Optimum Software Process Improvement Paradigm for Quality Practices in Software Industry Submitted by Faisal Tehseen Shah in accordance with the requirement for the degree of Doctor of Philosophy (August 2010) Supervisor: Dr. Niaz Ahmad Co-Supervisor: Dr. Shafay Shamail Institute of Quality and Technology Management Faculty of Engineering and Technology University of the Punjab Quaid-e-Azam Campus, Lahore-Pakistan
172
Embed
Optimum Software Process Improvement Paradigm for Quality ...prr.hec.gov.pk/jspui/bitstream/123456789/1535/1/1239S.pdf · Improvement Paradigm for Quality Practices in Software Industry
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Optimum Software Process Improvement Paradigm for Quality
Practices in Software Industry
Submitted by Faisal Tehseen Shah
in accordance with the requirement for the degree of
Doctor of Philosophy
(August 2010)
Supervisor: Dr. Niaz Ahmad
Co-Supervisor: Dr. Shafay Shamail
Institute of Quality and Technology Management Faculty of Engineering and Technology
University of the Punjab Quaid-e-Azam Campus, Lahore-Pakistan
CERTIFICATE
This is to certify that the dissertation is the original work of the author
and has been carried out under our supervision. We certify that the material
included in this thesis have not been used in part or full in a manuscript
already submitted or in the process of submission in partial / complete
fulfilment of the award of any other degree from University of the Punjab or
any other institution. We also certify that the thesis has been prepared
according to the prescribed format of University of the Punjab and we submit
for its evaluation for the award of Ph.D. degree through the official procedures
of the University of the Punjab.
Prof. Dr. Niaz Ahmad (Supervisor) Ex-Director
Institute of Quality and Technology Management University of the Punjab
Dr. Shafay Shamail (Co-Supervisor)
Department of Computer Science Lahore University of Management Sciences
Abstract Overall behaviour of local software industry towards quality if simply
phrased is, “No Quality Culture”. It is the cause of lower Information
Technology (IT) exports due to non competitive nature of local software
product development practices which are laden with delays, non-
conformances and inconsistency. Local quality culture lacks quality
awareness and is immature in following good quality practices and
implementing quality improvement standards. As a step further in this
direction the objective of this study is to map the actual environment and true
culture of Small and Medium Software Houses (SMSH) towards quality
improvement and process improvement by implementing Total Quality
Management (TQM) philosophy.
It was an exploratory research effort in the domain of Total Quality
Management (TQM) and Software Process Improvement (SPI). The research
begins with literature review of major quality standards implemented in the
local industry. The behaviour of international quality standards was
deliberated towards SMSH. A survey was conducted to evaluate the current
quality practices and develop a process improvement model within the local
SMSH. For this purpose software houses that were members of statutory and
professional organizations such as Pakistan Software Export Board (PSEB),
Pakistan Software House Association (PASHA) were selected. Listing of
commercially available directory of RozeePak was also referred. For this
survey the quality constructs and data collection instrument were designed
based on literature review about small and medium enterprises culture and
leading software quality models such as CMM, CMMI, ISO, SPICE and
PSP/TSP. The results of the survey were analyzed and reported to high light
quality problems being faced by SMSH to implement quality.
Study included descriptive as well as empirical analysis. Descriptive
analysis was based on comments via survey and personal interaction while
conducting the survey. The empirical analysis included correlation and
regression analysis of quality constructs. Structural Equation Modelling (SEM)
technique was used to develop an optimized Lean Quality Improvement
Model (LQIM) for standard quality practices in the local software industry.
Eight quality constructs were developed to ascertain the level of current
quality practices in the SMSH and evolve a LQIM. In correlation analysis all
seven independent constructs were found significant towards the dependent
variable Quality Improvement. Regression analysis revealed that only four of
these independent quality constructs contributed significantly towards the
dependant variable Quality Improvement. Through Structural Equation
Modelling (SEM) the LQIM was evolved. This model presented four quality
constructs and ten of their respective quality practices as significant.
LQIM evolved as a tailored and economized paradigm according to
the needs and perceptions of the local IT practitioners. Also LQIM evolved as
an indigenous model which when improvised in accordance to the SMSH
cultural and quality improvement recommendations is proven to be a fit model
for SMSH. The LQIM has already been ratified according to generally
accepted good fit indices in SEM analysis. In order to implement LQIM by
SMSH implementation of Indigenous LQIM was proposed using the Deming’s
philosophy of Plan, Do, Check, Act (PDCA) Cycle for continuous process
improvement. The set of recommendations for SMSH software process
improvement and proposed LQIM paradigm will give the innovative and
flexible directions for SMSH to change their culture and improve their
processes and software quality.
ACKNOWLEDGEMENTS
I am thankful to
Almighty ALLAH
who has given me the zeal to learn and seek knowledge.
my profound gratitude and regards to my visionary Supervisor for his support
Professor Dr. Niaz Ahmed.
and
my deep appreciation for my Co-Supervisor Dr Shafay Shamail
for his resolute commitment, deep insight and direction during the phase of research and especially his patience to match my pace of learning.
and
I am thankful to my class fellows for their priceless contribution in knowledge sharing especially Dr. Muhammad Usman Awan and Dr. Tajamal Hussain
and
my gratitude towards my colleagues for unprecedented support and motivation especially
Mansoor Shiraz and Muhammad Irfan
and
my thanks to young students who assisted me in conducting initial research during pilot studies especially
M. Shafique Khan (Late), and Faiza Dar
and
most importantly thanks to my wife and kids for their patience and sacrifice of their time that I consumed for this research.
CONTENTS LIST OF ABBREVIATIONS ................................................................................................................................ i
LIST OF FIGURES ................................................................................................................................................ iv
LIST OF TABLES .................................................................................................................................................... v
APPENDIX A – COVER LETTER .............................................................................................................. 145
Optimum Software Process Improvement Paradigm for Quality Practices in Software Industry ......................................................................................................................................... 145
APPENDIX B – QUESTIONNAIRE ........................................................................................................... 147
APPENDIX C QUESTIONNAIRE INDICATORS .............................................................................. 151
APPENDIX D INDICATORS & MAPPING ISO 9000.................................................................... 157
i
LIST OF ABBREVIATIONS
ACM Association for Computing Machinery
AMOS Analytical Movement of Structures
ANOVA Analysis of variance
BPR Business Process Re-Engineering
BSI British Standard Institute
CFA Confirmatory Factor Analysis
CFI Comparative fit index
CMM Capability Maturity Model
CMMI Capability Maturity Model Integration
CPI Continuous Process Improvement
CPI Capability process Index
CPI Continuous Process Improvement
CRM Customer Relationship Management
CUS Customer Supplier
DCS Data Collection System
FFRDC Federally Funded Research and Development Centre
GFI Goodness-0f-fit Index
GQM Goal Question Matrix
HQC High Quality software creation support virtual Centre
ICT Information & Communication Technology
IEEE Institute of Electrical and Electronics Engineers
ISO International Standard Organization
KBL Knowledge Base Library
KM Knowledge Management
KPAs Key Process Areas
LQIM Lean Quality Improvement Model
MAN Management
MAN Measurement & Analysis
NFI Normed Fit Index
OBQ Organization Behaviour Towards Quality
OBQ Behaviour Towards Quality
OCL Organizational Culture
ORG Organization
ii
OSS Organization Size & Structure
PA Process Attributes
PASHA Pakistan Software House Association
PDP Project development plan
PITB Punjab Information Technology Board
PMC Project Monitoring Tracking
PMC Monitoring and Control
PPL Project Planning
PQI Process / Quality Improvement
PQM Product Quality Management
PSEB Pakistan Software Export Board
PSP Personal Process Software
QA Quality Assurance
QC Quality Control
QFD Quality Function Deployment
QMPs Quality Management Principles
QMS Quality Management System
QoC Quality of Conformance
QoD Quality of Design
QoP Quality of Performance
RDM Requirement Development Management
RMSEA Root Mean Square Error Index
ROI Return on Investment
SCAMPI Standard CMMI Appraisal Method for Process Improvement
SCM Software Configuration Management
SDLC Software Development Life Cycle
SEI Software Engineering Institute
SEM Structural Equation Modelling
SME Small and Medium Enterprises
SMEs Small and Medium Enterprises
SMSHs Small and Medium Software Houses
SPI Software Process Improvement
SPICE Software Process Improvement Determination
SQM Software Quality Management
SUP Support
SW-CMM Software Capability Maturity Model
iii
TQC Total quality control
TQM Total Quality Management
TSP Team Software Process
VSEs Very Small Enterprises
WHO World Health Organization
iv
LIST OF FIGURES
FIGURE 1 STRUCTURE OF THE THESIS........................................................................................... 14 FIGURE 2 EVOLUTION QUAGMIRE OF QUALITY MODELS ..................................................... 34 FIGURE 3 RELATIONSHIP BETWEEN KEY SC7 STANDARDS ............................................... 37 FIGURE 4 THEORETICAL FRAMEWORK ......................................................................................... 80 FIGURE 5 THEORATICAL STRUCTURAL MODELING ............................................................... 84 FIGURE 6 QUALITY IMPROVEMENT DEPENDENCY MODEL ............................................. 108 FIGURE 7 SEM STANDARDIZED SOLUTION FOR SPI MODEL FIT. .................................. 113 FIGURE 8 IMPLEMENTATION OF LQIM MODEL ...................................................................... 127
v
LIST OF TABLES
TABLE 1 Structure of Quality Models .................................................................................................. 36 TABLE 2 OVERVIEW OF SPICE CAPABILITY LEVELS .................................................................. 57 TABLE 3 PSP Process Hierarchy ............................................................................................................ 61 TABLE 4 Structure of TSP ........................................................................................................................ 63 TABLE 5 CONSTRUCTS TABLE ............................................................................................................... 78 TABLE 6 ORGANIZATION SIZE & STRUCTURE ............................................................................... 87 TABLE 7 ORGANIZATION CULTURE .................................................................................................... 88 TABLE 8 ORGANIZATION BEHAVIOUR TOWARDS QUALITY .................................................. 89 TABLE 9 REQUIREMENT DEVELOPMENT & MANAGEMENT ................................................. 90 TABLE 10 PROJECT PLANNING............................................................................................................ 91 TABLE 11 PROJECT MONITORING TOOL ........................................................................................ 92 TABLE 12 MEASUREMENT & ANALYSIS ......................................................................................... 93 TABLE 13 PROCESS QUALITY IMPROVEMENT ............................................................................ 94 TABLE 14 QUALITY MODEL DEMOGRAPHICS .............................................................................. 95 TABLE 15 RESPONDENT’S PROFILE GROUPS .............................................................................. 96 TABLE 16 RELIABILITY OF CONSTRUCTS ................................................................................... 103 TABLE 17 RELIABILITY STATISTICS .............................................................................................. 103 TABLE 18 CORRELATION BETWEEN ALL CONSTRUCTS ..................................................... 105 TABLE 19 MODEL SUMMARY ............................................................................................................ 106 TABLE 20 ANOVA .................................................................................................................................... 107 TABLE 21 COEFFICIENTS .................................................................................................................... 107 TABLE 22 CMIN CHI‐SQUARE ............................................................................................................ 110 TABLE 23 BASELINE COMPARISONS MODEL FIT INDICES ................................................ 111 TABLE 24 RMSEA ..................................................................................................................................... 112 TABLE 25 EVOLVED SPI PARADIGM PRACTICES ..................................................................... 114 TABLE 26 SEM DELETED ITEMS FROM MODEL ....................................................................... 114 TABLE 27 LQIM CONCEPTUAL DETAIL ........................................................................................ 116 TABLE 28 LQIM DEPLOYMENT PLAN MAPPED WITH PDCA ............................................. 129
Chapter 1 Introduction
1
CHAPTER 1- INTRODUCTION
This Chapter starts with introduction and background to the software
quality improvement in the Pakistan’s software industry, and highlights the
efforts made by Government of Pakistan to promote Information Technology
(IT) industry. After that research questions are mentioned to support the
research objectives of the study. Following that role of government and IT
statuary bodies in developing IT sector of Pakistan is discussed. In the end
significance and structure of research are referred to develop further
understanding into this research.
After studying quality gurus like Joseph M. Juran who is considered to
be the pioneering authority in Quality Management1; Dr. Edward Deming who
is considered by many as father of quality2 and Crosby known for his book,
“Quality without tears” and his overall philosophy, “Quality is through
prevention and conformance to customer’s requirements only” (Sharon and
Shyrel, 1998). It was finally learnt that Quality was all about customer
satisfaction on one side and mindset change, prevention and culture change
on the other side, which is all cloaked in the phrase ” Total Quality
management” (TQM). As concept of quality claimed by Juran (1988) and
Crosby (1979 ) that “Quality is free”, is indeed the concept that the local
industry failed to apprehend where savings in rework cost are much more
than the amount invested in prevention costs. On the contrary trend showed
that companies only practiced quality when it is affordable or convenient to
the top management. It is realized that very little work has been done in the
area of TQM implementation in the local software industry, and also in the
third world developing countries. The study is to investigate local (Pakistani)
software industry quality practices in the Small and Medium Software Houses
(SMSH) in the light of TQM philosophy and benchmarking the world best
1 “Quality is planned; Product quality does not happen by accident; Quality product is fitness for use and free from deficiencies. (M. Juran, 1988)
2 Deming’s teachings mainly revolve around PDCA Cycle and Deming’s 14 principles that bought industrial revolution in Japan and Japan became the market leader in 1950s. Deming preached statistical quality control and emphasized that quality is management’s responsibility. (E. Deming, 1986) as cited by (Sharon, Sheryl, 1998)
Chapter 1 Introduction
2
quality practices for quality and process improvement. This chapter introduces
the problem, objectives and purpose of carrying out the research and benefits
and significance of the research. At the end of the chapter research
framework is explained with the help of a diagram.
1.1. PURPOSE OF THE RESEARCH
The software industry has become the backbone of a country’s
economic growth and prosperity and due to its nature of dealing, networking
and one to one linkage directly with foreign companies; it opens a virtual
corridor to develop industrial liaison and business linkages in the global
market. This research area deals with past and present state of software
industry in Pakistan and behaviour of software industries towards software
quality, Total Quality Management (TQM) and Continuous Process
Improvement (CPI) and above all Quality Improvement.
The main purpose of this study is to determine the level of quality
practices understood and implemented by the practitioners in Pakistan’s
software industry. The objective is to find out whether bear minimum quality
standards are being practiced. It does not matter which international standard
is being adopted by an organization, but it is important to find out if the quality
practices and processes are practically followed. If an organization utilizes its
stated standard spiritually then its product quality and product effectiveness
will match with that of software products produced at international level. The
cause of lower export rate of Pakistan software industry is the lack of
awareness with quality standards. The government of Pakistan is subsidizing
Information Technology companies to get certifications for ISO 9001 and
Capability Maturity Model Integration (CMMI) as reported on Pakistan
Software Export Board (PSEB) official website (PSEB, 2010). As a step
further in this direction the objective is to map the actual environment and true
culture of Small and Medium Enterprises (SME) towards quality improvement,
process improvement, and Continuous Process Improvement (CPI). It is an
exploratory research effort in the domain of Total Quality Management (TQM)
and SPI. A statistical industrial survey is conducted among the houses of the
local Industry. Mainly software houses which are members of Pakistan
Chapter 1 Introduction
3
Software Export Board (PSEB), Pakistan Software House Association
(PASHA) and SMSH in the major cities of Pakistan are targeted as population
sample for this study. A feedback from this survey will give a set of concrete
discrepancies between true SME culture and required culture of international
standards like CMM, CMMI, Software Process Improvement Determination
(SPICE) and ISO.
After identifying characteristics of a true SME culture, a set of guidelines
and a process improvement paradigm for SME is to be developed, which is
the basic purpose of this research. The set of guidelines for SME software
process improvement paradigm will give the innovative and flexible directions
for SMSH to change their culture and improve their processes and quality.
The aim is that organizations of all sizes especially of small size are able to
implement it for the improvement of their product and process quality.
Guidelines to change mindset of the employees and top management and
hence apply TQM philosophy to implement quality improvement and
measurement culture are proposed. Such guidelines will enable small and
medium sized software houses to build optimum quality culture and maintain
a beer minimum level of quality that will lead SMSH to become competitive,
as well as quality organizations through continuous process improvement. In
order to fully comprehend the aims and objectives of this research it will be
important to mention the research questions developed on the basis of
situation analysis. This situation analysis is based on critical analysis and
comprehensive literature review presented in Chapter-2 and Chapter3. In the
following section research questions are given that evolved during the
research design and questionnaire development phase as an output of pilot
study.
Chapter 1 Introduction
4
1.2. RESEARCH QUESTIONS
The research questions that evolved after the situation analysis and
pilot study to further highlight and support the research objective are the
following.
1. What is quality and concept of quality culture and process
improvement in the Small and Medium Software Houses (SMSH)?
Small and Medium Enterprises (SME) culture can be referred to as
behaviour of immaturity of SME towards software development that results in
threats for SME performance. SME culture concept becomes a vital issue in
the Performance of local markets especially in context of quality culture. A
detailed discussion is in Chapter-2.
2. What are the different types of leading models of Software
Process Improvement (SPI) being practiced world wide as best
practices to improve software quality?
These are the different types of leading software process models and
process management models being practiced locally and globally as best
practices in software quality improvement. These models comprise of ISO
9001, CMM, CMMI, SPICE, PSP and TSP. Detail discussion about these
models is given in Chapter-3.
3. What are the problems and issues faced by local IT practitioners
to implement quality for Software Process Improvement (SPI)?
