SUPPLIER PERFORMANCE ASSESSMENT TOOL IN AUTOMOTIVE
INDUSTRY USING MULTIVARIATE ANALYSIS
MOHD AZRIL BIN AMIL
UNIVERSITI TEKNOLOGI MALAYSIA
SUPPLIER PERFORMANCE ASSESSMENT TOOL IN AUTOMOTIVE
INDUSTRY USING MULTIVARIATE ANALYSIS
MOHD AZRIL BIN AMIL
A thesis submitted in fulfilment of the
requirements for the award of the degree of
Master of Engineering (Mechanical)
Faculty of Mechanical Engineering
Universiti Teknologi Malaysia
AUGUST 2009
iii
Dedicated to
My parents, Amil Yaacob and Ruzleina Ramly,
My wife, Rosnadia,
My brothers, Azriq & Azwan
My sisters, Awin, Wida & Wiza
All my family and friends for their immensurable support and love
iv
ACKNOWLEDGEMENT
Thanks to ALLAH, the most gracious and the most merciful, for His
guidance to accomplish this research. Without His help and mercy, this would not
been possible. HE is the one who knows the hardships and HE is the one I seek his
satisfaction and ask HIS acceptance.
I would like to express my deepest gratitude towards my advisor, Professor
Dr Sha’ri Mohd Yusof for his guidance, encouragement and valuable comments
during the research and writing of this dissertation. His ever willingness and kindness
for guiding me as well as providing me with valuable information are truly
appreciated. Without his continued support and interest, this thesis would not have
been the same as presented here.
I wish to express my appreciation to my FKM-E04 members, Halim
Abdullah, Muhammad Adil Khattak and Suhail Kazi for their generous cooperation,
hospitality, time and insight on related matters during this research. My fellow other
postgraduate students also be recognised for their support, whom one way or the
other, contributed to make this research a success and for their assistant in laboratory
work. I am also indebted to Universiti Teknologi Malaysia (UTM) for giving
opportunity for further my master study.
Special thanks go to my parents, Amil Yaacob and Ruzleina Ramly for their
patience and sacrifice during my academic career. Their concern, encouragement,
moral and financial support over the years has always been a source of motivation
that enables me to achieve this degree. To my brothers and sisters, Awin, Wida,
Wiza and Azriq and Azwan, I’m grateful to have all of you supporting me.
Finally, and most importantly, special thanks to my beloved wife, Rosnadia
Suainbon, for her unconditional love and support during my education. Thanks for
understanding my situation and for tolerating my absence during this challenging
time.
v
ABSTRACT
The supplier evaluation process is complicated because a variety of criteria
must be simultaneously considered. In some approaches to supplier evaluation, only
quantitative factors are allowed in the model, or qualitative factors can be used in the
model but the data are replaced by the assigned numbers. In practice, different goals,
multiple criteria, constraints and parameters that involve conflicting quantitative and
qualitative criteria make the decision making complicated. This thesis presents a
development of a supplier performance assessment tool to evaluate automotive
suppliers based on multivariate analysis. A questionnaire was prepared and sent to
278 companies from automotive sector in Malaysia. 5 forms were provided for each
of the company to perform total of 1390 recipients, thus giving 24.3 percent response
rate. Attempts were made to find the extent of practices in 5 different factors; quality
system, in-process quality, logistics and management, shipping and delivery, and
after sales services. The results were analyzed using the SPSS software. Factor
Analysis, one of the tools for multivariate analysis was used to design the assessment
tool for Supplier Performance Assessment and Evaluation. It can also be used for
Supplier Control. The procedure utilizes a proposed assessment tool; constructed
from factor analysis to do the evaluation. Since the supplier evaluation is a decision
problem combining multiple criteria or attributes into a single measure of supplier
performance, the objective is to find a method that can be used to objectively
evaluate the best supplier. It was also found that it is advantageous to use the
proposed instrument as it requires minimal manual interferences. If the proposed
instrument is executed and controlled regularly, the performance level of automotive
suppliers may be improved continuously. As a conclusion, this study may be able to
assist automotive suppliers to maintain and improve their performance. This will
support our Malaysian automotive industry as a whole.
