in the Figures 6 & 7 for the two Houses of

Quality, we can conclude that Frequency of the QMS training, Conducting regular

Quality Quiz and Instructor to be trained for QMS training are the factors that can

largely satisfy the CTQs, and thus result in having higher awareness levels about

Quality / QMS among the process users.

Figure 3 Existing flow process chart of induction process

Is a batch of 5

new employees

waiting for

QMS training?

Employees go through QMS training in

batch of 6. (Induction)



27

Figure 4 First House of quality

� H : High relationship between customer expectation and process requirement.

� M : Medium relationship between customer expectation and process requirement.

� L : Low relationship between customer expectation and process requirement.

Numerical equivalent of these variables are H = 9, M = 3 and L = 1.

Process Requirement

Customer Expectation

Imp

ort

ance

.

Ex

per

ien

ced

em

plo

yee

s R

efre

sher

Qu

alit

y t

rain

ing

for

thei

r dep

t.

Rev

ampin

g o

f Q

I p

age

(tra

inin

g m

ater

ial,

sea

rch

fu

nct

ion

alit

y).

QM

S T

rain

ing

Eff

icie

ncy

.

Dep

artm

ent-

wis

e Q

MS

tra

inin

g.

QM

S T

rain

ing

Att

endee

lis

t.

Dep

t. s

pec

ific

exam

ple

s in

th

e Q

MS

tra

inin

g.

Kn

ow

led

ge

shar

ing

rel

ated

to

qu

alit

y b

y t

he

pro

ject

s.

Dep

artm

ent

wis

e ca

teg

ori

zati

on

of

pro

cess

es o

n t

he

QI

Pag

e.

To

tal

Frequency of QMS Training 5 H L 50

QMS training for everyone 5 M M H 75

Search Functionality on the QI page 5 M H 60

Different links for different departments 4 H L 40

Guidance for the usage of templates 4 L H 40

Relevance of the training topic 4 H L 40

Time lag between joining the org and QMS

training

4 L L 8

Accessibility of QMS training material 2 M L 8

More examples in the QMS training

material

2 L H 20

Total 64 57 56 51 45 38 26 4



28

Figure 5 Second House of Quality

How’s

Process Requirement

Imp

ort

ance

QM

S t

rain

ing

wee

k e

ver

y 2

mo

nth

s.

Co

nd

uct

reg

ula

r q

ual

ity

qu

iz.

Su

pp

ort

fro

m Q

TM

and

QR

o

f th

e d

ept.

Inst

ruct

or

to b

e tr

ain

ed f

or

QM

S t

rain

ing

.

Inv

ite

em

plo

yee

s sc

ori

ng

lo

w i

n q

uiz

fo

r Q

MS

tra

inin

g.

Rew

ard

th

e P

roje

ct T

eam

fo

llo

win

g t

he

bes

t q

ual

ity

pra

ctic

es.

Rew

ard

exp

erie

nce

d P

M /

PL

fo

r tr

ainin

g.

To

tal

Experienced employees-refresher Quality

trainings for their dept.

5 H M 60

Revamping of QI page (training material, search

functionality).

5 M 15

Department-wise QMS training. 4 L L 8

Dept. specific examples in the QMS training. 4 H M 48

Knowledge sharing related to quality by the

projects.

4 H 36

QMS Training Attendee list. 4 H 36

QMS Training Efficiency. 4 M H H L 88

Department-wise categorization of processes on

the QI page.

3 M 9

Total 61 51 49 48 40 36 15



29

1st

House – Pareto

19%17% 16%

15%13%

11%

08%

01%

0

10

20

30

40

50

60

70

80

1 2 3 4 5 6 7Legend

1 : Experienced employee – refresher quality trainings for their department.

2 : Revamping of QI page (training material, search functionality).

3 : QMS Training Efficiency.

4 : Department-wise QMS training.

5 : QMS Training Attendance list.

6 : Department specific examples in the QMS training.

7 : Knowledge sharing related to quality by the projects.

8 : Department-wise categorization of processes on the QI page.

2nd

House - Pareto

21%

18% 16% 15%13%

12%

0

10

20

30

40

50

60

70

80

1 2 3 4 5 6



30

Legend

1 : QMS training week every 2 months.

2 : Conduct regular quality quiz.

3 : Support from QTM and QR of the department.

4 : Instructor to be trained for QMS training.