There are many issues and constraints that local SMEs are facing in
the local industry. The attitude of the top mangers is more towards producing
bulk of code and making money and less towards solving quality issues at
work place. Some of the problems faced by the practitioners as found in the
survey and literature review are given in Chapter-2 and Chapter-5
4. What can be a proposed Software Process Improvement (SPI)
paradigm which can best fit to solve the problems of quality
Chapter 1 Introduction
5
improvement in the local Small and Medium Software houses
(SMSH)?
The proposed Optimum Software Quality Improvement Model and a
set of guidelines for software quality improvement for Small and Medium Size
Software Houses (SMSH) are reported in the chapter-7. These guidelines
are derived from the Results of descriptive analysis chapter – 5 and proposed
SPI model through Structured Equation Modelling (SEM) in chapter 6.
1.3. BACKGROUND OF PAKISTAN SOFTWARE INDUSTRY
When a brief look is taken on the Pakistan’s software/IT industry, it is
observed that local software industry has shown a very uneven pattern of
growth through its very short history. Before early nineties the Government of
Pakistan (GOP) showed a cold stance toward the Information Technology and
software industry. The IT industry was not in the GOP priorities though
software houses have existed in the country since 1970’s. From early-to-mid
1990’s, it has been stated that the IT industry is promoted and supported by
the government. It started getting attention when the software industry of the
developing countries started to groom and rose to prominence. Since then
several policy actions and infrastructure development and up-gradation
projects have been undertaken by the GOP to promote not only the local
software industry but also to export the software from Pakistan. Many national
IT policies along with their action plans are documented according to the
Ministry of Science and Technology (MoST). Our local software industry
needs a face lift and deliberation in policy making as our local software
houses do not show the kind of vitality and growth as that is expected by
contemporary software houses of international repute. To export the quality
software where there is a strong competition across the globe. IN order to
improve Pakistan has to be aware of the importance of development of local
IT industry (Osama, 2005).
There are many reasons for the poor quality of IT business in the
country. For instance there is a brain drain hence Pakistan looses to take
advantage of those qualified professionals. There is a lack of entrepreneurial
skills and managerial know-how of IT professionals. IT industry faces serious
Chapter 1 Introduction
6
problems in securing financing and credit. Unwillingness of local business
managers to pay appropriate prices for locally developed software (Salim,
2001) is another cause of poor growth of local software market. Beside all the
problems and threats faced there still are many opportunities in grooming the
software industry of Pakistan. PSEB is willing to provide all possible help to
facilitate setting up of Call Centres. The global and domestic Internet
explosion helps a lot in worldwide growth not only in communication
infrastructure but also in distance learning and education. Also the customer
awareness and empowerment has increased through IT and Internet. PSEB
and Pakistan National Accreditation Council (PNAC) have extensively trained
and developed personal of software houses through their quality improvement
trainings and certification trainings, but a lot more effort in this direction is
desirable.
1.4. BUILDING BLOCKS OF PAKISTAN SOFTWARE INDUSTRY
The building blocks of IT industry in Pakistan are Pakistan Council for
Science & Technology (PCST), National Commission for Science &
Technology (NCST), Pakistan Software Export Board (PSEB), and Pakistan
Software Houses Association (PASHA), such organizations are the major hub
of activity and are taking steps to improve local software industry to make it a
candidate in the global software industry.
PASHA is an association to promote the software industry in Pakistan
and to protect the rights of its members. There were nine software houses
that formed PASHA in the last quarter of 1992. By 2007 it has grown to a
membership of over 350 software houses. The efforts of PASHA have
resulted in the formation of IT Policy and Action Plan of the government. The
guiding theme for the IT Policy is that “the government shall be the facilitator
and enabler to encourage the private sector to drive the development in IT
and telecommunications”. The government IT Policy covers the development
of human resource, IT infrastructure, software and hardware development
(PASHA, 2010).
Pakistan local software industry growth has been sluggish due to slow
development of supportive policies towards IT and local software houses in
Chapter 1 Introduction
7
last 3 decades by the GOP. For most of the 1990s, Government's policies
towards the IT Industry were a little misaligned and did not match the direction
and needs of local software houses. The hype of the IT global bubble also
affected Pakistan like other developing countries and it was believed that "all
you need is a computer and an Internet connection" to join the race (Osama,
2005). Similarly capacity building and creation of new jobs was also slow. It
was not until 2005 that GOP finally launched Digital Electronic Government
Directorate (DEGD) which gave a boost to local IT industry by creating mega
projects like E-government Portal, NADRA, and the web portals and web sites
of 34 Ministries/Divisions developed in 2002.
According to PSEB presently there are at least 1763 active IT
companies registered with PSEB in Pakistan with around 611 active
companies in Karachi, 544 in Lahore and around 479 in Islamabad. These IT
companies specialise in the domains of software development, networking,
printing, multimedia, call centres and Business Process Outsourcing (BPO).
Growth in these major cities has been three fold over the last five years.
PSEB reported the size of local market U.S. $2.6 billion and IT enabled
exports registered at State Bank of Pakistan raised to $1.6 billion (PSEB,
2010).
Since 2007, The IT sector has got greater attention from the GOP. The
National Commission for Science and Technology (NCST) is the top decision
making body that provides directions to the scientific and technological
development of the nation through the office of the Pakistan Council for
Science and Technology (PCST). The focus of NCST is on the acceleration of
scientific and technological capacity building for rapid and sustainable
economic growth. PCST is responsible to ensure proper linkage between
science and technology and production sector in the local industry. During the
years 2006-2007 PCST has scheduled to launch of more than 300 projects for
the development of science and technology in general and for the promotion
of information technology in particular. This represents a major step forward
towards building an indigenous science and technology capacity and a
knowledge-based economy in Pakistan (PCST, 2010). According to PCST in
the private sector Worldcall, Wateen Telecom and Micro Broad Band have
Chapter 1 Introduction
8
also laid down their fibre Optic networks in Islamabad, Lahore and Karachi. A
fully integrated international standard fibre optic & fibre optic cable
manufacturing facility is also functioning in Pakistan and second new foreign
investments are expected in this sector too. Pakistan is still thinking to go to
T1 bandwidth connectivity (Sulkani, 2007).
1.5. ROLE OF MINISTRY OF SCIENCE & TECHNOLOGY IN IT
SECTOR
The Ministry of Science and Technology (MoST), GOP has been taking
key measures to encourage Foreign Direct Investment (FDI) in the country.
The aim is to make the proposition financially attractive and simplify the FDI
process to open up the opportunities in Pakistan's IT sector. In this regard,
several policy measures have been taken that include for example Ministry of
Science & Technology National IT Policy and Action Plan (MoST, 2000 ).
Pakistan Educational Research Network (PERN) by Higher education
Commission (HEC) is established to enable sharing, among educational
institutions, global digital libraries of teaching and learning materials and to
promote faculty research collaboration among local and international
educational institutions. Special concessions like lower bandwidth rates for
universities, educational institutions, software exporters and Internet Service
Providers (ISP) are part of the package.
A few of the policy inducement laid down in the IT Policy include Income Tax
holiday for IT companies and IT Professionals. Along with that software
exporting companies are allowed to retain 35% of their earnings in foreign
currency accounts. Educational grants for scholarships and enhancement of
IT infrastructure in the public sector universities have been provided. Initially
import duty on computers and parts was exempted which is again reverted to
15% in 2009. National Computing Education Accreditation Council (NCEAC)
was also established to ensure quality of training and IT education provided
by the training institutes and Higher Education Institutions (HEI) (NCEAC,
2010). Higher Education Commission (HEC) also made available foreign
expatriate faculty to improve the quality of faculty and students (MoST, 2000).
Chapter 1 Introduction
9
The Pakistan Technology Board (PTB) was established with the
Minister for Science and Technology as its head to appraise technology
needs and view national and international trade and technology implications
(PTB, 2010). PTB advises technology transfer and fosters the public/private
partnership in commercializing locally developed technologies. Provincial IT
Boards established to ensure quality IT education, strengthen IT educational
institutions, develop databases, and build capacity in IT job markets and also
to establish linkages with industry. In addition it will provide services to Small
and Medium Enterprises (SME) for training, product development,
consultancy and quality improvement (MoST, 2000).
The Pakistan National Accreditation Council (PNAC) accredits
agencies providing certification of ISO-9000 and ISO-14000 standards,
laboratories for testing and calibration, and register’s auditors and offers
training courses in the area of quality control. So far 3000 Pakistani firms have
acquired ISO 9000 certification under this program (PNAC, 2010 ).
1.6. PAKISTAN’S INITIATIVES FOR QUALITY ENHANCEMENT IN IT
SECTOR
In 2002, Pakistan Software Export Board (PSEB) came up with a
quality enhancement plan for IT Sector of Pakistan. ISO 9001:2000 (now
revised ISO 9001:2008) had international acceptability in 165 countries
including western countries who would want to invest in Pakistan. PSEB
offered selected 80 IT Companies from all over Pakistan a packaged deal
where each company was to receive a full consultancy service from selected
and reputed ISO 9001 consultancy firms, and get certified by an authentic and
reliable certification body. The certification program was supposed to provide
much needed boost to the IT market and grab attention of western market
once again. The response to this step by PSEB was so overwhelming, that
their financial limit was increased by the Government to support 20 more
companies, making a total of 100 IT companies.
A full detail of the project plan of PSEB and the total 100 companies
(all of which were certified successfully) can be found at PSEB official website
(PSEB, 2010). Currently there are number of consultants and certification
Chapter 1 Introduction
10
bodies available for ISO 9001. By 2004, ISO 9001 had become quite
affordable and many software houses achieved certification. A detail about
software houses in Pakistan can be viewed at official site of Pakistan
Software Houses Association (PASHA, 2010)..
PSEB introduced another plan especially for software development
companies for achievement of CMMI (Capability Maturity Model Integration)
certification. This is still one of the most expensive and difficult certifications,
so very few organizations have been able to invest in it. Under the sponsored
programs of PSEB, by the end of 2007, Pakistan was expected to have 20
CMMI assessed IT companies as predicted by. (Sulkani, 2007). As per latest
statistics from PSEB there are only two CMMI maturity level 5 companies, 3
The main purpose is to provide structured approach to the software
process assessment which involves the organization to improve its own
processes and to determine its capability for the particular requirement. Also
acquire to determine a supplier’s capability for particular requirement. The
SPICE component has nine parts and each part has its own task to perform in
order to determine the capability of the processes of the organization. These
are briefly explained below (Emam and Jung, 2001).
Part 1: Concepts and introductory guide
It is an entry point into SPICE. It guides to select and use of SPICE
parts and their requirements and applicability of assessment.
Part 2: A reference model for process and process capability
Defines two dimensional reference model that identifies a set of
processes in terms of their purpose and a framework for evaluating
capability of processes through assessment of process attribute
structured into capability levels. (ISO/IEC 15504) , (Jung, 2005)
Part 3: Performing an assessment
It defines a framework for conducting an assessment, and sets out the
basis for rating, scoring and profiling process capabilities, and how to
get reliable outcomes.
Chapter 3 Quality Models
56
Part 4: Guide to conducting assessment
It leads to select and use of an assessment model. It is generic for all
organizations that perform assessment using different methods and
tools.
Part 5: An assessment model and indicator guidance
It provides a prototype model for performing an assessment that is
compatible with reference model.
Part 6: Guide to competency to assessors
It describes the relevant competence, education, training and
experience of assessors.
Part 7: Guide for use in process improvement
It provides the guidance for process improvement by using results of
process assessment for the purposes of process improvement. It is
also supported by relevant case studies.
Part 8: Guide for use in determining supplier process capability
It provides the guidance for process capability determination by using
results of process assessment. . It addresses process capability
determination in both straightforward situations and in more complex
situations involving constructed or future capability. It is also supported
by relevant examples.
Part 9: Vocabulary
It is a consolidated vocabulary of all terms defined in SPICE. SPICE
defines a reference model (ISO/IEC 15504: Part 2) for process capability
determination. It is a two dimensional reference model enclosing both process
and capability. The associated software processes are classified into five
categories in process dimension and capability dimension comprises of 6
capability levels (0 – 5) indicating Process Attributes (PAs). PAs are
Chapter 3 Quality Models
57
applicable to any process with measurable characteristics necessary to
manage a process and to improve its performance capability. “An ISO/IEC
15504 assessment is applied to an Organizational Unit (OU) (ISO/IEC 15504:
Part 9). An OU is the whole or part of an organization that owns and supports
the software process.” According to ISO/IEC 15504 (Part 2, 5), the capability
TABLE 2 OVERVIEW OF SPICE CAPABILITY LEVELS Capability Level Description of Capability Level and Process Area Level 0 There is general failure to attain the purpose of process. Incomplete Process
Level 1 Performed Process
Process Performance : The purpose of process is generally achieved without planning and tracking. There are identifiable input work products that testify to achieve output work products.
Level 2 Performance management: The performance of the process
is planned, tracked and managed. The process delivers work products of acceptable quality, conform to specified standards and objectives.
Managed Process
Work product management: Process performance is documented, managed and controlled to produce work products.
Level 3 Process definition: The process is performed and managed using a defined definition. Individual implementations of the process use approved, tailored version of standard and documented processes to achieve outcome.
Established process
Process resource: The extent to which processes utilize appropriate resources to deploy the processes in order to achieve out comes.
Level 4 Process Measurement: The process is quantitatively understood and controlled. Process performance is measured to achieve process and business goals. The detailed measures of performance are collected and analyzed continuously.
Predictable process
Process control: Extent to which process remains within control limits through continuous process measurement to achieve process and product goals.
Level 5 Process change: Change management process is controlled to achieve optimized Process performance to meet the business and process improvement goals.
Optimizing process
Continuous improvement: Extent to which changes to process are managed and controlled through continuous improvement to fulfil business goals of organization. Continuous process monitoring against defined goals is enabled by obtaining quantitative feedback and improvement and analysis of results.
Chapter 3 Quality Models
58
level of each process instance is determined by rating PAs. Table 1
elaborates the capability levels and process attributes. (Jung, 2005) and
(Emam and Jung, 2001).
Each PA is measured by an ordinal rating ‘F’ (Fully Achieved), ‘L’
(Largely Achieved), ‘P’ (Partially Achieved), or ‘N’ (Not Achieved) that
represents the extent of achievement of the attribute as defined in ISO/IEC
15504: Part 2. In the process dimension, the processes associated with
software are defined and classified into five categories known as the
Customer-Supplier (CUS), Engineering (ENG), Support (SUP), Management
(MAN), and Organization (ORG). The above dimensions of SPICE were
reported by (Jung, 2004), (Emam and Jung, 2001).
3.4.1 STRENGTHS OF SPICE
SPICE can be used to inform process improvement within a technology
organization. Process improvement is always difficult, and initiatives often fail,
so it is important to understand the initial baseline level, and to assess the
situation after an improvement project. SPICE provides a standard for
assessing the organization's capacity to deliver at each of these stages.
In particular, the reference framework of SPICE provides a structure for
defining objectives, which facilitates specific programmers to achieve these
objectives. Process improvement is the subject of part 7 of SPICE.
There are few benefits of the SPICE which include: (Emam and Jung,
2001).
For acquirers: An ability to determine the current and potential capability
of a supplier's software processes.
For suppliers: An ability to determine the current and potential capability
of their own software processes; An ability to define areas and priorities for
software process improvement; A framework that defines a road map for
software process improvement.
Chapter 3 Quality Models
59
For assessors: A framework that defines all aspects of conducting
assessments.
The techniques of assessment are flexible enough that SPICE can be
adopted by small and large organizations. Its techniques can be extended to
include the quantitative measures for monitoring improvement of processes.
The use of process assessment encourages the culture of constant
improvement and develops the proper ways to support that culture. SPICE
helps to meet business requirements by engineering the processes and
optimizes the resources. SPICE helps the organization to satisfy the
customer, minimize the full life time cost, and maximize the responsiveness to
customer and market requirements. Software suppliers only submit just one
process assessment scheme where as presently numerous schemes are
used. SPICE not only beats the change management for improvement but
also provides the process capability determination at every single process of
change effected in software process. SPICE provides the basis in the
organization to assess and evaluate their limited area of software
development. The organization has the tool to initiate and sustain a
continuous process improvement. The programme managers can ensure that
their software development is aligned and supported by the needs of the
organization. SPICE has taken the initiative to support small companies.
SPICE is supported by the international community (Jung, 2004), (Emam and
Jung, 2001).
3.4.2 WEAKNESSES OF SPICE
SPICE has some weaknesses too. These are:
SPICE is not the framework to set out the specific standards, it only
assesses the capability provided by the organization's defined process
definitions and management commitment. SPICE is not a methodology, it sets
out a list of activities but does not set out the order in which the activities
should be carried out. It is expensive and not readily available for
implementation. (Emam and Jung, 2001).
Chapter 3 Quality Models
60
3.5 PERSONAL SOFTWARE PROCESS
The employee’s personal skills and abilities to manage work crisis
largely determine the results of software development process. It is a
challenge for top management to improve personal performance of an
employee. The personal performance of an employee is very important
because the personnel’s cost constitutes 70 percent of the cost of software
development (Lakha, 1994). Personal Software Process (PSP) is a disciplined
and structured methodology to software development for an individual. It
evaluates the personnel skills and provides a regimented approach to improve
personnel performance. (Pomeroy-Huff et al., 2005) It edifies employees
about managing projects quality, make commitments they can meet, improve
planning abilities, how to define process, measure quality and productivity
and reduce defects in their products. PSP deals with individual employees.
According to Mike Grasso in a seminar, PSP can be applied to many parts of
the organization including small program development (SEI, 2010)b, (Hayes
1997) and (Grasso, 2005). Personal software processes can also be applied
to SME due to small size of the organization and small nature of the project
that are usually done in SME.