vi
ABSTRAK
Pengukuran prestasi pembekal atau vendor adalah agak sukar dan kompleks
kerana ia melibatkan banyak faktor dan ciri-ciri yang perlu dipertimbangkan secara
serentak. Di dalam sebahagian kaedah yang digunakan untuk mengukur prestasi
vendor, hanya faktor-faktor kuantitatif yang diambil kira dalam model penilaian,
ataupun hanya faktor-faktor kualitatif yang diterjemahkan dalam bentuk angka
supaya ia dapat digunakan sebagai pengiraan. Dari segi amalan, matlamat yang
berbeza, kriteria yang pelbagai, kekangan dan parameter yang melibatkan angkaangka kuantitatif dan kualitatif bertentangan menyebabkan sukar untuk membuat
keputusan bagi permasalahan ini. Tesis ini menerangkan prosedur untuk
membangunkan alat/kaedah untuk mengukur keupayaan prestasi pembekal/vendor
melalui analisis multi-variasi. Borang soal-selidik telah disediakan dan telah
diedarkan kepada 278 syarikat pembekal dalam sektor automotif di seluruh negara.
Setiap syarikat itu telah diedarkan 5 set borang soal-selidik dan menjadikan sebanyak
1390 penerima borang soal-selidik seluruhnya. Kadar respon adalah 24.3 peratus.
Soal-selidik ini adalah untuk mencari keluasan penggunaan 5 faktor yang berbeza
iaitu; kualiti sistem, kualiti proses, pengurusan logistik, penghantaran dan juga servis
selepas jualan. Keputusan soal-selidik itu kemudiannya di analisa dengan perisian
SPSS. Analisa Faktor, salah satu kaedah Analis Multivariasi telah digunakan bagi
mengukur prestasi vendor di dalam kajian ini. Di samping itu, kaedah ini juga boleh
digunakan sebagai kaedah pengawalan vendor. Prosedur ini menggunakan alat yang
dicadangkan hasil daripada analisa tersebut. Oleh kerana pengukuran prestasi vendor
merupakan masalah keputusan yang melibatkan pelbagai ciri, maka keputusan
tersebut seharusnya berkisar untuk mencari vendor yang terbaik. Didapati juga
bahawa kaedah/alat yang dicadangkan mempunyai kelebihan lain iaitu ia hanya
memerlukan pengolahan data yang minima. Jika kaedah yang dicadangkan
diimplementasi dan dikawal dengan baik, tahap prestasi vendor automotif dijangka
dapat ditingkatkan dengan lebih baik. Sebagai rumusan, kajian ini diharapkan dapat
membantu untuk meningkatkan prestasi vendor-vendor automotif. Seterusnya ia akan
menambah-baik industri automotif Malaysia secara keseluruhannya.
vii
CHAPTER
1
PAGE
DECLARATION
ii
DEDICATION
iii
ACKNOWLEDGEMENT
iv
ABSTRACT
v
ABSTRAK
vi
TABLE OF CONTENTS
vii
LIST OF TABLES
xi
LIST OF FIGURES
xiii
LIST OF APPENDICES
xiv
INTRODUCTION
1.1
Introduction
1
1.2
Background of the Problem
2
1.3
Research problem
3
1.3.1
Statement of Research Problem
3
1.3.2
Research Question
3
1.4
Objectives of the Research
4
1.5
Scope of the Research
4
1.6
Significance of the Research
5
1.6.1
6
1.7
2
TITLE
Why Automotive Industry?