5 : Employees scoring low in quiz for QMS training.

6 : Reward the Project Team which follows the best quality practices.

7 : Reward experienced PM / PL for training.

• Failure Modes and Effects Analysis (FMEA)

FMEA is a structured approach to identify the ways in which a process can fail

to meet critical customer requirements. In this study, FMEA is performed to identify

the potential failure modes in the Quality / QMS awareness process. The potential

failure effects of these failure modes, the causes for these failures and the controls that

currently exist over the causes are identified. The severity of the effects of the failure is

rated on a scale of 1 to 10, with 1 being the case when the failure has no effect on the

customer requirements and 10 being the case when the failure largely affects the

customer requirements. The probability of occurrence of the causes of these failures is

also on the same scale, with 1 being the case when these causes are unlikely to occur

and 10 being the case when the probability of occurrence of the causes are very high.

The detection certainty of the causes is rated on a scale of 1 to 10, with 1 being the case

when the cause can be easily detectable and 10 being the case when the causes usually

are not detectable. The performed FMEA is shown in the Figure 8.

• Definition of Performance Standards

The operational definition for the study is that process users are expected to be

at least 55% aware about the organization's QMS. Anyone having an awareness level

below 55% is considered as a defect for the current process. The data collection

methodology that was used for this study is survey. This survey was conducted in a

form of questionnaire consisting of QMS-related questions. The data obtained from the

survey was used for calculating the current Sigma level for the awareness level of the

process users about the organization's QMS.

• Measurement System Analysis -Data Collection Plan

The measures used for this study are the scores in the questionnaire. A survey

was conducted in the form of a questionnaire consisting of QMS-related questions.

Each question had four options, out of which only one was correct. Each question

carried different weights, which were arrived at in a discussion with the Quality Team

members. The designing of the questionnaire involved a brainstorming session with the

Quality Team members. The measurement system tool used is MINITABRelease

14.12.0, Statistical software.

International Journal of Advanced Research in Management (IJARM), B.P. Mahesh, Dr. M.S. Prabhuswamy & Mamatha. M

31

Figure 8 FMEA Table

Potential

Failure

Modes

Potential

Failure

Effects

Sev

erit

y

Potential

Causes

Occ

urr

ence

s

Current

Control

Det

ecti

on

RPN

Action

Recommended

Responsibi-

lity

and Target

Date

S O D

RP

N

QMS

induction

training not

happened

No

awareness

about QMS

10 Trainer busy with

other project

1 Stand by trainer 2 20

Trainee not attending 4 None 4 160 Get non-attendee for

next training

HR dept. 10 3 2 60

Frequency of QMS

training very low

8 Training only

when batch size

reaches 6

members

4 320 QMS training week

every 2 months

Quality team 10 3 4 120

Training

not

effective

Lack of

QMS

awareness

among

attendees

9 Poor instructor’s

presentation skills

2 None 6 252 Instructor to be trained

for QMS training


Examples not

included

4 4 9 1 4

Lack of attendee’s

interest for quality

6 None 3 162 Reward highest scorer

in quiz


Topic irrelevant to

the attendees

2 Department wise

trainings

5 90 Training requested by

QR, PM / PL

QRs, QTMs 9 2 3 54

Process

users not

filling the

templates

Lack of

QMS

awareness

among

process

users

9 PM/PL fills all the

templates

8 None 3 216 Initiate project

knowledge sharing for

best practices related to

quality.

SQAs 9 5 3 135

Process

users not

visiting QI

page for

searching

the

processes

or

templates

available in

QMS

Lack of

QMS

awareness

among

process

users

8 QI page structure not

user friendly

7 None 4 224 Add search

functionality to QI page

EPG 8 5 3 120

Too much data 5 None 3 120 Include and elaborate

the QI page during

QMS training

Instructor 8 4 3 96

Poor process users

motivation for quality

8 None 4 256 Conduct regular quality

quiz

SQAs 8 7 2 112



32

Even if one person repeatedly measures the awareness level of process

users using the survey questionnaire, there will be no variation in the result and

even if two or more people evaluates the process users' awareness revel using

this questionnaire, there will be no variation. Thus, the questionnaire used as

the measurement system satisfies the Repeatability and Reproducibility (R&R)

conditions.

The survey is conducted over a number of process users spread through

various departments of the organization. This sample size is to be sufficient

enough as the organization consists of around 150 process users out of which

around 30 are students who are not directly involved in the projects.