The Personal Software Process (PSP), is a product of SEI developed
by Watts Humphrey. This methodology is meant to bring discipline,
consistency and efficiency into software project development for achieving
high quality product output in a small program development environment. It
helps the IT practitioners to manage quality at work place and make capable
commitments for project deadlines that they can meet (Ferguson et al., 1997).
According to Pomeroy-Huff et al. (2009) PSP follows a process level hierarchy
namely Planning, Design, Code, Testing and Post-mortem.
Chapter 3 Quality Models
61
TABLE 3 PSP Process Hierarchy
3.6 TEAM SOFTWARE PROCESS
The Team Software Process (TSP), developed at the SEI, is designed to
facilitate superior performance of software development teams in SMEs. The
TSP along with the Personal Process Software (PSP) helps the high
performance engineer to ensure quality software products and improve
process management in an organization. TSP uses integrated team concept
with 3-20 engineers to develop software intensive products. An organization
using TSP can built self directed teams that can plan their processes and
track their established goals, and own their processes and plans (SEI, 2010)b.
According to Ferguson and Kitson (1997) the TSP is a methodology that helps
organizations implement processes and best practices at the team and project
level. Both TSP and PSP have been successfully implemented for SPI in
small settings. It is used as supporting methodology for management,
planning and tracking activities. It covers most of the requirements of the
Quantitative Process Management (QPM) and Quality Management (QM)
KPAs of the CMM Maturity level-4. The usage of TSP as a foundation for
implementing the CMM in a small organization has shown that TSP makes
the CMM implementation easier. TSP has good coverage of the CMM at a
LEVEL PSP Description
O Not performed Current process are running on ad‐hoc basis
1 Planning
Define the process, create conceptual design, estimate product size, estimate resources and schedule of product development
2 Design & Review Design program, and implement design according to
developed schedule.
3 Code & Review Compile the program, , fix and log all the errors in the defect
log.
4 Testing
Test the program and fix and log all the defects found.
5 Post‐mortem
Record all the data in the project summary form including, time, defects and size on actual basis for comparison. Lessons documented for future
Chapter 3 Quality Models
62
project level. It has good coverage of CMM at team level. However if all teams
are exactly following the TSP, there are still many other uncovered
organizational aspects of CMM (Serrano et al., 2006). It is suggested by
Ferguson (1997) and Hayes (1997) that smaller organizations should consider
PSP and TSP models for process improvement.
According to Wall etal., (2005) CMMI is meant for building organizational
capability where as TSP and PSP represent a complimentary position to
support CMMI practices. TSP is for building self directed teams within an
organization and PSP framework is meant for building and transforming skills
and habits of individuals. Experiences of large body of evidence depict that
TSP addresses key goals of both Software CMM and CMMI, namely,
delivering high-quality software products, on time and within budgeted costs
(McAndrews, 2000). Furthermore, TSP processes in industry practices have
depicted a close correspondence to CMMI practices (McHale 05). TSP is
also efficient in staging software organizations to accelerate their
accomplishment of high maturity and good business performance (Pracchia,
2004 and Switzer, 2004).
3.7 SIX SIGMA
According to Basu and Wrightt, (2002) the goal of Six Sigma is to
increase profit by eliminating variability, defects and wastes and total
orientation towards customer satisfaction. Six sigma is a holistic approach
that integrates all organizational functions like staff, culture, quality system to
collectively strive towards continuous process improvement and achieving
virtual perfection. Stone (2006) defines Six Sigma as an effective method
which aims to reduction in variation, prevent defects and continuous
improvement towards achieving selected targets and goals. .
Chapter 3 Quality Models
63
TABLE 4 Structure of TSP
MATURITY LEVEL
TSP STRUCTURE DESCRIPTION
0 Not performed Ad‐hoc based current processes
1 Launch
Review Project Objectives, prepare customer needs policy, define TSP team structure, assign team roles and establish goals
2 Define Strategy
Create product conceptual design, define development strategy and decide what will be produced in each TSP cycle. Prepare project size and effort man hour estimates, configuration management plan and risk management and monitoring plan.
3 Plan
Estimate the size of software artifacts like , SRS, UAT, code etc. Establish Software development plan, and allocate weekly tasks to team. Prepare a Quality management Plan.
4 Requirement & Design
Develop functional and non functional specifications according to customer needs. Review the requirements and develop user acceptance test plan for the system. Create functional and non functional design and review the design as per requirements. Prepare an integrated system test plan
5 Implement
Use PSP to implement modular components. Prepare a detailed modular design and review the design. Implement the design by coding and review the code. Compile and test each module and assess the quality of each module.
6 Test Develop integrated system testing plan and carry out system testing. Generate user documentation.
7 Postmortem
Conduct Postmortem analysis and prepare summary for each TSP cycle. Generate peer review and team evaluation reports.
References: (William. 2009) & (Humphrey, 2000)
Six Sigma Metric level is for measuring defects and improving quality;
and a methodology to reduce defects levels below 3.4 Defects per (one)
Million opportunities (DPMO). According to Harry Mikel (1988) Six Sigma
methods ensure processes to produce output within specification. With the
help of Six Sigma method processes can reduce below than 3.4 defects per
million opportunities. Six Sigma makes the organisation more goal oriented
and aggressive towards quality objectives.
The fundamental objective of Six Sigma is implementation of a
measurement-based strategy that focuses on process implementation and
Chapter 3 Quality Models
64
variation reduction through the application of Six Sigma improvement projects.
According to Stone (2006) Six Sigma is implemented with DMAIC
methodology which is a formal analytical system for incremental and
continuous process improvement. It is an innovative, systematic, and close
loop measurement process that eliminates unproductive steps. It has basically
five phases (define measure, analyze, improve and control).
1. DEFINE: Involves team charter, process mapping and total focus on
customer needs and expectations
2. MEASURE: It is based on data collection, process measurement and
control of process performance variation.
3. ANALYSE: It is based on data analysis, process analysis and customer
focus. Performance of all artefacts is compared with standards, is
quantified for goal refinement.
4. IMPROVE: Problem solutions and alternatives are generated and
optimized solution is selected through regression. Solution is implemented
by first preparing an implementation plan and doing pilot testing.
5. CONTROL: All the processes are monitored and documented. Good
practices are institutionalized throughout the organization. Post-mortem
summary report prepared for future implementation.
3.8 BRIEF COMPARISON SOFTWARE QUALITY STANDARDS
In the following section a brief comparison between ISO and CMM,
CMM and CMMI and ISO and CMM is provided.
3.8.1 ISO AND CMM
This ISO standard has many versions for the software industry and
they are still improving their standard according to the new needs of the
organizational cultures. The ISO 9001 has given the less emphasis on the
processes in the organization although its focus has been to develop quality
product. The CMM in contrast has emphasis on the process improvement and
also in the continuous manner. The ISO 9001 emphasizes more on product
engineering and the hardware whereas CMM deals with the development of
Chapter 3 Quality Models
65
the software processes. ISO 9001 does not need detailed documentation
during different phases of the processes in the organization. On the other
hand the CMM requires detailed documentation for the processes.
The CMM is more suitable for the development of the quality product
and quality improvement in the standard of the organization. CMM still has
limitations that it cannot be fully implemented in the SME due to requirement
of more resources for documentation which makes the implementation more
costly as compared to ISO 9000 certification. ISO 9000 model addresses
minimal criteria to establish a Quality Management System (QMS), where as
CMM has a detailed approach to address to quality improvement paradigm. It
can be said that every KPA of CMM is weakly related to ISO 9000 to some
extent (Paulk, 2005). There are strong correlations between IS0 9001 and
CMM level 2, specifically in relation to process quality improvement .and
transition from CMM to ISO 9000 is possible without disturbing the integrity of
the ISO 9000 certified organization. This transition can be accomplished in
easy five steps namely Establish Software Engineering Process Group,
Perform Gap Analysis, Make a Plan, Provide Training and Establish Metrics
Program ( McGuire and McKeowen. 2001) It can still be argued that spirit of
TQM culture is not found in the practices. May be this issue will be addressed
in the latter additions of these models.
3.8.2 CMM AND CMMI
Both Software CMM and CMMI models are based on the pretext that
organization will follow process improvement journey in small incremental
steps rather than bringing radical change through large scale sweeping
changes. Quality improvement through reengineering can be bought through
department wise small evolutionary steps and by repeating small wins
successively across the organization (Wall et al., 2005). As asserted by Paulk
et al., (1993) that software CMM and CMMI (staged) quality improvement
models provide a baseline for incremental SPI by defining five maturity levels
that lay down a framework with measurement and assessment criteria for an
organization’s software process maturity and for assessing its SPI capability.
Chapter 3 Quality Models
66
Each of the five levels is composed of a set of KPA with component goals and
practices, that, when satisfied, provide considerable improvement in
organization’s software processes. CMMI continuous grants freedom in
improving only those process areas which are critical for organization to
improve and mitigate the risk where as CMM has a defined path for over all
The reliability of the data is determined from practical considerations.
The questionnaire was pretested during the pilot study in order to remove
ambiguities, replication of questions and research rationale. Purpose and
objectives were clearly stated in the cover letter from the researcher. Data
was collected through multiple channels and sources to get accurate and
complete unbiased sample, as emphasized by (Stake, 2005). Data quality
was checked statistically to filter the questionnaires which were skewed at
extremes. Partially filled questionnaires were left out of the analysis. During
field visits decorum and manners were insured to develop the interest of the
participants so that they are not bored (Lincoln and Guba, 1985) and (Stake,
2005). In order to ensure confidentiality of the respondents as advised by
Patton, (2002) respondents were reminded not to fill name or email address.
Assurances were also given to respondents about confidentiality and
anonymity (Cohen and Manion, 2000). In order to further clarify the concepts
and interpretations, all questions were pre-tested by discussing them with
several IT Professionals, in an attempt to eliminate biases of ambiguity
(Elphinstone, 1990).
4.5 THEORATICAL FRAMEWORK FOR DEPENDENT VARIABLE
To find out that what are the most significant best practices that effect
SPI so that a set of minimal practices for the local software industry can be
proposed as a SPI paradigm for implementation of quality, a framework
shown in FIGURE 4 was proposed. In this framework SPI constructs are
obtained from research design that included a detailed process to develop a
questionnaire for this study. These constructs were then finalised after pilot
study. Process Quality Improvement (PQI) is the dependent variable and all
other constructs are independent variables.
Chapter 4 Methodology
80
FIGURE 4 THEORETICAL FRAMEWORK
4.6 SURVEY ADMINISTRATION
For the purpose of data collection necessary measures were used to
pilot test data collection tools to ensure accuracy, relevance and reliability for
quality of data. IT practitioners working in software houses who were the
members of Pakistan Software Export Board (PSEB) or Pakistan Software
House Association (PASHA) and IT practitioners working in companies that
are registered with ROZEE, (2010) were contacted for survey. The exercise to
implement survey was conducted through following sampling technique.
4.6.1 SAMPLING PROCEDURE
A systematic random sampling technique was adopted to collect the
sample. A list of 1031 software houses was compiled by collecting names and
emails of the companies which were registered with PASHA, PSEB and
www.rozee.com.pk (ROZEE, 2010) A small sampling interval of n=3 was chosen to
attain maximum population. Systematic sampling tolerates better
representativeness as compared to simple random sampling, assuming that
there is no cyclic pattern in the distribution list. Through systematic random
sampling technique good geographical distribution according to population
PROCESS QUALITY IMPROVEMENT
QUALITY BEHAVIOUR
REQUIREMENT
PLANNING
MONITORING
MEASUREMENT
STRUCTURE
CULTURE
Chapter 4 Methodology
81
density is achieved ( Iwan Ariawa, 1998). It was insured that there were no
duplicate names or emails in the list and it was not an ordered list. A
systematic sample was drawn by selecting every 3rd name from the list and a
sample was compiled of 343 names in first run. Following the same technique
229 and 153 names were selected in the second and third rounds of
systematic selection. In order to improve the response rate the respondents
were also given option to fill out the questionnaire placed online at
www.tecnologiz.com/quality (Sheraz, 2010). A total of 725 questionnaires
were administered on local IT practitioners through mail. The response rate
through mail was 19.8%, 144 out of 725 of the respondents filled out complete
questionnaires.
To further improve the response rate data collection was also administered in
person. As an additional effort during the period of last six months more than
90 software houses from the compiled sample list were approached from time
to time. A telephonic appointment was taken before approaching the software
houses. The response was quite encouraging from the lower management but
response from the top management was poor. Majority of the respondents
showed keen interest and answered the questions very carefully. These
interviews helped a lot to find out the attitude and concern of the organizations
about quality improvement. The response rate was fairly good and 83
completed questionnaires were achieved. The total number of questionnaires
completed for the study was 227, and over all response rate of 31.3% was
achieved.
4.6.2 POPULATI ON SAMPLE
In empirical studies it is important to select an unbiased sample and
hence an unbiased response (Salant and Dillman, 1994). A systematic
random sampling technique was used for this study. In order to choose
representative population for the research on Pakistan software industry
companies registered with Pakistan Software Export Board (PSEB), Pakistan
Software Houses Association (PASHA) and IT companies registered at
www.rozee.com (ROZEE, 2010) were selected. Most of the respondents
Chapter 4 Methodology
82
were from four biggest metropolitan cities of Pakistan namely Lahore,
Karachi, Islamabad and Rawalpindi. Altogether, the total number of software
houses situated in these four cities represents approximately more than 80%
of the software industry in the country. Examining the response from the IT
practitioners working in these organizations, who represent the target
population sample, is expected to help us to produce information about the
general nature of sample population characteristics.
4.6.3 SAMPLE SIZE DETERMINATION
The optimum sample size determination technique is taken from
Lwanga, S.K. et al, (1991). According to this formula the calculated minimum
sample size for this study comes out to be 199, therefore sample size of 227
completed questionnaires obtained for this study is justified.
Equation:
Where
z = 3.84 at 95% confidence Interval (CI) P = 0.2, proportion of anticipated study population N = 1031, Population Size, d= 0 .05 absolute precision (spread+_ 5 %)
4.7 DATA ANALYSIS
In order to do data analysis frequencies were generated for indicator
variables, and measures of central tendency, Mean, Median, Mode were
calculated for numerical variables where applicable. For Inference: Chi Sq
was applied to generate inference for categorical variables was applied and
students t-test for continuous variables where applicable. Model was
generated using linear regression and structure equation modelling (SEM).
Chapter 4 Methodology
83
Data was presented using Statistical Package for System Simulation (SPSS)
version 1.6 generated values in the form of figures and tables.
4.8 STRUCTURAL EQUATION MODELING In order to further test the model fitness Structural Equation Modelling (SEM)
technique was applied A specialized statistical software called Analysis of
Moment Structures (AMOS) was used to further refine the model through
empirical analysis to come up with an optimum SPI paradigm. SEM basically
describes relationship between variables. SEM technique is similar to
regression modelling and factor analysis and is effective in a way for
removing multi-co-linearity in the model. AMOS has a graphical interface .and
is an excellent tool to use for SEM model fitting (Wei, 2009). A framework of
structural modelling is given in FIGURE 5. Discussion on empirical analysis
and SEM is given in CHAPTER 6 ANALYSIS AND .
Chapter 4 Methodology
84
FIGURE 5 THEORATICAL STRUCTURAL MODELING
Chapter 4 Methodology
85
4.9 SUMMARY
The chapter describes how the data collection instrument was
designed and how it went through a pilot testing to further increase its
reliability, and subsequent development of final quality constructs. Survey
administration section discusses in detail the considerations regarding
population sample selection and systematic random sampling. In the last
section theoretical frameworks for regression modelling and structure
equation modelling (SEM) are proposed. The next chapter gives the
descriptive results.
Chapter 5 Descriptive Results
86
CHAPTER 5 DISCRIPTIVE RESULTS
In this chapter descriptive results are reported from the statistical
analysis that was conducted on the data collected through the questionnaire
implemented to determine the nature of local software process improvement
practices in the SMSHs. The data was mainly collected from the four main
metropolitan cities of Pakistan namely Karachi, Lahore, Islamabad and
Rawalpindi. Total number of (227) completely filled questionnaires were
received from respondents, who were IT practitioners working in the local
software industry. Likert’s scale was implemented in the data collection
instrument. The results were divided into 8 sections according to the
constructs and percentage frequencies answered by respondents against
each indicator were reported as results in this chapter. The reported
percentage in text is based on the following criteria. The likert’s scale values
corresponding to 1 and 2 are accumulated to represent “LOW”, and likert’s
scale values corresponding to 4 and 5 are accumulated to represent “HIGH”
as both concepts either give negation or acceptance of indicator. The
perceptions of IT practitioners working in the local software industry are
presented in the following section..
5.1 FREQUENCY ANALYSIS
In this section, percentage frequency and simple frequency of
responses against each item in the questionnaire based on eight quality
constructs have been presented from TABLE 6 to TABLE 13. These percentages
have been discussed to check the perception level of each quality assurance
practice in local software houses. Total numbers of Indicators included in the
descriptive results were 47 according to the items statements presented to
respondents through the questionnaire. Results are reported through negation
and acceptance of respondent’s perception for indicators where applicable.
For management commitment for quality improvement and
process assessments as indicated in TABLE 13 48.9% of IT practitioners don’t
agree that management supports and allocates separate budget for quality
improvement whereas 31.3% agree that management does allocate funds for
quality improvement efforts. Mostly 52.4% of the organizations do not have
Knowledge Base Library (KBL) nor do they have configuration management
tool to maintain digital shared repository. 26.9% respondents agree that they
Chapter 5 Descriptive Results
95
do have KBL as shared repository and they do apply SCM practices. For
quality assurance and process improvement as indicated in TABLE 13 47.2%
companies do not have dedicated staff only for quality improvement and
29.9% of companies agree that they do have separate staff for QA to do
process tailoring and quality improvement of standard processes. For
assessment of maturity level as shown in TABLE 13 56.8% claim that their
maturity level is high, while 14.1% said that their maturity level with respect to
quality is low.