Layout of Thesis
6
LITERATURE REVIEW
2.1
Introduction
8
2.2
Supplier Performance Issues
9
2.3
Supplier Development Issues
11
2.4
Supplier Performance Evaluation and Performance 15
Assessment Categories
viii
2.4.1
Method of Supplier Selection and Supplier 16
Performance Evaluation
2.4.1.1
Rating Methods
16
2.4.1.2
Mathematical Methods
20
2.5
Criteria for Supplier Performance Assessment
28
2.6
Previous Research Studies on Criteria Assessment for 32
Supplier’s Performance Evaluation
2.7
3
Summary
40
MULTIVARIATE ANALYSIS THEORY AND ITS
APPLICATION
3.1
Introduction
41
3.2
Factor Analysis
42
3.2.1
Steps in Factor Analysis
44
3.2.2
Examining the Correlation Matrix
45
3.2.3
Factor Extraction
47
3.2.3.1
48
3.2.4
3.3
4
Methods for Factor Extraction
The Rotation Phase
50
3.2.4.1
52
Factor Scores
Summary
52
RESEARCH METHODOLOGY
4.1
4.2
4.3
Introduction
53
4.1.1
53
Overview of Research Methodology
Survey Methodology
56
4.2.1
Introduction
56
4.2.2
Survey Questionnaires Development
56
4.2.3
Expert Validation and Pilot Study of Survey
58
Data Collection
59
4.3.1
Mail Surveys
59
4.3.2
Sampling Frames
60
ix
4.3.3
Population of the Study
61
4.3.4
Management for Non-response and Steps to 62
Increase Response Rates
4.4
Data Analysis
63
4.4.1
Reliabilty
63
4.4.2
Multivariate Analysis
63
4.4.2.1 Using
Factor
Analysis
for
64
Supplier Performance Assessment
and Evaluation
4.4.2.2 Steps in Using Factor Analysis in
Supplier Performance Assessment
and Evaluation
4.5
5
Summary
68
SURVEY RESULTS AND ANALYSIS
5.1
Introduction
69
5.2
Response rate
70
5.3
Descriptive Statistics of Respondents
71
5.4
Reliability Test
74
5.5
Non-Response Bias
75
5.5.1
Hypotheses
75
5.5.2
T-test for equality of mean
76
5.6
5.7
Examination of data
78
5.6.1
78
Validation of Data Entry
Data Processing with Multivariate Analysis
82
5.7.1
Initial Output layout
87
5.7.1.1
KMO and Bartlett’s Test
87
5.7.1.2
Communalities
88
5.7.2
Output for rerun
5.7.2.1
90
KMO and Bartlett’s Test with 90
Communaltities
x
5.7.3
Total Variance Explained
92
5.7.4
Scree Plot and Component plot in rotated
93
space
5.7.5
Rotated Component Matrix
95
5.8
Interpreting Result
97
5.9
Constructed Supplier Performance Assessment Tool for 101
Automotive Industry
5.10
6
Discussion and Summary
104
CONCLUSION AND RECOMENDATION
6.1
Introduction
106
6.2
Summary and Conclusions of the Research
106
6.3
Limitation of the study
109
6.4
Future Research Recommendations
110
REFERENCES
113
APPENDICES
Appendix A-C
119
xi
LIST OF TABLES
TABLE NO.
2.1
TITLE
Classification of suppliers (Muralidharan and Anantharaman,
PAGE
10
2001)
2.2
Illustration of the
Factors Ranking Method for supplier
17
performance evaluation (Enriquez et al., 2005)
2.3
Delivery cost for supplier X (Dahel and Nasr, 2003)
19
2.4
Illustration of the Cost-Ratio Method (Dahel and Nasr, 2003)
19
2.5
Pair-wise Comparison Matrix : Evaluation Criteria (Second
23
Level in the Hierarchy) (Kahraman,2003)
2.6
Illustration of the weight-point method for supplier
25
performance evaluation
2.7
Criteria proposed by Dickson (1966)
29
2.8
Chronological summary of the literature review
34
4.1
Population Size
61
5.1
Summary of responses
70
5.2
Classification of the supplier
71
5.3
Supplier’s Main Business to the Automotive Manufacturer
72
5.4
The length of time in conducting business
73
5.5
Quality Certification
73
5.6
Internal consistency of the questionnaire
74
5.7
Comparing Early to late Respondents
77
5.8
Means and Standard Deviation of responses
79
5.9
Total element and their nomenclatures
84
xii
5.10
KMO and Bartlett’s Test (a)
87
5.11
Communalities (a)
89
5.12
KMO and Bartlett’s Test (b)
90
5.13
Communalities (b)
91
5.14
Total variance explained (a)
93
5.15
Rotated Component Matrix
96
5.16
Factors and their elements
99
5.17
Constructed Assessment tool for Automotive Industry
101
xiii
LIST OF FIGURES
FIGURE NO.