4.3 ANALYZE PHASE The Analyze Phase consists of Establishing Process Capability,

Defining the Performance Objectives and Identifying Variation Sources.

• Establishment of Process Capability

The scores obtained by the process users from the survey which was

conducted during the Define phase is plotted (Figure 9). This graph shows

pictorially the score obtained by the process users. The red bars are the defects.

These bars show the process users scoring below the average score, i.e. below

55%.

Figure 10 shows the summary of statistics for the score obtained. The

histogram is shown along with the normal curve fitted to it. The box plot shows

that there are no Outliers. The P-value calculated is 0.038, which is below 0.05

(i.e. 5%). This result signifies that the scores are normally distributed. Thus the

process capability calcu1ations are performed.

The current average awareness level of the process users as per the

survey conducted is found to be only 55%. The defect definition for the process

is decided to be "an employee scoring less than the mean score, i e. less than

55%". Thus, for the current process, the defects in the process are the process

users scoring below 55%.

0

10

20

30

40

50

60

70

80

90

100

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 57 59 61 63 65

Sco

re o

bta

ine

d (

%)

Emp. No.

Score obtained (%) v/s

Figure 9 Plot of score obtained vs. Emp. No.



33

Figure 10 Summary of Statistics for the Quality Awareness Score

The calculations of the process capability of the current process are shown

below.

Total number of process users surveyed (o - opportunities) = 65

Average Score of the process users = 55%

Number of process users on or above the average score (c) = 33

Number of employee below the average score (d -defects)= (o)-(c) = 65-33= 32

Defects per opportunity (dpo) = (d / o) = (32/65) = 0.49230769

Defects per million opportunities (dpmo) = (d/o)*1000000 = 492307.6

For the calculated dpmo, the current Sigma Rating† =1.52σ

Process Capability of the current process = 1.52σ

• Definition of Performance Objectives

The goal of the study can be defined statistically as follows.

“To increase the average awareness level of process users (process target)

from 55% to 70% and the process capability from 1.52σ to 2.1σ”

† = The Sigma Rating is obtained from the standard Sigma and DPMO Conversion Table.

Anderson-Darling normality test

A- Squared 0.79

P- Value 0.038

Mean 55.477

St. Dev. 22.456

Variance 504.253

Skewness -0.05419

Kurtosis -1.13341

N 65

Minimum 13.000

1st Quartile 36.500

Median 56.000 3rd Quartile 76.000

Maximum 95.000

95% Confidence Interval for

Mean

49.913 61.041

95% Confidence Interval for Median

45.121 66.000

95% Confidence Interval for St.

Dev.

19.150 27.152



34

• Identification of Variation Sources

The P–value calculated signifies that the scores obtained are normally

distributed (for 95% confidence level). P-value may be formally defined as the

probability of being wrong if the alternative hypothesis is selected. The P-value

is calculated here by considering the null hypothesis as “the data follows

normal distribution”. Thus, P-value of less than 0.05 indicates that this null

hypothesis is true. The graphs as shown in Figure 11 show the effects of the

critical ‘X’ on the ‘Y’. This ‘Y’ is the Quality / QMS awareness level of the

process users. These are the critical ‘X’s which were obtained as a result of

QFD and FMEA.

The ‘X’s are:

� Frequency of training

� Instructor to be trained for training

� Conducting regular quality quiz

� Happening of Project knowledge sharing

� Search functionality on the QI Page

� Null Hypothesis statement

� The present process is better than the new proposed process.

4.4 IMPROVE PHASE

The Improve Phase consists of Screening the Potential Causes,

Discovering Variable Relationships and Establishing Operating Tolerances.

• Screening the Potential Causes

This step involves determination of the vital few ‘X’s that affect the ‘Y’.

In this study, the screening of the potential causes identified in the Measure and

Analyze Phases, using basic tools like QFD and FMEA, is being done in the

Improve Phase. Five major factors or ‘X’s that affect the Quality Awareness

among the process users of the organization have been identified.

The Main Effects Plot is used when one have multiple factors. The

points in the plot are the means of the Quality / QMS Awareness at various

levels of each factor (i.e ‘X’s). The plot in Figure 11 is used for comparing the

magnitude of effect, various factors have on the Quality / QMS Awareness (i.e

‘Y’). The slope of the lines depicts the effect of the factors on the ‘Y’. The

higher the slope of the line, higher is the effect of the particular ‘X’ on the ‘Y’.