5.16 QUALITY MODELS PRACTICED IN LOCAL IT INDUSTRY
In order to ascertain the nature of quality practices and vision of
software industry a direct question was put in the survey that what kind of
quality model is being followed at your work space. 42% of respondents
selected ISO option which means that most of the companies are following
ISO Practices. It is interesting to notice as shown in TABLE 14 12% are
following CMM and 7% claimed to follow CMMI. Overall 35% of the
practitioners selected “Other” option which may be interpreted as that most of
them may be following other quality models or indigenous organizational
processes model approach. Only 4% reported using TSP/PSP for quality
management.
TABLE 14 QUALITY MODEL DEMOGRAPHICS
MODEL ISO CMM CMMI TSP/PSP OTHERS
PERCENTAGE 42% 12% 7% 4% 35%
5.17 RESPONDENT PROFILES
The data collection instrument was filled by IT practitioners working in the
local industry who were working at different levels of management like top,
middle and lower. The designation profile of respondents is given in TABLE 15.
The total sample size was 227 out of which 7% were filled by top
management, 39% filled by middle management and 54% questionnaires
Chapter 5 Descriptive Results
96
were completed by lower management. Detail of management groups is as
follows;
TABLE 15 RESPONDENT’S PROFILE GROUPS
TITLE PERCENTAGE DESIGNATION
TOP 7% CEO, Chairman
MIDDLE 39% Managers, Assistant (manager ,directors) Program Managers, Line Manager
LOWER 54% Software Engineer, Developers, Sys-Analyst, QA Person and Web developer
5.18 PROBLEMS AND ISSUES RAISED
According to research question 3: What are the problems and issues faced by the local practitioners to implement SPI quality practices?
Many comments received during the survey. A few respondents also filled
the comments section of the questionnaire as feedback. During the personal
visits enlightening discussions were held with IT practitioners.. Following is
the general idea of issues and problems raised regarding Software Quality
and top management in the local SMSHs.
Culture Gap: A fundamental problem was identified regarding culture
gap between software industry culture and manufacturing and service
industry culture. The later practice quality principles rigorously and all
employees are accounted for punctuality and production non
conformances. If the employees sit late they are given extra salary for
overtime. Whereas there is very little consideration for punctuality in
software industry and employees are made to sit late to complete
deliverables and iterations to meat poorly estimated client deadlines.
Software practitioners are not acknowledged and paid for such late
sittings and no over time is given. Local software practitioners should
Chapter 5 Descriptive Results
97
develop a quality culture where they should follow their written
processes religiously.
Documentation: The software industry blindly follows quality
processes much of time is wasted for documentation. The
management has not enough domain knowledge or have domain
experts to train in quality or practice quality excellence. Management
should follow a lean process measurement policy to only measure
critical process areas that need high supervision. A culture of
measuring everything does not promise to reveal productive results.
Quality Tools: There is no awareness and implementation of quality
tools among the management of SMSHs. Quality tools are available in
international market but their prices are so high that acquiring highly
expensive quality tools for SMSHs is not feasible due to their low
stream of cash flows. Secondly management considers quality as a
cost. SMSHs should develop separate budget for purchasing quality
tools or start developing indigenous tools for quality management and
start offering it as a product for local SMSHs.
Low Salaries: Majority of the IT practitioners also pointed out the
problem of low salaries which they were being offered in the local
industry. SMSHs fail to retain employees due to their short term
planning and policies. The high turnover rate in IT industry is due to
tough local competition among SMSHs. The salaries for senior
management are less as compared to salaries in manufacturing or
service industry when compared on long term basis. Later gives more
benefits like car and residence and tries to retain the employees for
long-term basis. One of the reason may be that most of the
manufacturing industrial zones are situated in remote locations.
Quality Check at End: The local SMSHs have a culture practice to
check quality of a software product at the end when it is completed.
Chapter 5 Descriptive Results
98
Any requirement changes or functional and non-functional non-
conformances lead to high change management costs which lead to
project delays, customer dissatisfaction and budget overruns. Such
projects do not remain profitable for SMSHs. Top management should
develop new quality philosophy and should develop a paradigm to
measure product quality throughout the software development life cycle
and not just at the end.
Rethink Quality: Trainings are not offered for awareness and
implementation of quality. Quality is just offered theoretically with high
load of documentation. Management considers quality as a burden.
Leadership should develop commitment towards quality and start
rethinking quality, culture change and adopt total quality philosophy
and start considering quality as an integral part of organization on long
term basis.
Poor Baseline Knowledge: There is an acute shortage of quality
domain experts at lower or baseline level in local SMSHs which is one
of the reasons for poor implementation of quality. Processes should be
developed to involve lower level employees to learn and develop
domain expertise in quality. Templates, Standard Operating
Procedures (SOP) and easily accessible manuals should be provided
to develop and enforce quality culture.
Change Management: Requirement management and change
management processes are not followed using quality guidelines which
leads to non conformance due to incomplete information and delays in
projects. Management should develop proper processes for change
management and customer requirement gathering which should be
governed by an effective quality control system.
Lack of Planning: The project quality and delivery depends upon
schedule planning through Project Development Plan in industrial
Chapter 5 Descriptive Results
99
practices that project plans are only made for the sake of
documentation and are not followed in letter and spirit. The variance in
schedule planning may affect timelines and project cost. There should
be strict vigilance in schedule planning to prevent from rework and
project delays. Project resources should be planned before project start
to avoid inconvenience from available resources.
5.19 PROBLEMS IN IMPLEMENTATION OF QMS IN PAKISTAN’S IT
INDUSTRY
Tight budgets do not allow most of the organizations to invest on QMS
(or related practices) as most of them only adopt it to create a better
market impression rather than improve their system. Most of the SME’s
only adopt QMS when external financial support is involved. PSEB and
Business Support Fund (NGO) are some of the organizations providing
support to SME to implement such practices.
Lack of financial resources leads to few investments on human
resource training and more on technological solutions, untrained staff
leads to unpolished Quality System.
Lack of trainings for most of the staff (usually a selected few are trained
from external sources and are provided with certificates that increases
their academic qualification level and hence motivates them), leads to
lack interest in QMS.
Many QMS activities can be easily recorded and documented with use
of software solutions (or deployment of ERP applications), but financial
constraints force organizations to use manual recording methods which
leads to human errors, missing entries, slower responses, etc.
eventually effecting overall performance of QMS.
Re-work is a major issue as most of the organizations cannot afford to
invest higher on Quality Assurance and focus more on development &
delivering, which leads to problems at customer’s end and reboot of
development process. Quality Assurance is mostly done with support
Chapter 5 Descriptive Results
100
of QA tools that are either free or very old as the new ones with better
features are very expensive.
5.20 SUMMARY In this chapter descriptive results of the survey are reported for each
quality construct. Problems and issues raised by the IT practitioners are also
presented at the end. Problems in implementation of quality management
systems in local IT industry are highlighted. Analysis and findings are
presented in the next chapter.
Chapter 6 Analysis & Findings
101
CHAPTER 6 ANALYSIS AND FINDINGS
In this chapter empirical analysis and its findings are presented. The
empirical analysis includes correlation and regression analysis of quality
constructs. Results of Structural Equation Modelling (SEM) are elaborated to
evaluate the evolved LQIM model. IN the end graphical representation of
LQIM model and its conceptual detail is given.
The following section addresses the fourth research question.
Question 4. What can be a proposed SPI paradigm which can best fit to
solve the problems of quality improvement in the local
software houses (SMSHs)?
In order to address this question following two statistical techniques are
used to come up with a reliable and measurably fit optimum model for SPI in
local SMSHs. First section includes output from SPSS ver. 16.0 as proof of
empirical analysis, using linear regression and correlation analysis to depict
the interrelationship between the dependent variable (QualityIMP) and 7
independent constructs as discussed in Chapter 4 (Methodology). Important
components of empirical analysis like reliability, internal validity of constructs
and data collection instrument, and external validity are also discussed. In the
second section in order to propose an optimum SPI Model Structure Equation
Modelling (SEM) analysis is carried out using Analytical Movement of
Structures ( AMOS) to further validate the results obtained through the linear
regression modelling. The objective is to test the stability of the relationships
between the measurement variables and the constructs and measure the
goodness of fit of the proposed SPI Model.
6.1 RELIABILITY ANALYSIS
Reliability Is one of the main pillars of research methods and techniques meant to endorse and put into practice the authenticity of methodology on longitudinal scale to guarantee similar research outcomes by different researchers (Yin, 2003). The instrument design and data collection procedures have been reported in detail in Chapter-4 (Methodology). The
Chapter 6 Analysis & Findings
102
instrument was pretested and reviewed by experts in order to filter out all types of misinterpretations and ambiguities for respondents and surveyors. Reliability means that a test, method or an experiment yields the same results on repeated trials (Carmines and Zeller, 1979). In order to ascertain the reliability of data collection instrument and the data collected against each indicator variable, a technique developed by Cronbach, (1951), is used during the data analysis. The SPSS ver.16.0 was used to run the scale reliability Cronbach’s Alpha test. This technique gives a value of Cronbach’s Coefficient Alpha for measuring reliability. According to Murphy and Balzer, (1989) generally Cronbach’s Coefficients value of greater than 0.70 is considered adequate. In reference to
TABLE 17 the Cronbach’s Alpha for this research is 0.839 based on 47
items of questionnaire and 227 respondents. It is a measurement of the
overall reliability of instrument. The Cronbach’s test was also applied to all
eight constructs individually and as indicated in TABLE 16 the maximum
range of Cronbach’s Alpha for (Structure= 0.77) and the minimum Range of
Cronbach’s Alpha for (Monitoring & Control =0.51 and. Therefore, reliability of
instrument is valid and good as according to Nunnally, (1978) Cronbach’s
Coefficient Alpha value of more than (0.5) is also acceptable. A study was
carried out by Van de Ven and Ferry (1980) suggested that Crobach’s Alpha
value of 0.35 is also an acceptable benchmark. In order to ascertain
construct reliability the Cronbach’s Coefficient Alpha for the eight quality
constructs is given in TABLE 16 for reference.
Chapter 6 Analysis & Findings
103
RELIABILITY TEST SPSS OUTPUT
TABLE 16 RELIABILITY OF CONSTRUCTS
CONSTRUCT ABBREVIATION Cronbach’s Alpha
Organization Size & Structure OSS 0.77 Organization Culture OCL 0.53 Organization Behaviour Towards Quality OBQ 0.67 Requirement Management RQM 0.61 Project Planning PPL 0.62 Project Monitoring & Control PMC 0.51 Measurement & Analysis MAN 0.72 Process Quality Improvement PQI 0.68
TABLE 17 RELIABILITY STATISTICS
Cronbach's Alpha N of Items
RELIABILITY /VARIABLES=SIZ TND STA OST RTN OVT SHD LRN COM TM TRA TST ASS RA TR SQI QA TPQ PR EST CS CM OSR CNF UAT PPL RM CSTD AR PM PT PER TW PMR DCS PPR DL KBL RP PI TQM QPL CPI OCM SCM SPI ML /SCALE('ALL VARIABLES') ALL /MODEL=ALPHA.
0.839 47
6.2 INTERNAL VALIDITY CONSTRUCTS
Validity measures the quality of answers provided against the research
questions. Internal validity of the data and data collection instrument includes
the determination of cause and effect relationships introduced by Dilanthi at
al., (2002) in the design of experiment by fitting a theoretical framework
behind each construct for finding out effects on dependent variables.
According to Carmines and Zeller, (1979), validity also means that if the
instrument measures what was intended to measure then that instrument is
simply valid. Development of 47 indicators of quality practices from ISO 9000:
2001 and CMM (CMU/SEI) KPAs provides a solid foundation on which to
build a methodology to assess quality improvement practices in local industry.
Chapter 6 Analysis & Findings
104
A number of similar studies have used SEM and proposed the same cause
and effect relationship between constructs and have made recommendations.
Among contemporary studies, a recent study by Khung Kok Wei, (2009) uses
the same model. On the basis of literature review, interaction with academic
quality experts and data collection from IT practitioners resulted in creation of
8 quality constructs to measure the quality improvement practices. In general
these constructs resolve the issue of evaluating interrelationships between
large number of items. It is a way of condensing and summarizing the
information into new dimension of composite size called constructs. It is also
called factor analysis (Flynn et al, 1994). The reference list representing
literature review along with detail discussion on research design and research
methodology section gives detailed discussion on engineering of data
collection tool, development of quality constructs and subsequent survey
administration.
6.3 EXTERNAL VALIDITY
External validity investigates whether the research findings in a
particular environment can be generalized for other situations in which a
sample population is investigated. Such a generalized behaviour can be a
critical characteristic of research that further signifies the research’s scope
and make it contributory to multidimensional fields and body of knowledge
(Dilanthi et al, 2002). To further enhance the external validity of the research a
systematic random sampling technique is adopted to select an unbiased
sample size from a list containing Software houses located mainly in four
major cities of Pakistan namely Lahore, Karachi, Islamabad and Rawalpindi.
The data was collected from members of PSEB and PASHA and other
software houses. The methodology to select samples systematically was
followed to get a true probability sample which can be justified statistically. It
is envisaged from the above discussion that this research has fulfilled the
prerequisites, and has good external validity
Chapter 6 Analysis & Findings
105
6.4 CORRELATION ANALYSIS
The correlation matrix for 8 quality constructs for quality improvement and
quality implementation is given in TABLE 21. where results are reported
along with Pearson Correlation Coefficient Alpha value denoted by “R”,
Level of Significance denoted by “P” and Sample size N. Correlation is
considered significant at level 0.0l if “P” value is near to 0.01, similarly
correlation is considered significant at level 0.05 in case “P” value is near
to 0.05.
TABLE 18 CORRELATION BETWEEN ALL CONSTRUCTS ** Correlation is Significant at the 0.05 level (2‐tailed)
PLANNING REQ_MGT STRUCTURE CULTURE QUALITY CONTROL MEASURE QUALITYIMP
PLANNING
R 1 P N 227
REQ_MGT
R .534** 1 P .000 N 227 227
STRUCTURE
R .204** .356** 1 P .002 .000 N 227 227 227
CULTURE R .384** .517** .283** 1 P .000 .000 .000 N 227 227 227 227
QUALITY
R ‐.236** ‐.372** ‐.488** ‐.460** 1 P .000 .000 .000 .000 N 227 227 227 227 227
CONTROL R .513** .548** .254** .449** ‐.222** 1 P .000 .000 .000 .000 .001 N 227 227 227 227 227 227
MEASURE R .514** .623** .324** .499** ‐.336** .618** 1 P .000 .000 .000 .000 .000 .000 N 227 227 227 227 227 227 227
QUALITY_IMP R .486** .547** .427** .502** ‐.341** .595** .648** 1 P .000 .000 .000 .000 .000 .000 .000 N 227 227 227 227 227 227 227 227
Chapter 6 Analysis & Findings
106
In TABLE 18, it is obvious that there is a strong correlation between the
dependent variable and all other variables in the table. Hypothetically if the
correlation is high between two variables then it is said that the two variables
have strong interrelationship characteristics. If we try to study the correlation
of dependent variable “QUALITY_IMP” with the all other quality constructs,
the TABLE 18 shows significant correlation with 7 constructs at significance
level 0.05%. The significance value of Planning is 0.486, that of REQ_MGT is
0.547, that of STRUCTURE is 0.427, that of CULTURE is 0.502, that of
QUALITY is 0.341, that of CONTROL is 0.595 and that of MEASURE is
0.648, The subsequent step in the empirical analysis is to find out the impact
in variability of dependent variable outcomes due to the independent
variables. According to Mahour, (2006) regression analysis explains which
variables(s) is significant among other variables in explaining the quantum of
variability on dependent variables. The following section gives
implementation details of the theoretical model to test the approximation of
the model already developed in chapter 4 (Methodology).
6.5 REGRESSION ANALYSIS
Following the theoretical framework for regression analysis given in
Methodology Chapter-4, regression analysis was done using SPSS ver.16.0.
The output results are given in, TABLE 19, TABLE 20, and TABLE 21. Quality
Improvement (QualityIMP) construct was put as dependent variable and all
other 7 constructs namely Structure, Culture, Planning, Control, Quality ,
Req_Management and Measurement were put in the category of independent
variables.
TABLE 19 MODEL SUMMARY
Model R R Square Adjusted R Square
Std. Error of the Estimate
1 .744 .553 .539 .52844
a. Predictors: (Constant), MEASURE, QUALITY, Structure, CULTURE, PLANNING, CONTROL, REQ_MGT b. Dependent Variable: QUALITYIMP
According to the output four independent variables are strong
predictors of dependent variable QualityIMP. As indicated in TABLE 19 R-
Square = 0.553 which means that these constructs explain 55.3% variability of
dependent variable QualityIMP. As indicated in TABLE 21 Significant
predictors of QualityIMP are; significance of Structure (OSS) is 0.000, that
of Culture (OCL) is 0.019, that of Control (PMC) is 0.000), that of Measure (MAN) is 0.000 are significant at P value less than 0.05. The value in TABLE
21 also endorse that the relationship between QualityIMP is linear with the 4
predicting independent variables.
Chapter 6 Analysis & Findings
108
Analysis of variance (ANOVA ) was also performed through SPSS ver.
16 and as indicated in TABLE 20 ANOVA significance (P- Value) is 0.000
which signifies that the model is statically significant at Alpha= 0.05.
It can be deduced from this analysis that in order to make an effective
Quality Improvement Model and guidelines for the local industry the
following 4 critical success factors will play a very important role. Dependency
model for quality improvement is given in FIGURE 6. Implications and
guidelines to implement these constructs shall be discussed in Chapter 7
where the final model is evolved by using SEM.