2.1
TITLE
PAGE
The Buying Firm's Perspective of Supplier Development
13
Program (Hahn et al, 1990)
2.2
Supplier Evaluation Hierarchy
2.3
Overview
of
Supplier
Selection
22
and
Performance
27
Evaluation categories
4.1
Overview of the research Methodology
55
4.2
Using Factor Analysis as a Supplier Evaluation and
65
Controlling
4.3
Factor analysis steps
67
5.1
Scree Plot
94
5.2
Component plot in rotated space
95
xiv
LIST OF APPENDICES
APPENDIX
TITLE
PAGE
A
Questionnaire
119
B
Letters
125
C
Reliability Test
131
CHAPTER 1
INTRODUCTION
1.1
Introduction
It is a well established fact that an organization can perform no better than its
suppliers. This fact along with greater demands for lower prices and continuous
improvement in all aspects of supply management make supplier assessment and
performance measurement a critical process in world-class organizations. Yet it is
acknowledged that a majority of enterprises are less than satisfied with their ability to
consistently select the best suppliers or measure and manage supplier and contractor
performance. Nowadays, there are many techniques had been used such as those that
focus on accounting techniques, auditing techniques and quality certificates for
performance measurement.
However, there is no theoretical or generic approach to studying the practice
of ongoing companies’ performance measurement, in particular on how companies
use performance measurement to manage their relationships and interactions with
suppliers and how suppliers respond to the measurement (Schmitz and Platts, 2003).
Supplier Performance Assessment is a technique of measuring a supplier’s
actual performance against a set of agreed criteria then awarding "marks" according
2
to the quality of that performance. These criteria are often called "Performance
Indicators". The transformation of qualitative data to quantitative data means that it
can be measured and evaluated. It is an objective way of assessing a supplier’s
performance.
1.2
Background of the Problem
The supplier evaluation process is complicated because a variety of criteria
must be simultaneously considered. In some approaches to supplier evaluation, only
quantitative factors are allowed in the model, or qualitative factors can be used in the
model but the data are replaced by the assigned numbers. However, the assigned
numbers may not directly reflect the impreciseness of the performance data. In order
to obtain an effective evaluation, the impreciseness of data should be accurately
reflected.
Three traditional techniques are designed to evaluate suppliers: the
categorical method, the weighted-point method, and the cost-ratio method
(Muralidharan and Anantharaman, 2001). These traditional evaluation processes
have the disadvantage of being either intuitively judged by the evaluator or too
expensive to use. Operations-research-oriented approaches may also be available for
dealing with the problems of the supplier evaluation process. In general, however,
they are not only too complex for practical use but they are used to solve
optimization-oriented problems. Since the supplier evaluation is a decision making
problem combining multiple criteria or attributes into a single measure of supplier
performance, the objective of selecting suppliers is to find a method that can be used
to objectively evaluate the best supplier. Unfortunately, the methods listed above do
not provide a generally applicable methodology, are too complex for practical use by
operating managers, or do not fit to this type of problem.
3
1.3
Research Problem
1.3.1 Statement of Research Problem
More often than not, a supplier assessment is based on the lowest bid, and in
some cases on unsystematic and incomprehensive subjective evaluation and
interviews. Therefore, it becomes too late to proactively avoid supplier issues or
divest production flow of their symptoms. If causes of the suppliers’ issues (i.e
quality, delivery, etc) are accounted for early in the supplier assessment process, the
associated risk could be minimized.
1.3.2 Research Question
The general question this study attempts to answer is this: is there a more
comprehensive and effective supplier performance assessment model that minimizes
the risk associated for their end product to automotive manufacturer? The general
question subsumes several related questions:
1. Which practices contribute the most to the suppliers’ end products to the
automotive manufacturer?
2. How are we doing in supplier assessment and performance measurement?
It will motivate suppliers with performance measurement and receive
feedback from the supplier’s point of view.
4
1.4
Objectives of the Research
The objectives of the study:
 To design an assessment tool that can be used as a generic approach to
measure supplier performance in automotive industry.
 Using multivariate analysis as a method for supplier performance
assessment development process. This is to prove that multivariate
analysis can be used to perform the assessment.
 To give benchmark from which to measure improvement. Suppliers need
to know how well they are performing and to have the opportunity to meet
the needs of the customer better. In the rare event that supplier
performance is so poor that the contract needs to be terminated and/or
damages sought, supplier rating provides objective documented evidence
of unsatisfactory performance.
1.5
Scope of the research
The scope of this study is to develop supplier performance assessment tools
using multivariate analysis approaches. The focus is limited to companies which are
suppliers from automotive industry manufacturing sectors in Malaysia. Also, the
focal point process in this research is the evaluation of suppliers and benchmark it
based on qualitative and quantitative data from the measurement.