In the Figure 11, it can be clearly seen that the slope of the line for

‘Frequency of Training’ is highest. Thus it can be concluded that the Quality /

QMS Awareness among the process users is largely affected by the ‘Frequency

of Training’. The factor ‘Conducting Quality Quiz’ has the second highest

slope, i.e Quality / QMS Awareness among the process users can also be highly

affected by ‘Conducting Quality Quiz’. The factor ‘Instructor Training’ also

affects the Quality / QMS Awareness among the process users. However,

adding a ‘QI Page-Search’ and ‘Project Knowledge Sharing’ would not affect

the awareness level among the process users as much as the other 3 factors.



35

Table 1 Data for Regression Analysis

Frequency

of Training

Instructor

Training

Regular

Quality

Quiz

Project

Knowledge

Sharing

QI Page

Search

Quality /

QMS

Awareness

1 1 1 1 1 1.00

0 1 1 1 1 0.75

1 0 1 1 1 0.80

1 1 0 1 1 0.79

1 1 1 0 1 0.83

1 1 1 1 0 0.83

0 0 0 0 0 0.00

0 0 1 1 1 0.55

1 0 0 1 1 0.59

1 1 0 0 1 0.62

1 1 1 0 0 0.66

0 1 1 1 0 0.58

0 0 0 1 1 0.34

1 0 0 0 1 0.42

1 1 0 0 0 0.45

0 1 1 0 0 0.41

0 0 1 1 0 0.38

0 0 1 0 1 0.38

1 0 0 1 0 0.42

0 1 0 0 1 0.37

0 1 0 1 0 0.37

1 0 1 0 0 0.46

0 0 0 0 1 0.17

0 0 0 1 0 0.17

0 0 1 0 0 0.21

0 1 0 0 0 0.20

1 0 0 0 0 0.25

Figure 11 Main Effects Plot



36

Interaction plot (data means) for Quality / QMS Awareness

Figure 12 Interaction Plots

• Discovering Variable relationships

The variable relationships were discovered using the main effects plot

and the interaction plots. Interaction plots are useful for judging the presence of

interaction among the factors. Interaction is present when the response at a

factor level depends upon the level(s) of other factors. Parallel lines in an

interactions plot indicate no interaction. The greater the departure of the lines

from the parallel stage, higher the degree of interaction.

Figure 12 shows a matrix of interaction plots for the five factors. It is a

plot of means for each level of a factor with the level of a second factor held

constant. In the full matrix, the transpose of each plot in the upper right is

displayed in the lower left portion of the matrix.

Figure 12 clearly shows that the ‘Frequency of Training’ is not affected

by the factors ‘Conducting Quality Quiz’ and ‘Project Knowledge Sharing’.

However, there is an interaction between the ‘Frequency of Training’ with the

‘Search functionality on the QI Page’ and ‘Instructor’s training’. Similarly it

can be seen that ‘Project Knowledge Sharing’ has an interaction with the

‘Search functionality’ on the ‘QI Page’. From the interaction plots as shown in

Figure 12, the variables or the factors affecting the quality awareness do not

have much effect on each other.

The prioritization of the factors that affect the awareness of

Quality/QMS among the process users as obtained from the Main Effects Plot

is shown in Table 2.



37

Table 2 Prioritization of factors affecting Quality awareness

Factors Priority

Frequency of QMS training

Conducting regular Quality Quiz

QMS training instructor’s presentation skills

Search functionality on QI page

Project knowledge sharing for best practices related to quality

1

2

3

4

5

This prioritization is used for arriving at an equation relating various

factors with the Quality / QMS Awareness among the process users. These

magnitudes of effect that the various factors have on the Quality / QMS

Awareness (i.e. ‘Y’) can be seen in the Main Effects Plot (Figure 11). The

slope of the lines depicts the effect of the factors on the ‘Y’. The higher the

slope of the line, higher is the effect of the particular ‘X’ on the ‘Y’.

Regression Analysis was executed for arriving at the equation. (Table 1)

Transfer Function between ‘Y’ and the vital few ‘X’s is

Where, Y Quality / QMS Awareness among the process users.

X1 Frequency of the QMS training.

X2 Regular Quality Quiz.

X3 Instructor to be trained for QMS training.

X4 Project Knowledge Sharing for best practices related to quality.