FIGURE 6 QUALITY IMPROVEMENT DEPENDENCY MODEL
Quality Improvement (QIMP) DepVar Organizational Culture (OCL) IndVar Project Monitoring and Control (PMC) IndVar Organization Size and Structure (OST) IndVar Measurement and Analysis (MAN) IndVar
Chapter 6 Analysis & Findings
109
At this point it is important to explore the covariance between each of
these constructs and also need to find out the impact of each unique variable
on the respective construct. For this Structural Equation Modelling (SEM)
technique is used that is discussed in the next section. It will help to analyze
this model further and help to reduce the set of variables into a lean and more
manageable Model.
6.6 STRUCTURAL EQUATION MODELING
Specialized statistical software called Analytical Movement of
Structures (AMOS) is used to further refine the model to come up with an
optimum SPI paradigm. AMOS has a graphical interface .and is an excellent
tool to use for SEM model fitting Khung Kok Wei, (2009). SEM basically
describes relationship between variables. SEM technique is similar to
regression modelling and factor analysis and is effective in a way for removing
multi-co-linearity in the model. AMOS has a graphical interface and is an
excellent tool to use for SEM model fitting. AMOS is distributed by SPSS Inc.
6.6.1 SEM IMPLEMENTATION
. A second order Critical Factor Analysis (CFA) is carried out using
SEM to further validate the results obtained through the linear regression
modelling in previous section. It tests the stability of the relationships between
the measurement variables and the constructs. At this stage all 8 latent
constructs are retained and the total numbers of measurement variable
indicators remain 47. The SEM completed 4 runs to reach acceptable
goodness of fit indices benchmark level. During the run 4 constructs got
completely deleted. In total 37 indicators were deleted from scale
development process. These deleted indicators were found to be inadequate
to load the model due to poor level of explained variance. List of the deleted
indicators is given in table 26 at the end of this section. The graphical output
of the model is given in FIGURE 7 and the recommended optimum paradigm
standard quality practices are shown in TABLE 25.
C
“G
G
c
M
Im
Im
re
6
T
e
st
w
A
hapter 6
This
GOODNES
Gerbing (19
riterion to d
Degre
Comp
Norm
Goodn
Accor
Mean Squar
FIGUR
mprovemen
mprovemen
esults and a
6.6.2 RATI
TABLE 22
Mode
Defau
Satur
Indep
The
specially if
tatistic whic
which is us
Anderson an
study fal
SS OF FIT”
988), and
determine th
es of freedom
arative fit in
ed fit index (
ness fit index
rding to Hu
re error Ind
RE 7 indic
nt Model (L
nt. The evo
all determin
IO OF CHI-
CMIN CH
el
ult model
rated mode
pendence m
value of c
the sample
ch measure
sed to mea
nd Gerbing
lowed var
of the me
Bentler (1
he model fi
m chi square
dex (CFI),
(NFI),
x (GFI),
u and Bentle
ex (RMSEA
ates the o
QIM) as so
olved mode
nistic goodn
-SQUARE
HI-SQUARE
NP
el
model
chi-square
e size is gr
es the ratio
asure the
, (1988), if
rious crite
asurement
1992) follow
t.
e /df
>0.8
er, (1999) a
A) is accept
optimized
olution and
l indicates
ness of fit in
: D.F
E
PAR CMIN
17 88.2
55 .0
10 370.6
is usually
reater than
of chi-squa
model fit.
the value o
rions for
model. Ac
wing are
< 3
8 to 0 .90, (als
> 0.90
>0.90
and Byrane
table under
evolved p
model fit fo
that CFA s
ndices pass
N DF
286 38
000 0
659 45
y very sen
200. There
are to the d
According
of this ratio
Analy
determinat
ccording to
acceptable
so acceptable
e, (1989) va
r (0.08).
aradigm L
or the Softw
showed goo
sed the mod
P CMI
.000
.000
nsitive to s
e is anothe
degrees of f
to Kline,
is less tha
ysis & Finding
11
tion of th
Anderson
e values fo
e 0.85)
alue for Roo
Lean Qualit
ware Proces
odness of f
del.
N/DF
2.323
8.237
sample siz
er chi-squar
freedom (d.
(1998) an
n 3 then it
gs
10
he
&
or
ot
ty
ss
fit
ze
re
.f)
nd
is
C
a
8
th
m
6
in
th
(I
w
is
th
T
S
6
c
m
a
m
re
hapter 6
good indic
8.286, d.f =
he benchm
model is with
.6.3 COM
Comp
ndex calcula
he null mo
ndicators)
when sampl
s 0.85. Acco
hus should
TABLE 23
SEM MODEL
MO
DefaultDefaultDefault
6.6.3.1 NOR
The N
hi-square v
model. One
ssumes th
measure an
ecommend
cator of mod
=38 and ra
mark fit inde
hin accepta
MPARATIV
parative fit
ates the de
odel with th
have no s
le size is sm
ording to Bo
be accepta
BASELIN
GOOD FIT IN
DEL
Model Model Model
RMED FIT
NFI index
value of the
e of its as
at all meas
nd it is se
ed by Bent
del fit. As in
tio Chi-squ
ex therefor
able range.
VE FIT IND
index was
egree of inte
he assump
ignificant in
mall. The S
ollen (1989
able.
NE COMPA
DICES
INDEX
GFI CFI NFI
INDEX (NF
is meant to
e proposed
ssumptions
sured varia
ensitive to
ler, (1990).
ndicated in
uare /
re the good
EX (CFI)
s first foun
errelationsh
tion that th
nterrelation
SPI Model r
9) a CFI val
ARISONS M
VALU
0.920.80.76
FI)
o determine
d model wit
is critical
ables are u
sample si
The resul
the results
d.f = 2.3
dness of fi
ded by Be
hip between
he latent m
ship. This
resultant va
ue of .85 re
MODEL FIT
UE
26 85 62
e the mode
th the chi-s
that as w
uncorrelate
ze less th
tant value o
Analy
TABLE 22 c
323. The va
t of the pr
entler, (199
n suggeste
measureme
index work
alue in TABL
epresents p
INDICES
el fit by co
square valu
worst case
d. NFI is a
an or arou
of NFI for S
ysis & Finding
11
chi-square
alue is unde
roposed SP
90). The CF
d model an
ent variable
ks well eve
LE 23 of CF
progress an
omparing th
e of the nu
scenario
an importan
und 200 a
SPI Model
gs
11
is
er
PI
FI
nd
es
en
FI
nd
he
ull
it
nt
as
Chapter 6 Analysis & Findings
112
as shown in TABLE 23 is 0.76 which is less than the benchmark and therefore
it is a mediocre indication of SPI Model fit. One reason for mediocre model fit
may be that sample size is small.
6.6.3.2 GOODNESS-0F-FIT INDEX (GFI)
The goodness of fit GFI was founded by Joreskog and Sorbom in
1984. Goodness of fit index is an alternative to the chi-square test and its
criterion is to assess the ratio of variance that is accounted for by the
approximation of the SPI Model covariance. Through such estimation it can
be predicted how close the proposed model is able to replicate the covariance
of population matrix (Tabachnick and Fidell, 2007). It has also been found that
as the value of GFI increases the number of parameters in the model also
increases.
In reference to Error! Reference source not found. the value of GFI =
0.926 which is greater than 0.9 benchmark, therefore it can be confirmed that
SPI model has a very good Model fit.
6.6.4 ROOT MEAN SQUARE ERROR INDEX (RMSEA) 6.6.4 ROOT MEAN SQUARE ERROR INDEX (RMSEA)
TABLE 24 RMSEA
Model RMSEA LO 90 HI 90 PCLOSE
Default model .077 .056 .097 .020
Independence model.179 .162 .196 .000
The RSMEA index was first developed by Steiger and Lind (1990).
RMSEA has evolved as a good measure for models fit with regard to model’s
economy or parsimony. It chooses the most optimal parameters that would fit
the population covariance matrix (Byrne, 1998). In other words it assesses the
divergence among the proposed and estimated covariance matrices per
degree of freedom. In recent years it has gained importance among
C
re
A
R
fo
b
o
a
hapter 6
esearchers
According to
RMSEA lies
or RMSEA
The r
enchmark t
f SPI mod
ffected by t
FIGUR
due to its
o MacCallu
s between 0
is < 0.8 the
resultant va
thus value
el is a bet
the size of t
RE 7 SEM
sensitivity
um et al, (
0.8 to 1.0 th
en shows a
alue for RM
asserts a g
tter approx
the sample
M STANDAR
to the num
1996) and
hen it is a m
good fit.
SEA in TAB
good model
imation of
e.
RDIZED SOLU
mber of mea
McQuitty,
mediocre m
BLE 24 is 0
l fit. It furthe
the popula
UTION FOR S
ChisquaX /d.fNFI=0.76CFI = 0.8RMSEA =GFI = 0.9
Analy
asurement
(2004), if
model fit and
.077 which
er proves th
ation and it
PI MODEL FI
are=88.29, (d
f = 2.32., P < 06 85 0.077 26
ysis & Finding
11
parameters
the value o
d if the valu
is below th
hat the valu
t will not b
IT.7
d.f)=38
0.001
gs
13
s.
of
ue
he
ue
be
Chapter 6 Analysis & Findings
114
The detail of the significant quality constructs evolved through the SEM
analysis and quality improvement indicators is given in TABLE 25
TABLE 25 EVOLVED SPI PARADIGM PRACTICES
CONSTRUCTS LQIM VARIABLE INDICATORS
Org. Size & Structure (OSS) TND Technology / Research & Dev. Org. Culture (OCL) COM Communication TRA Training TST Team Structure Org. Behaviour towards Quality (OBQ) TR Turnover Rate SQI Software Quality Improvement QA Quality Assurance TPQ Top Management Quality Approach Measurement & Analysis (MAN) KBL Knowledge Base Library DL Defect log
The item wise deletion of Questionnaire items (indicators) from the
model during Structural Equation Modelling (SEM) is given in table 26.
TABLE 26 SEM DELETED ITEMS FROM MODEL
No VAR QUESTIONS
1 RTN Employees are hired on long term basis and organization tries to retain employees
2 SPI The organization has dedicated (quality Assurance) QA group for tailoring And improving standard processes.
3 CPI CPI (Continues Process Improvement) is practiced dedicatedly and documented to improve quality.
4 SCM The organization has established a shared knowledge base library for configuration management, which can accurately reconstructing software items from scratch in development environment.
5 PI After process analysis and measurement, top management is quick at devising corrective actions for process improvement.
6 TQM Top management while investing in quality and process improvement considered it as a cost burden.
7 ASS Personnel performance assessment just documentary and everyone is given same bonus?
8 CS There is some gap between customer understanding and project team’s perception of customer requirement.
9 PM All variations in baseline of project plan implementation due to changes, quality, schedule and delays are well reflected in the updated project plans
Chapter 6 Analysis & Findings
115
10 CM Customer requirement changes that occur during development are incorporated through Change control Board..
11 OSR Organizational documented subcontract management procedure is available for out Sourcing projects to external firms.
12 CSTD Organization has common coding standards for all projects and all employees are trained.
13 DCS Data Collection System (DCS) has dedicated resources are available and DCS is in use for collecting process measurement data on continuous basis
14 STA Organization has hired data analysis experts like MSc / PhD statistics for organizational and quality performance analysis.
15 PT Project plans are tracked and different versions are maintained due to functionality changes.
16 SIZ Your organization have number of employees from
17 TM Task allocation and time management is strictly followed.
18 SHD Usually tasks are completed within working hours planned for the task: how much you agree?
19 RA Your organization can accept projects even if required resources are insufficient to complete the project.
20 CNF Software errors issues arise frequently after the project completion and handover.
21 AR Management conducts periodic quality audits and reviews for all stages in a project continuously for all projects.
22 TW In case of project delay (failure), Project performance Assessment and responsibility is emphasized on individual basis rather than team based responsibility
23 PMR Organization defined and documented procedure for measuring process performance is practiced?
24 LRN Top management and colleagues willingly sponsor learning to other employees usually.
25 EST A documented procedure is used for project cost, effort and size estimation.
26 ML The maturity level of your organization with respect to quality is somewhere at.
27 QPL Top management establishes plans for quality improvement activities and continuously gives a follow up.
28 OCM The organization has committed funds, staff and other resources for quality process development and process assessment.
29 RP Top management is equally willing to employ dedicated staff for quality control and process Improvement.
30 OST There are more than 3 levels of management hierarchy in the organization
31 PER Cost Performance Index (CPI) or Schedule Performance Index (SPI) or project earned value tracking are periodically calculated.
32 PR A documented review process exists at each stage to transfer project from one stage to another like Sign‐in, requirement , design, coding and to testing .
33 UAT Project test cases are prepared at design stage before implementation of design.
34 PPL Project development plan with resource allocation (PDP) is prepared and strictly followed throughout the SDLC.
35 RM Project risk assessment and mitigation is thoroughly documented and evaluated before start of each project.
36 PPR Process efficiency and effectiveness is measured individually for each process to optimize process performance
37 OVT Organization practices include Overtime hours offered, appreciated and paid for.
Chapter 6 Analysis & Findings
116
6.7 LEAN QUALITY IMPROVEMENT MODEL CONCEPTUAL DETAIL
The conceptual detail of Lean Quality Improvement Model (LQIM) that evolved
through SEM analysis is given in TABLE 27. This conceptual detail is based on
Deming’s TQM philosophy of Plan. Do. Check, Act, PDCA and Software Process
Improvement (SPI) guidelines from literature review.
TABLE 27 LQIM CONCEPTUAL DETAIL
MODEL VARIABLE INDICATOR DESCRIPTION
PLAN
Organization Size and Structure (OSS)
TND Tech. R n’ D
The research performed on existing and new applications, processes and hardware can result in new product development or up‐gradation of existing LQIM Process.
Organization Culture (OCL)
COM Communication
To make any changes in the culture of the organization, whether for the sake of quality or business development, freedom of speech and communication is to be established. Top Management is required to develop effective and easily available communication systems that can be used by all personnel involved so that they are able to express their opinion and also express their work progress (self‐monitoring) and problems / issues / violations (monitoring of others) through this system. Mode of communication can be email, phone, hand‐written, etc. but all communication is to be logged and recorded to ensure investigation of any problem and to rectify / correct any changes that are not garnering a positive response.
TRA Training
A system for training need assessments and skill development of personnel is to be established to increase the adaption rate of culture changes and personnel capabilities. Culture change refers to changes in processes and normal day‐to‐day tasks due to implementation of LQIM. Culture change is a holistic approach based on long term planning. Short internal awareness sessions can help people increase their confidence while keeping everything under budget.
TST Team Structure
Cross sectional teams based on domain experts from all departments involved in LQIM development and implementation need to be established in order to ensure that all departments are part of the implementation process and are able to comply with LQIM requirements.
Chapter 6 Analysis & Findings
117
Organization’s Behavior Towards Quality (OBQ)
TPQ
Top Management Quality Approach
Quality requires resources and investment from Top Management. It is the Management’s decision whether they want to provide all resources or limit them based on their budgeting. Timeline provided is also at the hands of Top Management. In the end, it is their decision about “Timeline” and “Resources” that will eventually support in achievement of quality as planned. Limited resources and short timelines may not provide the intended results, but long timelines and unlimited resources are also not the right solution. Top Management’s priority towards quality is the key.
TR Turnover Rate
A high turnover rate (more than 6% person leaving the organization annually) expresses insecurity of employees and can weaken the base of LQIM. Organization needs to motivate their personnel through incentives (bonuses and titles) and trainings that can support them not only in operations but also in their career development.
QA Quality Assurance
Plan must include methods for quality assurance so that products and services are under constant check while they are being processed, therefore leaving little chances for non‐conformities. Cost of Quality Assurance may be high, but through adopting TQM approach with long term benefits and almost no re‐work, makes up to the investment in this category.
DO
Organization Culture (OCL) TRA Training
Perform the trainings based on the requirements of LQIM as well as focused on personnel preferences and on the basis of training need assessment. Develop a learning culture through training and re‐training.
Organization’s Behavior Towards Quality (OBQ)
QA Quality Assurance
Quality Assurance goes side by side with product development processes in IT companies. This system is required to ensure that the required product and services are meeting customer requirements. Management’s approach is to gain customer’s loyalty and long term relationship.
CHECK
Measurement and Analysis (MAN)
KBL Knowledge Base Library
Organization is required to record all the issues and problems faced during a working year, and also record their rectifications and methods to avoid such issues / problems. This knowledge is distributed to all levels of organization so that personnel are able to learn from their past mistakes and avoid any problems that may have occurred previously.
DL Defect Log Defects are recorded separately so that they can be statistically analyzed to identify any trends
Chapter 6 Analysis & Findings
118
6.8 SUMMARY
In this chapter empirical analysis and its findings are presented. The empirical
analysis includes correlation and regression analysis of quality constructs.
Structural Equation Modelling (SEM) technique is used to develop an
optimized Lean Quality Improvement Model (LQIM) for standard quality
practices in SMSH. Eight quality constructs were developed to ascertain the
level of current quality practices and evolve a LQIM. In correlation analysis all
seven independent constructs were found significant towards the dependent
variable Quality Improvement. Regression analysis revealed that only four of
these independent quality constructs contributed significantly towards the
dependant variable Quality Improvement. Through Structural Equation
Modelling (SEM) the LQIM was evolved. This model presented four quality
constructs and ten of their respective quality practices as significant. At the
end the conceptual understanding of the LQIM model is presented for
implementation in SMSH using Deming’s TQM Philosophy of PDCA. The next
chapter provides recommendations for implementing quality practices.
or determine alternates to avoid / eradicate the issues.