5
1.6
Significance of the research
Manufacturers can attain multiple benefits by
measuring supplier
performance. Companies that fail to measure most of their suppliers risk large-scale
quality mishaps, service deficiencies, and cost overruns that can eat into bottom-line
profits and damage competitive positioning in the market. On the other hand,
companies that subscribe to such practices can reduce buffer inventory, cut cycle
times, and lower the total cost of ownership (TCO) of their supply chains.
There are also several reasons to evaluate suppliers. First, by evaluating
results of suppliers, buying firms can identify who fits the requirements best. Buying
firms can upgrade and obtain the greatest competitive advantage by cooperating with
better performers. Second, supplier evaluation also develops a better negotiating
position for the buying firm. Third, supplier’s performance directly impacts the
buying firm’s performance. For example, a better quality product may result from
higher quality material, a lower manufacturing cost may result from lower
purchasing cost, and a short production schedule results from shorter lead time of
orders, etc.
The concept of Supplier Performance Assessment has been developed since
10 to 20 years ago with so many control laws has been introduced but has yet to be
evaluated in a real application. In Malaysia, many suppliers and manufacturer lack
the expertise to perform such task and with this study, it hopefully will benefit all
parties. The study will hopefully be a platform for future research in a similar field.
1.6.1 Why Automotive Industry?
The automotive industry was selected due to the diversity of businesses and
because the relationships between suppliers and manufacturers are well developed
6
and fairly stable. The automobile manufacturer had indicated that the selected
suppliers were superior or critical for their business success. When a supplier is
critical to the buying firm, the buying firm is more inclined to utilize supplier
development of the problem. (Porter, 1997)
1.7
Layout of thesis
This research thesis is organized into six chapters:
1.
Chapter I : Introduction
2.
Chapter II : Literature Review
3.
Chapter III : Research Methodology
4.
Chapter IV : Multivariate Analysis and Its Application
5.
Chapter V : Result and Analysis
6.
Chapter VI : Conclusion and Recommendations
Chapter 1, describes the background of research, problem statement, purpose
of research, importance of the research, scope of the research and layout of the thesis.
Chapter 2, presents a review of the literature to understand the issues and
formulate the research problems. The review describes about Supplier Performance
and development issues, supplier assessment categories and methods that have been
used to perform the assessment. This chapter also described criteria that check upon
during the assessment.
Chapter 3, describes basic definitions of multivariate analysis and the original
set of factor analysis theory, as well as the steps of operation. It explains about
multivariate analysis, its uses and application. Details of steps for conducting Factor
Analysis, the multivariate tool used in this research are explained briefly.
7
Chapter 4, describes the research methodology employed in conducting the
study. Survey methodology is the main approached adopted. The survey method was
used to find-out the practices level of performance implemented in supplier’s firm.
The questionnaire is developed in order to fulfill the objectives of research.
Chapter 5, presents the results and findings from the survey. Using SPSS
software, data were processed with factor analysis techniques, mean test and t-test. It
explained the results which were relevant to research questions. From the result, an
instrument/tool for supplier assessment and performance evaluation was developed.
This instrument/tool was used in the case study.
Chapter 6, presents the conclusion of this study. The report culminates with
some suggestions and discussions for future research. It also presents the limitation
of the study.
CHAPTER 2
LITERATURE REVIEW
2.1 Introduction
In this chapter, the literature review on supplier performance evaluation and
performance assessment is presented. The purpose of this chapter is two-fold. The first is
to briefly review on the key criteria in the literature on supplier evaluation so that the
reader will understand the relevance of this study. The second purpose of this chapter is
to review and compare analytical approaches used in supplier evaluation and to note
deficiencies in these approaches.
Supplier selection is one of the most critical activities of purchasing management
in a supply chain, because of the key role of supplier's performance in achieving the
objectives of a supply chain. In all purchasing process, the main objective is to select a
good supplier that meet specifications for quantity, quality, delivery and price (Tsai,
1999). Najmi et al. (2005), suggested that the supplier selection is so important because
it will impact to the final cost of the products, production performance and a strategic as
well as tactical decision by the company. Therefore, the success of a company is
determined to a greater degree by the abilities of its suppliers.
9
2.2
Supplier Performance Issues
There are several issues that need to be addressed when discussing about supplier
performance evaluation. Shortage of raw materials, shorter lead time, high quality,
increasing the variety of products with smaller runs, inflation, productivity, and
introduction of a just-in-time (JIT) purchasing system, etc., has prompted the realization
of the importance of purchasing in many manufacturing firms (Muralidharan and
Anantharaman, 2001).