X5 Search functionality on the QI page.

• Proposed Process

Based on the results of the steps performed above, the proposed process

of making the employees aware of the organization’s QMS / Quality related

activities, is shown in the Figure 13.

4.5 CONTROL PHASE The Control Phase consists of Definition and Validation of Measurement

System for the 'X's in actual implementation, Determination of Process

Capability (i.e. Short Term Sigma or σST) and Controlling Long Term Sigma

(σLT).

• Definition and Validation of Measurement System for the 'X's' in actual

implementation

The proposed process needs a pilot study. The need for a pilot study is

to better understand the effects of the proposed solution and plan for a

successful full-scale implementation and to lower the risk of failing to meet

improvement goals when the solution is fully implemented. The measures for

the pilot study stage remains the same as were during the Measure Phase, i.e.

scores obtained in the questionnaire. This data collection plan is used to

confirm that the suggested solution meets the improvement goals.

Y = 0.25X1 + 0.21 X2 + 0.20X3 + 0.17X4 + 0.17X5



38

Yes

Yes

No

No

Figure 13 Proposed Process

• Determination of Process Capability

During the first few trials, in any process, the variability is small and

mean is centered at the target. It is called Short Term Sigma (σST). This is the

best the process is capable of. The survey used for measuring the Quality

Awareness levels of the process users again after implementing the suggested

improvements is the data for calculating the process capability of the new

process.

New Employee

Joins the

organization

Employee to undergo QMS

induction training, which will

happen bi-monthly and as per

need-basis

Is the score of the

employee above

70% in the quiz

conducted with

the QMS training?

The employee’s name is noted in the

invitee list of the next QMS training /

special attention to be given by the

SQA / PQA in the project he / she is

working.

Instructor is trained for

QMS training

Mention about URL

for QI Page and EPG

especially

Department specific

examples are included in

consultation with the

experienced PMs / PLs and

QR.

Employee continues to

work on his / her

project and prepare

necessary documents

Is the employee

scoring > 70% in

the regular

quality quiz (by

SQA / PQA)?



39

The defect definition for the process is modified as "employee scoring

less than the mean score, i. e. less than 70%". This change in defect definition

is due to the goal of this study, which aims at having an average score of 70%

in the questionnaire used for survey. Thus, the number of process users scoring

below 70% is the number of defects for the new process and the number of

process users being surveyed is the number of opportunities. Every possibility

of making an error is called an opportunity and in this process, an opportunity

is an employee who is being surveyed.

The number of defects and the number of opportunities are used to

calculate defects per million opportunities (dpmo). The process capability (σST)

of the new process is obtained using the "Sigma and DPMO Conversion Table"

corresponding to the calculated dpmo. If this sigma rating is around 2.1σ, the

new process is successful. The new process is then to be documented and

followed.

• Controlling the Long Term Sigma (σLT)

Over a period of time, assignable causes creep in and the capability of

the process to meet the requirements diminishes. This sigma which represents

the capability of the process to meet the requirements over a period of time

considering those extraneous conditions causes process shifts from that at

which it was set is called the Long Term Sigma. Normally, the short term

sigma is higher than long term sigma. Unless otherwise specified, long term

sigma is calculated as σLT = σST – 1.5.

There are various mechanisms that can be used to control a process

namely, Risk Management, Mistake Proofing, Statistical Process Control

(SPC) and Control Plans.

The key to controlling the process is frequent interval monitoring. The

ongoing measurements of the process variation and/or process capability are to

be used for monitoring. The ongoing measurements in this study are the regular

quality quizzes that need to be conducted by the Quality Team. Even random

auditing of the documents prepared by the process users for their projects can

give an idea of how much the process users are aware of the organization's

QMS. The responses obtained by these measurement systems indicate the

success of the new process.

5. SOLUTIONS FOR IMPROVING QUALITY AWARENESS The first four phases -Define, Measure, Analyze, and Improve -of the

DMAIC methodology have been applied successfully to this study. The

improvements suggested were planned for implementation, which essentially

forms the Control Phase. Rigorous efforts were made to get the required

approvals from the top management and co-operation from the process users

themselves to improve the Quality Awareness levels in the organization.

Some of the improvements suggested were

• To have QMS trainings every 2 months or on the need basis.

• To conduct regular Quality Quiz for all the process users of the

organization.

• To train the instructor who conducts QMS training.



40

• To add a search functionality on the QI Page on the organization's intranet.