ACT
Organization’s Behaviour Towards Quality (OBQ)
SQI Software Quality Improvement
The findings based on Quality Assurance Activities, Logs, reviews, internal audits and statistical analysis (using various tools), need to be corrected with most appropriate corrective measure. This can also help in determining better solutions to enhance system performance.
Chapter 7 Recommendations
119
CHAPTER 7 RECOMMENDATIONS
This chapter includes the set of recommendations given on the
following basis. Literature review on quality models and SME culture for small
and medium size software houses according to research questions 1 and 2;
Descriptive Analysis findings that gives answer to the problems faced by
SMSHs according to research question 3 and analysis & discussions to
measure the reliability and goodness of fit of the proposed SPI paradigm
according to research question 4.
7.1 QUESTION 1: HOW TO CHANGE ORGANIZATIONAL CULTURE IN SMSH
Many companies are not in favour of the cultural change. They resist
change as they had a fear of failure. They feel easy to the old environment.
They are unaware of the real meaning of implementing the quality in a
company. Only relying on training and certifications does not mean the
organization has achieved the quality level. According to Crouch (1998), “I do
wish I had more knowledge in areas such as identifying key business drivers
and processes as well as developing performance goals, measures and
standards”. As all you learn from the training and certification does not work
as it is too much academic to go by the book. It should be more flexible and
close to the real life examples.
But before implementing any strategy, standard and approach the
organization should develop clear understanding of processes and standards.
SMSHs shall realize the importance of the quality culture and bring the
change according to their respective business environment. Process
reliability, productivity, quality are usually measured by human perception in a
poor quality environment. Preferably, an effective way to measure process
performance and quality is to first understand the process at micro and macro
levels and then use statistical measurement techniques and automated
software tools for data analysis (Siok and Tian, 2007). Make the practitioners
and leadership believe that yes there is a need of the radical cultural change
Chapter 7 Recommendations
120
in the organization for the improvement of business performance. The local
SMSH still fail to gain the latent benefit of cost of quality concept that it is an
investment and not an over head cost (Deming, 1986). SMSH need to
reengineer their old processes according to the new upgraded technologies
and need to automate their quality management systems. In early 90s a
number of large scale change management projects failed for the reasons
related to technology up gradation as the managers continued to rely on the
old processes (Markus, and Keil, 1994). So there is a need of not only to bring
the change but the essential part is to reengineer their old processes and to
create process alignment with technology and organization culture. As a result
performance efficiency will increase due to SPI and reduction in rework costs.
Improvement in quality will achieve economies of scale and thus local
software products will be more competitive in global markets. That’s the
reason why SMSH should implement TQM principles along with ISO 9000 or
CMM /CMMI for continuous improvement of the quality.
To bring the change in the culture of our local software industry through
TQM a transition to total quality culture is required and behaviour towards
quality needs to be changed among the local IT practitioners. Top
management should delegate powers to empower employees to take
appropriate actions when the things go wrong or to take preventive action
before they go wrong instead of inspection and fire fighting afterwards. There
should be an open communication between the employees at all levels,
instead of having weak communication pattern based on the grapevine and
secrecy. As advised by Deming, (1986) break communication barriers and
learn from mistakes instead of hiding them or finger pointing on others. He
advised to adopt learning culture through training and retraining to develop
awareness in quality. Promote freedom of speech and open channels of
communication with internal and external customers. It’s top management’s
responsibility to create vision and lead the organization to success with
excellence in quality management (Anderson et al.,1994).
Chapter 7 Recommendations
121
7.2 RECOMMENDATIONS ON FINDINGS SMSH
All the responses from the IT practitioners were tabulated for
descriptive analysis and results. Based on the feedback, comments from the
respondents and in the light of literature review of quality models, quality and
quality culture in SME following recommendations are proposed.
7.2.1 ORGANIZATION SIZE & STRUCTURE
Employees should be hired on basis of available resources of the
organization. An organization needs to have at least one technology specialist
for each set of process areas and technology based departments because all
rounder cannot perform technical and trivial tasks in making high quality
products. A leader must be qualified, skilled and must have guidance and
leadership abilities in right directions. There should be a separate department
for Research and Development. The central and tall structure is not
appreciated because it can cause delays in decision making during project
development phases and also during quality improvement. There should be
criteria for employee harmony based upon their performance and assessment
and some beneficial services for employees. In this way the employees can
work with motivation and with their heart and soul which is a key point for
quality culture and a good environment. It also reduces turnover rate. SMEs
should make favourable human resource policies to retain employees to
prevent knowledge and brain drain due to high turnover rate.
7.2.2. ORGANIZATION CULTURE
Time management and time scheduling should be arranged according
to the project deadlines. Before committing and signing agreement with
customer regarding project cost and completion deadlines, top management
should consult project technical team leads. Offering overtime hours is not a
good approach as it signifies lack of time planning and time management.
There should be appropriate pay scales for IT professionals and working
overtime hours to be discouraged. Top management should take interest in
time scheduling and time planning as it will increase project success rate.
Chapter 7 Recommendations
122
Freedom of speech is one of Deming’s 14 points and free communication
across organization should be encouraged.
Top management should promote learning culture by encouraging
employees as well as for management to learn on continuous basis. Training
about new and upcoming technologies is necessary for the organization to
compete in local and global markets. A specific budget should be allocated for
such trainings.
The assessment results lead to quality and performance improvement.
An organization should practice 360 degree assessment technique for
employee assessment and should reward people on good performance which
will also motivate remaining employees to work hard and perform better.
7.2.3. ORGANIZATION BEHAVIOUR TOWARDS QUALITY
An organization should take only those projects for which required
technical resources are sufficient and currently available. Resource
optimization and workload management helps to achieve high quality at low
cost. Tests are conducted to remove errors and non conformances of
software product. Quality can never be achieved by increased number of
tests. Top management should adopt preventive approach rather than
corrective approach toward testing and quality improvement. There should be
a separate team or department for quality control and quality assurance with
qualified and trained senior level quality assurance professionals.
7.2.4. REQUIREMENT DEVELOPMENT & MANAGEMENT
Requirement development needs adequate time schedules to avoid
from rework. The accuracy of time planned for requirement analysis and
commitment should be done on realistic timelines and not just estimates that
make the management happy. Implementation of estimation tools and
techniques should be encouraged. The accurate time scheduling for
requirement development and management would also take less effort for
requirement change management and configuration management. It is difficult
to reach total customer satisfaction but if a documented procedure is followed
Chapter 7 Recommendations
123
for change management and software configuration management is practiced
as a tool, then most of the disputes and disagreements with the customer can
be settled.
Top management should be involved in time scheduling. There should
be frequent and close communication with the customer and all changes
should go through change control board. The requirement specification should
be made by the involvement of customer to achieve completeness. The
orientation of the SMEs should be towards total customer satisfaction. There
should be a change management agreement for requirement change to avoid
rework. Companies that are involved in offshore projects should hire legal
firms to develop agreements and policies to subcontract projects..
7.2.5. RECOMMENDATIONS: PROJECT PLANNING
Top management should develop comprehensive quality improvement
approach towards prevention of errors because non conformance after
handing over not only has high change management over heads, rework
costs and warranty claims, but it also maligns the repute of organization. User
Acceptance Tests (UAT) should be developed at design stage as required by
standard quality practices and software engineering models governing SDLC.
A given software development life cycle helps in defining a concrete way of
development and also prevents from rework. Developing test cases during
development lead to poor quality assurance practices. The project quality and
delivery depends upon schedule planning through Project Development Plan
(PDP) and Quality Management System. The variance in schedule planning
may affect timelines and project cost. There should be strict schedule
planning to prevent from rework and variance. Project resources should be
planned before project start to avoid from inconvenience from available
resources. Risk management is an important component of project
management (PMbok.)8 best practices, and risk management monitoring and
8 Project Management Book of Knowledge, Project Management Institute (PMI).
Chapter 7 Recommendations
124
mitigation plans should be made part of SDLC. Top management should
practice proper effort estimation and resource utilization techniques to give
equal importance to all the projects regardless of size or profitability to ensure
uniform quality output. Management should follow a standard policy for
assigning project estimation deadlines in consultation with technical team
leads.
Common coding standards should be established and shared through
knowledge base library. Orientation training should include training on Coding
standards. By implementing coding standards software code becomes more
readable and understandable and it helps the developers during change
management and maintenance.
7.2.6. MONITORING AND CONTROL
The management should put emphasis on project monitoring and
process assessment side by side during project work by using standard
procedures because only functionality reviews and audits do not assure the
quality without appropriate quality control procedures. The management
should have ability to reorganize their effected plans and should allocate
optimal resources to timely accommodate changes without disturbing the
schedule of PDP. The management should introduce a trend of using metrics
at project measurement level for monitoring and tracking to get accurate
figures of process and project performance. Capability Process Index CPI for
all critical processes should be regularly monitored and tracked. It will help to
not only enhance the performance of individual processes but will also
improve the synergy and alignment between the processes. Team culture and
spirit of team work should be promoted among team members. It will create
stronger bonding and more communication among team members.
Chapter 7 Recommendations
125
7.2.7. MEASUREMENT AND ANALYSIS
It is up to management to inculcate the performance measurement
culture in the organization. Secondly management should involve employees
in setting measurement goals.
The basic system and identifiers for process measurement and
process improvement should be developed for top managers and employees.
Mostly measures are not specified but if measures are specified, then these
measures should be linked with smart goals and objectives. Data collection
and storage procedures are not properly defined. In this undeveloped
environment, process measurement and process improvement is a big
challenge for SMSH. These software organizations need to have highly
qualified managers that are adequate to inculcate the measurement culture in
software houses and who should be able to develop a mechanism for data
collection System (DCS). Process efficiency and effectiveness is an important
measure to calculate process alignment and over all process synergy. As a
guideline top management has to assure that all processes are aligned
together and are working in synergy. Management should develop template
for post-mortem summary which should highlight lessons learned through
failure and should also list down mistakes which should not be repeated
again. Overall performance measurement is very weak area and effort should
be made to improve quality through performance measurement.
7.2.8. PROCESS QUALITY IMPROVEMENT
Management should adopt quality culture by initiating programs like
ISO/CMMI certification. IT should develop quality training programs to create
quality awareness among the employees. Management should allocate
separate budget for Quality improvement and should consider it as an
investment (Juran, 1984). Management should also invest in having qualified
QA resources for process tailoring and up-gradation. Processes and
procedures internally developed should be practiced in letter and spirit and
not just left alone in the files and folders. Process improvement is most
important process in order to achieve quality product. If processes are not
Chapter 7 Recommendations
126
improved frequently this means processes are not measured. Recent attitude
shows least concern towards process measurement. No check and control
towards quality because top management is interested towards quantity
rather than quality. If process is not measured, gap analysis cannot be
performed and hence processes improvement is not possible. But as it is
seen through following data organizations rarely improve its processes.
7.3. LQIM (PARADIGM) FOR LOCAL SMSH
In chapter 6, structural equation modelling (SEM) technique was used
to come up with an evolved LQIM to depict quality improvement paradigm and
standards practiced in the local software industry. LQIM is a tailored and
economized paradigm according to the practices and perceptions of the local
IT practitioners. The proposed LQIM is an indigenous model which when
improvised in accordance to the SMSHs cultural recommendations can
establish to be a fit model for SMSHs. The LQIM has already been ratified
according to generally accepted good fit indices in SEM analysis. It has been
established through literature review that to implement ISO 9000 SMSHs
should adopt TQM philosophy, long term planning and measurement culture
as such practices have proven to produce good results in the industry. The
main objective of the research was to propose a Lean Quality Improvement
model suitable for local software industry. This LQIM was derived through
SEM in the previous Chapter.
Implementation of Indigenous LQIM is proposed using the Deming’s
philosophy of “Plan DO Check Act”, PDCA Cycle for continuous process
improvement and is shown In
FIGURE 8 IMPLEMENTATION OF LQIM MODEL
The conceptual detail of the LQIM is given in TABLE 27. The LQIM Deployment plan mapped with PDCA is given in TABLE 28.
Chapter 7 Recommendations
127
TABLE 28
FIGURE 8 IMPLEMENTATION OF LQIM MODEL
7.3.1. LEAN QUALITY IMPROVEMENT MODEL DEPLOYMENT PLAN
The current model provides us with a Lean Quality Improvement Model
(LQIM) for a SMSH. The model does not include planning and monitoring of
projects being performed in an organization. Most of the software companies’
rely on their project management of software development, considering
project planning to be an important part of this activity; In a SME model
Chapter 7 Recommendations
128
development study in Finland Saastamoinen and Tukiainen, (2004) also
emphasized that planning and continual monitoring of quality activities as key
prerequisite for good quality products. The Finland study also covered Quality
practices like Planning, data collection, data validation, process reporting and
process measurement
Planning and Monitoring, its absence effect’s the prime objective of
quality achievement rather negatively. Moreover without any controls for
monitoring deployed, there is far less chance of timely identification of errors /
issues and their timely rectification. The model actually expresses use of
Quality Assurance at some stages for identification of issues, but without
planning various stages of project management and without their monitoring,
Quality Assurance will eventually lack a timely and planned response
therefore leading to rework and waste of resources. According to Allen,
Ramachandran and Abushama, (2003) in PRISMS study mentioned most
important Quality metrics like project tracking, monitoring and defect
detection. PRISMS study as well as literature review on SPI also emphasized
on automation of data collection activities to support planning and timely
decision making.
As indicated in LQIM deployment plan is mapped with Deming’s PDCA
cycle based on the conceptual understanding given in TABLE 27 and
guidelines given in the literature review.
Chapter 7 Recommendations
129
TABLE 28 LQIM DEPLOYMENT PLAN MAPPED WITH PDCA
PHASE I : REVIEW & GAP ANALYSIS (Performed Once Only) Company Wide LQIM Review & Gap AnalysisReview of Existing Processes, Policies and ProceduresIdentification of Gaps based on best management PracticesSubmission of Gap Analysis Report with recommendations and solutions TRAINING A: Introduction and Awareness on LQIMDELIVERABLE : LQIM GAP ANALYSIS & RECOMMENDATION REPORT PHASE II : SYSTEM DEVELOPMENT / REVISION / IMPROVEMENT Development of Quality Policy & Quality ObjectivesDevelopment of Process Flow Charts , Corporate Organization Chart and Departmental ChartsLQIM Development and Implementation Monitoring TeamDevelopment of Technology Development & Research and Development Department Development of Procedure for Corrective and Preventive Measures Development of HR Policies (to reduce turnover rate)Development of Knowledge base LibraryDevelopment of Procedure for control of non‐conformanceDevelopment of Procedure for Quality AssuranceDevelopment of Procedures for CommunicationDevelopment of Training and Awareness procedures and plans (Systematic Culture Change Acceptance)Development of Procedure for internal system auditing (development of Audit Checklist) TRAINING B: Documentation, Recording and Reporting based on LQIM requirements DELIVERABLE : LQIM POLICIES, PROCEDURES AND TEMPLATES PHASE III : IMPLEMENTATIONImplementation Team AssignmentImplementation of Records as per developed proceduresImplementation of Records & Awareness verification Relating to LQIM Correction of Documentation based on FeedbackImplementation Team AssignmentDELIVERABLES: LQIM MANUAL PHASE IV : AUDITING Selection of Internal AuditorsPlanning the Internal AuditCollection of Defect Logs and updating knowledge base library Company‐wide LQIM Internal AuditCorrective Preventive Actions/Audit Non‐Conformity ClosingPerform Trend Analysis on the basis of previous audit results TRAINING C: Internal Auditing TrainingDELIVERABLES:INTERNAL AUDIT REPORT
PLAN
DO
CHECK
ACT
Chapter 7 Recommendations
130
7.3.2. TQM SUGGESTIONS AND GUIDELINES In order to implement LQIM more effectively in the local software industry a set of
TQM implementation guidelines is presented which were developed during literature
review to fulfil the environmental and cultural requirements of proposed model.
Top management should assure total commitment towards optimum
resource allocation, training and CPI.
Planning should be made important part of all organizational and
project management activities continually on long and short term basis
Achievement of organizational goals through customer relationship
management (CRM ) and customer orientations.
Management to rethink quality as a way of doing business, and
resources spent on quality improvement should not be considered as a
cost but it is an investment which reaps higher profits in the long run.
Learning culture through training development programs to enhance
human resource performance skills, capabilities and quality awareness.
It will help to inculcate positive culture and work ethics.
Open Communication Channels across the organization for employees
to freely express ideas and share information and develop a sense of
team work within organization. (one man show to be discouraged). This
will help SMEs to capitalize on employee’s talent and innovativeness.
Selecting right people for the right job based on their academic
qualifications and skills; and selects the best cross-functional teams for
LQIM Development and Implementation Monitoring Team. Do not re-engineer on large scale, bring the change through small baby
steps ( wins), by forming a result oriented strategy and preventive
approach.
Management needs to provide enough resources for monitoring,
mitigation and management of assessed organizational risks.
Business process redesign should be carried out across the
organization to resolve the problems of resource contention which is
the biggest problem in SME. Business process redesign will reduce the
Chapter 7 Recommendations
131
synchronization delays in processes and hence improve overall
effectiveness and profitability of SME.
Performance incentives and rewards should be separate from regular
increments based on annual assessment in order to create value
driven employees.
To create Involvement and ownership, Top Management should share
development of annual quality objectives in consultation with
employees.
Improve the work environment and culture through benchmarking with
industrial leading best practices and develop opportunities through
innovation, change management and feedback from all stakeholders.
SPI activities should be linked with customer satisfaction and
organizational goals, and top management should prioritize to improve
key process areas accordingly.
All activities like process measurement, data collection and process
rating should be automated by SME as human perception is poor and
inefficient as compared to automated process measurement tools.