To ensure that incoming materials meet relevant quality standards, supplier
evaluation or rating must be carried out periodically. Suppliers should be selected on the
basis of how well they meet a variety of specific requirements – requirements that do not
depend solely on price, but the total ‘cost of ownership’ of materials (Li et al., 1997).
Because production begins with purchasing, and a purchasing program will not be
successful unless cooperative buyer/supplier relationships are established and
maintained, supplier selection is one of the most important activities in purchasing.
Effective supplier selection can also help companies achieve ‘just-in-time’ (JIT)
production.
The search for new suppliers is a continuous priority for companies in order to
upgrade the variety and competitiveness of their products range. This is essentially due
to two reasons. First, product life cycle is becoming very short (3 to 4 years or less) and
new models must often be developed by using completely renewed material or with new
technologies. Second, many industries are labor intensive. These aspects are expressed
through a complex pattern of demand for material and labor. There are different aspects
that characterize the supplier performance measurement problem.
The first aspect is the determination of the number on the suppliers and modes of
relations with them. Considering the characteristics of the company, product and market,
company’s strategy can encourage a large number of suppliers or not. On the other hand,
10
the company can have a hierarchical relation and a significant number of suppliers for
the standard parts in order to establish a competition between them and thus to reduce
purchasing costs. Several authors, Schimtz and Plattz (2003), Sucky (2001) and Chen et
al. (2005) are interested by the suppliers classification problem.
The second aspect is related to the selection of the best suppliers among the
existing criteria. Several European companies have found it wiser to keep two suppliers
with a 80-20 per cent split-up of orders, so that one is a major supplier (may be the
approved supplier) while the other (may be the preferred supplier) has large enough
business to keep him interested in the buying firm (Muralidharan and Anantharaman,
2001). Table 2.1 shows the classification of suppliers according to them.
Table 2.1 : Classification of suppliers (Muralidharan and Anantharaman, 2001)
Less than or equal to 20 per cent of the total number of suppliers and who
Approved
secure greater than or equal to a minimum prescribed composite score
40 per cent to 70 per cent of the total number of suppliers and who secure
Deferred
less than or equal to a minimum prescribed composite score
20 per cent to 40 per cent of the total number of suppliers and who secured a
Preferred
composite score in between that of approved and deferred supplier
Based on this approach, suppliers can be classified as approved suppliers,
preferred suppliers and deferred suppliers. If the current approved supplier (80 per cent
supplier) turns against the buyer-company, which can wreak havoc (Tsai, 1999), then the
20 per cent supplier (preferred supplier) could be built up quickly to take up a major
position.
11
Another important aspect about supplier evaluation is the complexity in decision
making. According to Muralidharan and Anantharaman (2001), the evaluation of
suppliers is a complicated decision problem, because of the following reasons:

The relative difficulty to conceptualize and structure and numerous components
of the evaluation problem into an analytical framework which may facilitate
understanding.

The nature of the components in this process which some are quantitative
whereas others are subjective.

There are multitudes of factors/attributes involved in a selection process which
are often conflicting and sometimes complementary. Many times, such
factors/attributes
are
non-expressible
in
commensurable
units.
Some
factors/attributes might reflect psychological aspects such as qualitative
considerations and intangibles.
It is therefore crucial to define and characterize those factors that will be useful in
conducting supplier performance evaluation.
2.3
Supplier Development Issues
The supplier performance evaluation and performance is a crucial factor in
supplier development program. Most of the supplier development literature has utilized
case study, such as Giunipero et al. (1998) and Larson and Kulchitsky (1999) to describe
business practices in integrating supplier performance evaluation and performance
assessment in supplier development. There is a need to explore the impact of supplier
development programs within the buyer-supplier relationship and on supply chain
12
performance. In this section, the importance of supplier development programs to the
buying firm will be presented.
Buying firms emphasize the development of long-term, cooperative relationships
with their critical suppliers (Millington et al., 2005). When a supplier is unable to
conform to the buying firm's expectations, the buying firm must determine the most
appropriate action, which could include: (1) motivating the suppliers to improve their
performance abilities, (2) modifying the objectives to a more realistic standard, (3)
switching suppliers or developing alternative suppliers, or (4) producing the product inhouse (Millington et al., 2005).