• To initiate regular project knowledge sharing sessions by the SQAs/PQAs

highlighting the best practices related to quality.

• To involve QRs and experienced PMs/PLs of all the departments to suggest

good examples that can be included in the QMS training material.

• To involve experienced PMs/PLs to conduct refresher QMS/Quality-

related trainings for their departments.

• To welcome constructive comments, so that the Quality Awareness process

can be improved continuously.

6. POST IMPLEMENTATION RESULTS

In a span of three months, all solutions recommended were

implemented. Then, the research scholar repeated the Measure and Analyze

phases. The scores obtained by the process users in the post implementation

study are plotted (Figure 14). The red bars are the defects. These bars show the

process users scoring below the average score, i.e. below 70%.

In the improved process, for 17 defects out of 65 opportunities, the

dpmo is found out to be 261538. i.e. the sigma rating or the process capability

of the improved process is found to be 2.13σ.

0

10

20

30

40

50

60

70

80

90

100

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 57 59 61 63 65

Sco

re o

bta

ine

d (

%)

Emp. No.

Score obtained (%) v/s Emp.No.

Figure 14 Plot of score obtained vs. Emp. No.

7. CONCLUSION All the phases - Define, Measure, Analyze, Improve and Control - of the

DMAIC methodology have been successfully applied to the study. The

solutions implemented resulted in increasing the awareness level of the process

user’s form 55% to 70% and increasing the sigma level from 1.52σ to 2.13σ

about the organization's QMS. Similarly, efforts can be put for achieving

higher and higher level of Sigma, until the organization reaches Six Sigma

level.



41

REFERENCES 1. Doug Sanders and Cheryl R Hild (2000-01), “Common Myths about Six

Sigma”, Quality Engineering, Vol 13, No 2, pp 269-276.

2. Cheryl Hild, Doug Sanders and Tony Cooper (2000-02), “Six Sigma on

continuous processes: How & why it differs?” Quality Engineering, Vol

13, No1, pp 1-9.

3. Goh T.N (2001), “Information Transformation Perspective on

Experimental Design in Six Sigma”, Quality Engineering, Vol 13, No 3, pp

349-355.

4. Piere Bayle, Mike Farrington, Brenner Sharp, Cheryl Hild & Doug Sanders

(2001), “Illustration of Six Sigma Assistance on a Design Project”, Quality

Engineering, Vol 13, No 3, pp 341-348.

5. Spencer Graves (2001-02), “Six Sigma Rolled Throughput Yield”, Quality

Engineering, Vol 14, No. 2, pp 257-266.

6. Goh T.N (2002), “The role of Statistical Design of Experiments in Six

Sigma: Perspectives of a Practitioner”, Quality Engineering, Vol 14, No 4,

pp 659 – 671.

7. Dana Rasis, Howard. S. Gitlow & Edward Popouich (2002-2003), “A

fictitious Six Sigma Green Belt, case study”, Quality Engg; Vol 15, No 1,

127-145.

8. Charles Ribardo and Theodore T Allen (2003), “An Alternative

Desirability Function for achieving Six Sigma Quality”, Quality and

Reliability Engineering International, Vol 19, pp 227-240.

9. Goh T N and M Sie (2003), “Statistical control of a Six Sigma Process”,

Quality Engineering, Vol 15, No 4, pp 587-592.

10. Rick L. Edgeman & David Bigio (Jan 2004), “Six Sigma in Metaphor:

Heresy or Holy Writ?” Quality Progress, pp 25 -31.

11. Mohammed Ramzan and Goyal (Jan 2006), “Six Sigma: An introduction

for Industrial Engineers”, IIIE Journal, Vol 35, No. 1, pp 13-15.

12. Peter S. Pande & Larry Holpp (2001), “What is Six Sigma?” Tata Mc Graw

Hill Company Limited 1st edition.

13. Peter S. Pande, Robert P. Neuman, Roland R. Cavanagh (2000), “The Six

Sigma Way: How GE, Motorola, and Other Top Companies are Honing

Their Performance”, McGraw- Hill Companies.

14. Greg Brue (2002), “Six Sigma for Managers” Tata McGraw-Hill.

Improvement of quality awareness using six sigma methodology for achieving higher cmmi level

Documents

sigma level

sigma management

quality level

sigma quality initiative

process users

sigma framework

sigma methodologies

sigma dmaicdefine