7.3.3. LIMITATIONS OF PROPOSED LQIM PARADIGM
As LQIM is Culture changing process, normally the level of acceptance
expressed by human resource is very low in the beginning.
Not all personnel can be part of Skill Development & Training
Programs due to limitation of resources
Lack of Consistency by Management in long term planning for system
improvement.
Customers may prefer (and enforce in some cases) their own process
improvement policies over LQIM Policies.
Departments get resources according to their priority in LQIM,
therefore some departments are ignored intentionally in the beginning
Based on observations from local IT industry culture, employee
empowerment to take decisions is discouraged by higher management
Chapter 7 Recommendations
132
Lack of proper inflation adjustment to annual increments which
eventually leads to higher turnover rate.
. All processes should be spiritually followed so that organization is
able to deliver quality software products.
. TQM culture is not found in these models therefore TQM should be
made part of SPI activities to reduce schedule delays, cost over runs,
and rework costs..
7.3.4. SUMMARY This chapter provides a set of recommendations based on literature
review and the empirical analysis of the research conducted. It also presents
an implementation and deployment model of the proposed LQIM. In the end
TQM guidelines are proposed for deploying LQIM based on Deming’s PDCA
cycle. The next chapter concludes the thesis.
Chapter 8 Conclusion
133
CHAPTER 8 CONCLUSION AND FUTURE WORK
The main rationale behind this reading is to develop an optimum Lean
Quality Improvement Model (LQIM) and a set of recommendations as
guidelines to Implement total quality culture and standard quality practices in
the local software industry. Indigenized LQIM is destined to give innovative
and flexible directions for SMSH to change their culture and improve their
processes economically by following TQM philosophy.
As a first step the study exposed the local IT industry’s behaviour
towards quality and its notion of quality through studying previous and
contemporary quality improvement practices in local SMSH. An exploratory
research effort in the domain of total quality management (TQM) and
Software Process Improvement (SPI) was conducted with the help of an
extensive literature review of major quality standards and models being
implemented in the local industry. The behaviour of international quality
standards was deliberated towards quality improvement culture and SMSH
practices. It’s cited in literature that Quality Culture plays an important role in
developing maturity, learning and improvement in the business performance
of an organization. Organizational quality culture groups people together with
an orientation to work towards achieving their common goals by being united.
The idea is to align all efforts towards achieving organizational set
performance goals by creating process synergy through TQM philosophy. The
literature review about quality, TQM, SMSH culture and quality improvement
is implicated to achieve a set of implementation guidelines for an indigenous
LQIM model. The results of the survey are analyzed and are reported to high
light the SMSH cultural and quality problems being faced to implement quality.
Structural Equation Modelling (SEM) technique was used to optimize
the theoretical structural framework and evolved an indigenized LQIM to
implement quality improvement paradigm and standard quality practices in the
local software industry. LQIM is a tailored and realistic paradigm according to
the needs and perceptions of the local IT practitioners.
Chapter 8 Conclusion
134
The proposed LQIM is an indigenous model which when improvised in
accordance to SMSH cultural and quality improvement recommendations, is
established to be a fit model for SMSH in the local industry. On Similar
footings implementation of Indigenous LQIM is recommended using the
Deming’s philosophy of PDCA Cycle for continuous process improvement. A
set of guidelines based on questionnaire results and literature review are also
proposed in order to improve the quality and culture of local SMSH.
LQIM and a set of quality improvement guidelines and practices
achieved as a paradigm through this research, is an economized and proven
(Good-Fit ) paradigm for implementation of true quality culture in local SMSH.
It is a first step towards rethinking of quality implementation based on TQM
philosophy and long term planning and measurement culture. Such practices
have proven to transform to quality culture and bring improvement in quality
processes and software quality culture, and above all produce optimal
business performance results in the industry.
In future this research can be extended to explore additional quality
dimensions which are recommended by Project Management Body of
Knowledge (PMBoK) and other quality models. This research can further be
replicated in other developing countries like Bangladesh, Nepal and Sri Lanka
to develop Lean Quality Improvement Model (LQIM) to cater the needs and
cultural requirements of SME in a respective developing country.
References
135
BIBLIOGRAPHY Addison-Wesley. (1995). Maturity Model for Software.
Ahire, S., Golhar, D., & Waller, M. (1996). Development and Validation of TQM Implementation Constructs. Decision Sciences , 27 (1), 23-56.
Akao, Y., & Hoshin, K. (1991). Policy Development for Successful TQM.
Allen, P., A., R., & Abushama, H. (2003). PRISMS: An Aprroach to Software Process Improvement for Small to Medium Enterprises. Third International Conference on Quality Software (QSIC"03). IEEE, Computer Society.
American Society for Quality. (2000). Quality MAnagement Systems, Fundamentals and Vocabulary ANSI/ISO/ASQ Q9000:2000. Milwaukee: ASQ.
Anderson, J. C., & Gerbing, D. W. (1988). Structural Equation Modeling in Practice: A Review and Recommended Two-Step Approach. Psychological Bulletin , 103 (3), pp. 411-423.
Anderson, J. C., Rungtusanatham, M., & Schroeder, R. G. (1994). A Theory of Quality Management Underlying the Deming Management Method. Academy of Management Review , 19 (3), pp. 472-509.
Arent, J., Iversen, J. H., Andersen, C. V., & Bang, S. (2000). Project Assessments: Supporting Commitment, Participation, and Learning in Software Process Improvement. 33rd Hawaii International Conference on System Sciences. Hawaii.
Arthor. (1991). Does Investment in IT Pay off. Computer World Magazine .
Ashrafi. (2003). The Impact of Software Process Improvement on Quality: In Theory and Practice. Information & Management , 40 (7), 677-690.
Bacon, L. D. (1997). Using AMOS for structural equation modeling in market research. Lynd Bacon & Associates, Ltd. And SPSS Inc.
Bacon, L. D. (1997). Using AMOS for Structural Equation Modeling in Market Research. Lynd Bacon & Associates, Ltd. And SPSS Inc.
Baskerville, R., & Heje, J. P. (1999). Knowledge Capability and Maturity in Software Management.
Basu, R. A., & Wright, N. (2003). Quality Beyond Six Sigma. Elsvier Ltd., LinacreHouse, Jordan Hill, Oxford OX2 8DP.
Bentler, P. M. (1990). Comparative Fit Index in Structural Models. Psychological Bulletin , 107 (2), 238-246.
Bentler, P. M., & Bonnet, D. C. (1992). Significance Tests and Goodness of FIT in the Analysis of Covariance Structure. Psychological Bulletin , 88 (3), 588-606.
Beth, M. C., Konrad, M., Shrum, S., Smith, K., & Wemyss, G. (2004). Software Engineering Process Management. Interpretive Guidance Project: What We Learned. Pittsburgh: Carnegie Mellon.
Biberoglu, E. a. (2002). A Survey of Industrial Experiences with CMM and the Teaching of CMM Practices. Southeastern Conference Proceedings, CCSC. Vol. 18, no.2.
Biberoglu, E., & Haddad, H. (2002). A Survey of Industrial Experiences with CMM and the Teaching of CMM Practices. Conference Proceedings JCSC, 18.
Black, S., & Porter, L. (1996). Identification of Critical factors of TQM,Decision Sciences. , 27 (1), 1-21.
References
136
Boldyreff, C., Newman, J., & Taramaa, J. (1996). Managing process improvement in virtual software corporations. Enabling technologies: Infrastructure for collaborative enterprises. proceedings of the 5th workshop on, (p. 142).
Bollen, K. (1989). Structural equations with latent variables. In Bollen, Structural equations with latent variables. NewYork: Wiley.
Bollen, K. A. (1989). Structural Equations with Latent variables. In Structural Equations with Latent variables. New York: Wiley.
Bradford, K. C. (1997). The Effects of Software Process Maturity on Software Development Effort, PhD Dissertation. Computer Science Department. University of Southern California.
Brodman, J. G., & Johnson, G. L. (1994). What small business and small organizations say about the CMM. 16th international conference on Software engineering (pp. 331 - 340 ). Italy: IEEE Computer Society Press .
Byrne, B. M. (1989). A Primer of LISREL: Basic Applications and Programming for Confirmatory Factor Analytic Models. New York: Springer-Verlag.
Byrne, B. M. (1998). Structural Equation Modeling with LISREL, PRELIS and SIMPLIS: Basic Concepts and Applications in Programming. New Jersey: Lawrence Erlbaum Associates.
Carmines, E. G., & Zeller, R. A. (1979). Reliability and Validity Assessment. Beverley Hills: Sage Publications.
Cattaneo, F., Fuggetta, A., & Lavazza, L. (1995). An Experience in process Assessment. ACM.
CAVE, M., HANNEY, S., KOGAN, M., & TREVETT, G. (1988). The Use of Performance Indicators in Higher Education. London: Jessica Kingsley.
Charles, A. S. (2004). Classification and Taxonomy of TEISMEs.
Claude, Y., Alain, A., & Renault, A. (2006). Applying ISO/IEC Software Engineering Standards in Small Settings: Historical Perspectives and Initial Achievements. SPICE 2006. Luxembourg.
CMMI-Team. (2006). CMMI ® for Development, Improving Processes for better products. (1.2), 4.
CMU/SEI. (2006). Retrieved from www.sei.cmu.edu/cmmi
CMU-SEI, (2010). Software Engineering Institute, Carnegie Mellon University. (n.d.). Retrieved 07 10, 2010, from www.smu.sei.edu: www.smu.sei.edu
Cohen, I. M., & Morrison, K. (2000). Research Methods in Education (5 ed.). London and New York.
Cronbach, L. (1951). Coefficient Alpha and the Internal Structure of Test, Psychometrica. 16 no.3, 297-300.
Crosby, P. (1989). Let’s Talk Quality: 96 Questions You. Always Wanted To Ask . McGraw-Hill.
Crosby, P. (1979). Quality if Free: The Art of Making Quality Certain. MC Graw-Hill. New York.
Crouch, J. (1998). Essential Tools for Quality Managers.
Damij, N., Damij, T., Grad, J., & Jelenc, F. (2008). A methodology for business process improvement and IS development. Information and Software Technology . 50 (11), 1127-1141 .
Davis, N., & Mullaney, J. (2003). The Team Software Process (TSPSM) in Practice: A Summary of Recent Results.
References
137
Deming, W. (1986). Out of the Crisis. Massachusetts Institute of Technology, Center for Advanced Engineering Study: Cambridge, M.A.
Demiriirs, E., & et.al. (1998). Process Improvement Towards ISO 9001 Certification in Small Software Organizations. Proceeding of 20th International Conference on Software Engineering. Washington, DC., USA: IEEE Computer Society Press.
Dilanthi, A., David, B., Narjan, S., & Rita, N. (2002). Quantitative and Qualitative Research in the Built Environment: Application of Mixed Research Approach, Work Study. 51 (1), 17-31.
ELPHINSTONE, L. (1990). The development of the Course Experience Questionnaire.
Emam, K. E., & Jung, H. (2001). An Empirical Study of ISO/IEC 15504 Assessment Model. 23-41.
Feigenbaum, A. V. (1991). Total Quality Control. (T. Edition, Ed.) McGraw-Hill, Inc., New York.
Fenghueih, H., Ching, H., & Cleve, C. (1999). A Study of ISO 9000 Process, Motivation and Performance. 10 (7), 1009-1025.
Ferguson, P., & Kitson, J. (1997). CMM-Based Process Improvement Supplemented by the Personal Software Process in a Small Company Environment. Proceedings of the Software Engineering Process Group Conference, (pp. 17-20). San Jose, CA.
Ferreira, A., Santos, G., Cerqueira, R., Montoni, M., Barreto, A., Barreto, A. O., et al. (2006). Improving Processes in Small Settings.
Flynn, B. B., Schroeder, R. G., & Sakakibara, S. (1994). Framework for Quality Management Research and Associated Measurement Instrument. Journal of Operations Management , 11, 339-336.
Garcia, S. (2003). Case Studies and Presentation. Retrieved from Software Engineering Institute: http://www.sei.cmu.edu/cmmi/casestudies/presentations/
Garcia, S., & Greattinger, C. (2006). Proceedings of the First International Research Workshop for Process Improvement in Small Settings.
Grasso, M. The Personal Software Process (PSP), A Software Engineering Seminar.
Grasso, M. (2005). The Personal Software Process. University of Maryland, Baltimore County, U.S.A.
Grunbacher, P. (1997). A Software Assessment Process for Small Software Enterprises. Proceedings of EUROMICRO 97.IEEE Computer Society Press. Budapest, Hungary.
Gulbro, R., Shonesy, L., & Dreyfus, P. (2000). Are Small Manufacturers Failing the Quality Test, Industrial Management and Data Systems. (2), 76-80.
Habra, Naji, & Alexandre, S. (2007). Initiating Software Process Improvement in Very Small Enterprises.
Hanna, O. A Software Process Model for Small Enterprises. Proceedings of First MoProSoft, (p. 93).
Hayes, W., & Over, J. W. (1997). An Empirical Study of the Impact of PSP on Individual Engineers. Software Engineering Institute, Carnegie Mellon University.
Herbsleb, J., Zubrow, D., Goldenson, D., Hayes, W., & Paulk, M. (1997). Software Quality and the Capability Maturity Model, Communications of the ACM. 40 (6), 30-40.
Hofstede, G. (2001). Culture’s Consequences:Comparing Values, Behaviors, Institutions and Organizations across Nations. Thousand Oaks, CA: Sage Publications.
Hu, L., & Bentler, P. M. (1999). Cutoff Criteria for Fit Indexes in Covariance Structure Analysis: Conventional Criteria Versus New Alternatives, Structural Equation Modeling. 6 (1), 1-55.
References
138
Huarng, F., & Chen, Y. (2002). Relationship of TQM Philosophy, Methods and Performance: A Survey in Taiwan. Industrial Management and Data Systems. 4, 226-234.
Humphrey, W. S. (2000). Retrieved from SEI: http://www.sei.cmu.edu/reports/00tr023.pdf
Humphrey, W. S. (1989). Managing the Software Process. Addison Wesley, Reading, MA.
Huq, Z. (2005). Managing Change: A Barrier to TQM Implementation in Service Industries, Managing Service Quality. 15 (5), 452-469.
Hyde, K., & Wilson, D. (2004). Intangible Benefits of CMM-based Software Process Improvement. Software Process Improvement and Practice , 9 (4), 217-228.
Ince, D. (1994). An Introduction to Software Quality Assurance and its Implementation. McGraw-Hill.
Introduction and Support Package: Guidance on the Documentation Requirements of Document: ISO/TC 176/SC 2/N525R2. International Organization for Standardization ISO .
Irigoyen A., F. F. (2007). Applying ISO 9001:2000, MPS.BR and CMMI to Achieve Software Process Maturity. 29 International Conference in Software Engineering. Minneapolis, MN, USA.
ISO 9001-2008 (E). (n.d.). Retrieved 05 25, 2010, from International Organization for standardization: www.iso.org
ISO/IEC 15504. (1996). Information Technology - Software Process Assessment -Part 5: An assessment Model and Indicator guidance. Geneva: ISO Press.
(2000). IT Policy and Action Plan. IT & Telecommunication Division. Ministry of Science and Technology.
Iwan, A. (1998). Sample Size and Sample Design for Nutritional Research. Department of Biostatisticss, School of Public Health, University of Indonesia, , Kampus FKM-UI, Depok 16424, Indonesia.
Jalote, P. (1999). CMM in Practice: Processes for Executing Software Projects at Infosys;The SEI Series in Software Engineering. Addison-Wesley.
Jansen, S., & Brinkkemper, S. Evaluating the Release, Delivery, and Deployment Processes of Eight Large Product Software Vendors applying the Customer Configuration Update Model. International Conference on Software Engineering Proceedings of the International Workshop on Workshop on Interdisciplinary Software Engineering Research, (pp. 65-68). NY, USA.
Jansma, P. (2005). When management gets serious about managing software.
Jung, H. (2005). Evaluating the Ordering of the SPICE Capability Levels. An Empirical Study. Information and Software Technology , 47 (3), 141-149.
Juran, J. (1993). A Renaissance in Quality. USA: Harvard Business.
Juran, J. M. (2004). Architect of Quality: The Autobiography of Dr. Joseph M. Juran. New York: McGraw-Hill.
Juran, J. M. (1988). Juran on Planning for Quality. New York: The Free Press.
Juran, J. M. (1988). Managing for Quality. The Journal of Quality and Participation (11), 8-12.
Juran, J., & Gryna, F. (1988). Quality Control Handbook (4 ed.). New York: McGraw-Hill.
K.K., W. (2009). Service Quality Index; A study on Malsian Banks. Contemporary Management Research , 5 (2).
Kanji, G. K. (1998). Measurement of Business Excellence. Total Quality Management , 9 (7), 633-643.
References
139
Kelly, G. Barriers to Adoption of the CMMI Process Model in Small Settings. First International Workshop on Process Improvement, (p. 18).
Kishore, S., & Naik, R. (2003). ISO 9001:2000 for Software Organizations. New Dehli: McGrawHill.
Krasner, H. (1994). The Payoff for SPI: What it is and How to Get it. (1) , 1-6. IEEE Computer Society.
Lakha, S. (1994). The new International Devision of Labour and the Indian Software Industry. Modern Asian Studie , 28 (2), 381-408.
Laporte, C. Y., & April, A. (2006). Applying Software Engineering Standards in Small Settings: Recent Historical Perspectives and Initial Achievements. First International Workshop on Process Improvement, (p. 39).
Lawlis, P. K., Flowe, R. M., & Thordahl, J. B. (1996). A Correlational Study of the CMM and Software Development Performance. Journal of Defense Software Engineering , 8 (9), 21-25.