Usually buying firms recognize when there is a disparity between their objectives
for the incoming product/service and the performance of their suppliers. The question
that many purchasing managers have is, "Now what?" Ideally, the managers should predetermine the possible actions that they would take in the event of various outcomes
(Saunders, 1994). However, many professionals continue to struggle with the question
of which action is preferable in a particular scenario.
Supplier development activities are some of the buying firm's possible responses
to the disparity between their objectives and their evaluation of the supplier's
performance. As shown in Figure 2.1, the development activities are typically triggered
by an evaluation of the supplier (Hahn et al., 1996). The literature on supplier
evaluations will be discussed in the following section.
13
Buying Firm’s Objectives
Buying Firm’s Response –
Supplier Development
Programs
Buying Firm’s Evaluation of
the Supplier’s Performance
Incentive
Process- Based
Evaluation
PerformanceBased Evaluation
Competitive
Pressure
Direct
Involvement
Supplier
Assessment
Figure 2.1: The Buying Firm's Perspective of Supplier Development Program
(Hahn et al. 1990)
Supplier development programs are defined as "any set of activities undertaken
by a buying firm to identify, measure and improve supplier performance and facilitate
the continuous improvement of the overall value of goods and services supplied to the
buying company's business unit" (Krause et al., 2000). From the buyer's perspective,
supplier development programs are warranted when the supplied product or service is
essential for success (Handfield, 2003).
Krause et al. (2000) developed four categories of supplier development activities:
incentives, competitive pressure, direct involvement, and supplier assessment. Incentives
are the positive reinforcement mechanisms used to entice suppliers to improve their
performance according to the goals and objectives established by the buying firm.
Examples include:

promises of current benefits

promises of future benefits
14

awards and public recognition, such as the annual publication of "General
Motor's Suppliers of the Year" in the USA Today newspaper; and

long-term contracts
Competitive pressure is the negative reinforcement mechanisms used to pressure
suppliers to improve their performance. Examples include: (1) increased competition by
using multiple sourcing or emphasis on switching suppliers (Hahn et al., 1996) and
emphasis on reducing the supply base (Krause et al., 2000).
Direct involvement is the assistance and hands-on direction provided by the
buying firm to the supplier. Examples like training and education of the supplier's
personnel (Williams and Oumlil, 1987) and site visits to help the supplier improve its
performance.
The final supplier development category is the supplier assessment. Supplier
assessment is the communication of the performance evaluation to the supplier. There
are two categories of the evaluation: process-based evaluation, which assesses the
supplier's capabilities, and performance-based evaluation, which assesses the supplier's
achievements. The definitions of these categories will be further developed in section
2.4.
The focus of this research is on improving an existing supplier's performance and
capabilities through the supplier evaluation process. Therefore, this research is trying to
develop a tool for supplier assessment.
15
2.4
Supplier Performance Evaluation and Performance Assessment Categories
The buyer's evaluation of the supplier's performance is a catalyst for the supplier
development activities. According to Hartley and Choi, (1996) there are two main
categories for the supplier evaluation: process-based evaluations and performance-based
evaluations.
The process-based evaluation is an assessment of the supplier's production or
service process. Typically, the buying’s firm will conduct an audit at the supplier's site
to assess the level of capability in the supplier organization's systems for costing,
quality, technology, and other specific factors. Process flow charts can be developed to
identify the non-value added activities that should be eliminated to improve the business
efficiency (Cebi and Bayraktar, 2003a). Increasingly, large buying’s firm are demanding
that their suppliers become certified through a third-party organization, such as ISO
9000 certification or Malcolm Baldrige National Quality Awards (Porter, 1997).
The performance-based evaluation is an assessment of the supplier's actual
performance on a variety of criteria, such as delivery reliability, cost, and quality defect
rate. It is a more of a tactical assessment and measures the day-to-day actual
performance of the supplying firm; hence it is an after-the-fact evaluation. The
performance-based evaluation is more common than the process-based evaluation
(Porter, 1997), perhaps because it is reactive and easily measured. Once completed, the
evaluation can be compared to either the buying firm's stated goals or benchmarked to
the performance evaluations of the supplier's competitors.
Various methods have been proposed with regard to supplier performance
assessment. In section 2.4, the studies available on supplier performance assessment are
reviewed for their application and development.