Lee, H., & Ahn, S. (2008). Assessment of process improvement from organizational change. Information & Management , 45 (5), 270-280.
Lee, M. C. (2006). Applying TQM, CMM and ISO 9001 in Knowledge Management for Software Development Process Improvement. Int. J. Services and Standards , 2 (1).
Lincoln, Y. S., & Guba, E. G. (1985). Naturalistic inquiry. Beverly Hills, California: Sage Publications.
Liu, X., & Sun, Y. (2005). QFD Application in Software Process Management and Improvement Based on CMM. International Conference on Software Engineering. NY, USA.
Lwanga, S. K., & Manual, L. (1991). Sample Size Determination in Health Studies.
MacCallum R. C., B. M. (1996). Power Analysis and Determination of Sample Size for Covariance Structure Modeling. Psychological Methods , 1 (2), 130-149.
Mahour, M. (2006). The Effect of Quality Management Practices on Operational and Business Results in the Petroleum Industry in Iran. USA: PhD Research Report, University of Nebraska.
Mark C. Paulk, B. C. (1993). Capability Maturity Model for Software. Software Engineering Institute, Pittsburgh.
Markus, M. L. (1994). If We Build, They Will Come. Designing Information Systems That Users Want to Use, Sloan Management Review.
McAndrews, D. R. (2000). The Team Software Process (TSP): An Overview and Preliminary Results of Using Disciplined Practices. Software Engineering Institute, Carnegie Mellon University.
McGuire, E. G., & McKeown, K. A. (2001). 5 Critical Steps for Adopting CMM in an IS0 Environment. PICMET '01 InternationalConference, 1, pp. 430-431. Poland, OR, USA.
McHale, J. &. (2005). Software Engineering Institue. Carnege Melton University.
McHale, J. C. (2008). Accelerating CMMI Adoption with PSP/TSP. Carnegie Mellon University.
McKeown, K. A. Critical Steps for Adopting CMM in an IS0 Environment. New York: 'American University, Washington, D.C. 'Keane, Inc.
McQuitty, S. (2004). Statistical Power and Structural Equation Models in Business Research. Journal of Business Research , 57 (2), 175-183.
Mikel, H. (1988). The Nature of six sigma quality. Motorola University Press.
References
140
Miluk, G. (2006). Results of a Field Study of CMMI for Small Settings Using Rapid Applied Ethnography., (p. 27).
Mitra, A. (2005). Introduction Quality Control and the Total Quality System. In Fundamentals of Quality Control and Improvement (2 ed.).
Mondragon, O. (2006). Addressing Infrastructure Issues in Very Small Settings., (p. 5).
Mondragon, O. (2006). Addressing Infrastructure Issues in Very Small Settings., (p. 5).
Montazemi, A. (2006). How They Manage IT. SMES in Canada and the U.S.
MoST, (2000). IT Policy and Action Plan. IT & Telecommunication Division, Ministry of Science & Technology, government of Pakistan, Islamabad. . (n.d.). Retrieved 08 18, 2000
Murphy, K. a. (1989). Rate Error and Rating Accuracy. Journal of Applied Psychology , 71 (4), 619-624.
(2000). National IT Policy. Government of Pakistan, Islamabad.: Ministry of Science & Technology (MOS&T). .
NCEAC, (2010). National Computing Education Accreditation Council. (n.d.). Retrieved 07 25, 2010, from www.nceac.org: www.nceac.org
Nie, N. B. (1975). SPSS: Statistical Package for the Social Sciences. New York: McGrawHill.
Nunnally, J. (1978). Psychometric Theory. London: Mc Graw-Hill.
Oktaba, G. F. (2007). Software Process Improvement: The Competisoft Project, Computer. 40 (10), 21-28.
Oktaba, H. (2006). A Software Process Model for Small Enterprises. MoProSoft®, (p. 93).
Osama, A. (2005). Pakistan’s Software Industry Best Practices & Strategy Challenges. an Exploratory Analysis. Ministry of Information Technology Government of Pakistan. Islamabad.
PASHA, (2010). Pakistan Software House Association. . (n.d.). Retrieved 05 20, 2010, from www.pasha.org.pk: www.pasha.org.pk
Patton, M. (2002). Qualitative Research and Evaluation Methods (3 ed.). California: Sage Publications.
Paulk, M. C. (1994). A Comparison of ISO 9001 and the Capability Maturity Model for Software. Pittsburgh, Pennsylvania.
Paulk, M. C. (1995). How ISO 9001 Compares with the CMM. Feature. IEEE Computer Software Society.
Paulk, M. C., Curtis, B., & Chrissis, M. B. (1991). Capability Maturity Model for Software. Software Engineering Institute.
Paulk, M. (1998). Using the Software CMM in Small Organizations.
Paulk, M., Curtis, B., Chrissis, M., & Weber, C. (1993). Capability Maturity Model for Software, Version 1.1. Software Engineering Institute, Car-negie Mellon University, Pitsburgh PA.
PCST, (2010). Pakistan Commission of Science and Technology. (n.d.). Retrieved 06 20, 2010, from www.pcst.org.pk : www.pcst.org.pk
Pedreira, O., Piattini, M., & et, a. (2007). A Systematic Review of Software Process Tailoring. ACM SIGSOFT Software Engineering Notes , 32 (3).
References
141
Pennington, R. G. (2003). Change Performance to Change the Culture, Industrial and Commercial Training. 35 (6), 251-255.
Philips, M. (2001). CMMI Overview V1.1. CMMI Technology Conference. Pittsburgh: Software Enineering Institute, Carnegie Mellon University.
PITB, (2010). Pakistan Information Technology Board. (n.d.). Retrieved 06 20, 2010, from www.pitb.gov.pk: www.pitb.gov.pk
PNAC. (2010). Pakistan National Accreditation Council. Retrieved March 15, 2010, from www.pnac.org.pk
Pomeroy-Huff, M., Cannon, R., Chick, T. A., Mullane, J., & Nichols, W. (2009). The Personal Software Process SM. SEI.
Pomeroy-Huff, M., Mullaney, J., & et, a. (2005). The Personal Software Process (PSPSM) Body of Knowledge (Vol. 1.0).
Porter, L. J., & Parker, A. J. (1993). TQM- The Critical Success Factors. Total Quality Management & Business Excellence. , 4 (1), 13-22.
Pracchia, L. (2004, Jan). The AV-8B Team Learns Synergy of EVM and TSP Accelerates Software Process Improvement. pp. 20-22.
Praxiom Research Group, L. (2010). Retrieved from www.praxiom.com.
PTB. (2010). Pakistan Technology Board. Retrieved Machr 15, 2010, from Pakistan Technology Board Ministry of Science and Technology: www.ptbMoST.gov.pk
Quality Management and Quality Assurance Standard-part 3. International Organization for Standardization ISO , Geneva.
(2008). Quality Management Systems-Requirements. . International Standard Organisation ISO , Geneva.
Quazi H.A. & R. Padibjo Quazi, S. (1998). A journey toward total quality management through ISO 9000 certification. International Journal of Quality &Reliability , 15, 489-508.
Rad, A. .. Survey of Total Quality Management in IRAN. Successful Implementation in Health Care Organizations, Leadership in Health Services , 18 (3), 12-34.
Rad, A. M. (2006). The Impact of Organizational Culture on the Successful Implementation of Total Quality Management, The TQM Magazine. 18 (6), 606-625.
Ross, J.E, Total Quality Management: Text and Cases Readings, 2nd Edition (1994).
Rout, T. P., El Emam, K., Fusani, M., Goldenson, D., & Jung, H. (2007). SPICE in Retrospect. Developing a Standard for Process Assessment. Journal of Systems and Software , 80 (9), 1483-1493.
ROZEE. (2010). Retrieved January 15, 2010, from It is one of the biggest local online recruiting agency in the country.: www.rozee.com.pk
Saastamoinen, I., & Tukiainen, M. (2004). Software Process Improvement in Small and Medium Sized Software Enterprizes. (T. Dingsoyr, Ed.) Springer , 69-78.
Sahraoui, & Sofiane. Enabling a TQM Structure Through Information Technology. Competitiveness Review. 29 (3).
References
142
Salant, P. A., & Dillman, D. A. (1994). How to Conduct Your Own Survey. Wiley New York.
Saleh, A. S., & Ndubisi, N. O. (2006, 04 11). SME development in Malaysia: Domestic and Global Challenges. University of Wollongong Economics Working Paper Series 2006. [Online] Available, . Retrieved from http://www.uow.edu.au/commerce/econ/wpapers.html.
Salim. Software Exports and its Challenges.
Samat, N., Ramayah, T., & Yusoff, Y. M. (2008). Do ISO Certified SME's have Higher Quality Practices? Empirical Insights from the Northern Region of Malaysia, International Journal of Business and Management. , 3 (3), 66.
Sambrook, S., Greeetshuis, S., & Cheseldine, D. (2001). Developing a Quality Assurance System for Computer-Based Learning. Problems and Issues. Assessment & Evaluation in Higher Education , 26 (5), 417-426.
Saraph, J., Benson, P., & Schroeder, R. (1989). An instruement for measuring the critical factors of quality management. Decision Science , 20 (4), 810-829.
SEI. (n.d.). Retrieved July 4, 2010, from Software Engineering Institute, Carnegie Mellon University, CMMI®, : www.sei.cmu.edu/tsp
SEI. (n.d.). Retrieved July 10, 2010, from Software Engineering Institute, Carnegie Mellon University, CMMI®: www.sei.cmu.edu/
SEI. (n.d.). Retrieved 12 10, 2009, from www.sei.cmu.edu/tsp/tsp.html: www.sei.cmu.edu/tsp/tsp.html
Serrano, M., Oca, C. M., & al, e. (2006). Proceedings of First International Workshop on Process Improvement. An Experience on Implementing the CMMI in a Small Organization Using the Team Software Process , 81.
Shahmoradi, S. (2005). Study of Organizational Structure and Culture in Order to Implement TQM.
Sheard, S. A. (2001). Evolution of the Frameworks Quagmire, SOFTWARE MANAGEMENT. In B. B. [email protected] (Ed.), Software Productivity Consortium.. Computer, (p. 96).
Shen, B; Ruan, Tong, A case study of software process improvement in a chinese small company. Computer science and software engineering, 2008 international conference on, 609 (2008).
Sheraz, M. (2010). Technologiz. Retrieved 2010, from Questionaire Hosting Site: www.technologiz.com/quality
Shoniregun, C. (2004). Classification and Taxonomy of TEISMEs 5. (Richard Baskerville, Jan Pries Heje, 1999) Knowledge Capability and Maturity in Software Management.
Shull F, S. C., Zelkowltz, M., & Victor, R. basili's contributions to software quality. Software, IEEE. 23 (1), 16-8.
Sila, I., & Ebrahimpour, M. (2003). An Investigation of the Total Quality Management Survey Based Research Published Between 1989 and 2000. A literature review International Journal of Quality & Reliability Management. 19 (7), 902-70.
Siok, M. F., & Tian, J. (2007). Empirical study of embedded software quality and productivity. High assurance systems engineering symposium, 2007. HASE '07. 10th IEEE. 313.
Siviy, M. J., & Eileen, C. (2004). Accelerating CM Using Six Sigma: Pittsburgh: Carnegie Mellon, SEI.
Siviy, M. J., & Forrester, E. C. (2004). Accelerating CMMI Using Six Sigma. Carnegie Mellon, SEI., Pittsburgh.
References
143
Stake, R. E. (2005). Qualitative case studies. In N.K. Denzin, & Y.S. Lincoln (Eds.), The sage Handbook of Quality Research (3rd Edition ed.). London: Sage Publications.
STAMBOLLIAN, A., HABRA, N., LAPORTE, C., DESHARNAIS, A., & RENAULT, A. (May 4-5, 2006). OWPL: A Light Model & Methodology for Initiating Software Process Improvement. Proceedings of SPICE Conference. Luxembourg.
Steiger, J. H. (1990). Structural Model Evaluation and Modification. An Interval Estimation Approach, Multivariate Behavioral Research , 25, 173-180.
Stelzer, D., Mellis, W., & Herzwurm, G. (1996). Software process improvement via ISO 9000? In 1. p.-n. results of two surveys among european software houses. System sciences (Ed.)., (p. 703).
Sulkani, M. (2007). Software Technology Park. Monthly Progress Report. Punjab Information Technology Board (PITB) GOP.
Sun, Y., & Liu, X. (2010). Business-oriented software process improvement based on CMMI using QFD. Information and Software Technology. 52 (1), 79-91.
Switzer, J. (2004). Integrating CMMI, TSP, and Change Management Principles to Accelerate Process Improvement. Second Annual Meeting of the TSP User Group. Pittsburgh, PA.
Tabachnic, B. G., & Fidell, L. S. (2007). Sing Multivariate Statistics . (5, Ed.) New York: Allyn and Bacon.
Tamimi, N., & Gershon, M. (1995). A tool for assessing industry TQM practice versus Deming philosophy. Production and Inventory Management Journal , 36 (1), 27-32.
Tanaka, T., & Aizawa, M. O. (1998). Software quality analysis and measurement service activity in the company. Software engineering, 1998. proceedings of the 1998 international conference on., (p. 426).
Team, C. P. (2000). CMMI SM for Systems Engineering, Software Engineering, Integrated Product and Process Development, and Supplier Sourcing. Pittsburgh: SEI.
Team, CMM. (2006). CMMI ® for Development, Improving Processes for better products CMU/SEI-2006-TR-008. Carnegie Mellon University.
Team, CMM Product. (2006). Improving Processes for Better Products: CMMI ® for Development CMU/SEI-2006-TR-008/ ESC-TR-2006-08. no. 1.2. Pittsburgh: Carnegie Mellon.
Thorn, & Ron. (1999). The Influence of Organizational Structure on the Effectiveness of TQM Programs. Journal of Managerial .
Tick, I. T. (1992). A Guide to Software Quality Management System Construction and Certification Using. EN29001, U.K. Department of Trade and Industry and the British Computer Society. no. 2.
Van de Ven, A., & Ferry, D. (n.d.). Measuring and assessing organizations, Wiley.
Wadrsworth, H. M., Stephens, S., & Godfrey, A. B. (2002). Modern Methods for Quality Control and Improvement. Wiley, U.S.A.
Wall, D., James McHale, J., & M., P.-H. (2005). Accelerating Process Improvement by Integrating the TSP and CMMI. SEI, Pitsburgh, PA. USA.
Wei, K. K. Service Quality Index: A Study on Malsian Banks. Contemporary Management Research. 5 (2), 109-124.
Weltgen, W. (2004). Total Quality Management as a Structuring Task. Presented at the Work Shop of the Schülke & MayrGmbH. 26-27.
References
144
Wheeler, S., & Duggins, S. (1998). Improving Software Quality. Computer Science Department Southern PolytecPhnic State University. ACM.
William, R., Nichols, W. R., & Salazar, R. (2009). Deploying TSP on a National Scale. SEI.
Xiaoguang, Y., Xiaogang, W., Linpin, L., & Zhuoning, C. (2008). Research on organizational-level software process improvement model and its implementation. Computer science and computational technology, 2008. ISCSCT '08. international symposium ., (p. 285).
Yin, R. K. (2003), Case Study Research, Thousand Oaks, CA: Sage, London.
Yoo, C., Yoon, J., B, L. B., Lee, C., Lee, J., Hyun, S., et al. (2004). An Integrated Model of ISO 9001:2000 and CMMI for ISO Registered Organizations. Apsec, 11th Asia-Pacific Software Engineering Conference (APSEC'04)., (pp. 150-157).
Yoo, C., Yoon, J., Lee, B., Lee, C., Lee, J., Hyun, S., et al. (2004). An Integrated Model of ISO 9001:2000 and CMMI for ISO Registered Organizations. Proceedings of the 11th Asia-Pacific Software Engineering Conference (APSEC’04),, (pp. 150-157).
Yusof, S. M. (1999). Development of a Framework for TQM Implementation in Small Business, Unpublished Ph.D. thesis, University of Birmingham.
APPENDIX COVER LETTER
145
APPENDIX A – COVER LETTER
OPTIMUM SOFTWARE PROCESS IMPROVEMENT PARADIGM FOR QUALITY PRACTICES IN SOFTWARE INDUSTRY
This research project is about finding out whether bare minimum
quality practices are understood and implemented in the local software
industry. As a step further in this direction the objective is to map the actual
environment and true culture of Small and Medium Enterprises (SME)
towards quality improvement, process improvement, and CPI. The feedback
from this survey will give us the concrete discrepancies between true SME
culture and enforced culture of international standards like CMM, CMMI, and
ISO etc. After identifying characteristics of a true SME culture, we will provide
a set of guidelines and a process improvement paradigm for SME, which is
the basic purpose of this research. The set of guidelines for SME software
process improvement paradigm will give the innovative and flexible directions
for SMEs to change their culture and improve their processes and quality.
Indeed organizations of all sizes especially of small size can implement it for
the improvement of their product and process quality. The new guidelines to
implement quality will enable small and medium sized software houses to
build optimum quality culture and maintain a bare minimum level of quality
that will lead SMEs to become competitive and as well as quality
organizations through continuous process improvement.
We are therefore writing to you to solicit your help and support in this
matter. The experience of your organization in this field will be extremely
valuable to our research. It is appreciated that this questionnaire may take
some of your valuable time, however this survey should not take more than 10
minutes to complete. The findings of this research “SPI paradigm and
Guidelines”, will be shared on your request. If you need any further
APPENDIX COVER LETTER
146
information or clarification, please do not hesitate to contact the key
researcher. I appreciate your kind co-operation in this matter, and look
forward to receiving your input. Identity of the assessor and the name of the
organization is not the part of this research (optional), therefore your privacy