16
2.4.1 Method of Supplier Selection and Supplier Performance Evaluation
Two major approaches have been suggested by researchers in this area in term of
supplier selection and performance evaluation method (Prahinsky, 2001). It can be
grouped into the rating method and the mathematical approach. The first method, rating
method deals with factors rating and popular in traditional ways because of its focus on
factors evaluation. This is due to it is easy to implement and requires minimal data. It is
also good for firms with limited resources. Nowadays, the mathematical methods had
gained popularity among researchers as the computers are advanced enough to compute
and analyzed complex data which are affected by conflicting factors for supplier
selection and evaluation. The mathematical methods focus on factors tradeoff.
2.4.1.1 Rating Methods
Rating methods encompasses two subcategories, criteria rating and cost methods.
Rating methods is a subjective weights or dollar values to evaluate and select suppliers.
Factors such as quality and delivery are subjectively evaluated, rated, and ranked. A
supplier, then, is compared to other suppliers and selected. Rating methods are divided
into two subcategories: factors ranking and cost methods.
a) Factors Ranking
Factors ranking subjectively weigh and ranks suppliers’ selection factors. The
main strength of subjective weighting of factors is its capability to evaluate suppliers’
performance through an unsophisticated evaluation process. Factors ranking methods are
generally conducted through different mechanisms such as plant visits, interviews, and
17
audits. For example, a buying company team member subjectively evaluates and rate
supplier’s performance factors. After some basic mathematical calculations, a number is
assigned to each supplier. Accordingly, the buying company team ranks each supplier
from high to low.
Factors ranking is the most popular method in supplier evaluation and selection
in a traditional ways. Different authors have provided various ways to apply factors
ranking. Some of the studies rate factors by assigning simple subjective weights to each
factor. For instance, Millington et al. (2005) used a simple weighted factor matrix
approach to select suppliers. They identified three major evaluation criteria and their
subjective weight: quality, cost, and delivery. The supplier with a higher score will be
selected over the others. Similarly, Enriquez et al. (2005) presented a simple approach
to evaluate medical devices suppliers through the use of linear averaging, which is based
on assigning weight to different suppliers’ selection criteria such as quality and delivery.
This concept is illustrated in Table 2.2. The supplier with the highest score is
selected. Clearly supplier Z in Table 2.2 is preferred since this supplier obtained the
highest total scores in this example.
Table 2.2: Illustration of the Factors Ranking Method for Supplier Performance
Evaluation. (Enriquez et al., 2005)
Supplier
Performance Attributes Product
Cost
Quality
X
Good (+)
Y
Neutral (0)
Z
Good
(+)
Neutral
(0)
Delivery
Unsatisfactory (-)
Unsatisfactory (-) Unsatisfactory (-)
Good (+)
Total
Neutral (0)
0
-
-
+ +
18
One problem with this approach is that attributes selected are equally weighted,
perhaps not truly reflecting reality. In addition, this method is mainly an intuitive
process, it is more qualitative than quantitative and lacks precise evaluation because it
depends heavily on the personal judgment, experience, and ability of the evaluators.
b) Cost Method
Cost methods supplement the factors ranking by assigning a dollar value to
selected and significant factors. For this reason, cost methods are easily communicated
to top management, who will ultimately make decisions. Very few authors have
employed cost analysis in their supplier factors evaluation. Monczka and Trecha (1988)
is one of the few authors that have elaborated on using cost methods in selecting a
supplier. Monczka and Trecha’s cost method evaluates suppliers based on different
factors, such as quality and delivery. The method assigns dollars per time spent, or lost,
due to quality or delivery problems.
Another approach to the cost factors estimation is through Activity Based
Costing (ABC) analysis. ABC accounts for all direct and indirect activities’ cost. Dahel
and Nasr (2003) focuses on selecting an optimal number of qualified suppliers using
ABC information.
Table 2.3 shows the delivery cost for supplier “X”. If the total delivery and the
purchase costs are $500 and $25000, respectively. The delivery cost ratio is 2% or
(500/25000). The other cost ratios can be found in a similar way. Table 2.4 illustrates the
procedure used to compare suppliers based on the calculated cost ratios and the bid price
for every supplier.
The net adjusted cost for each supplier is listed on the last column in the table.
The net adjusted price for supplier Z is $24840 or {$23000+(8%*23000)} which is the
lowest. Thus, supplier Z is recommended in this example.
19
Table 2.3: Delivery cost for supplier X (Dahel and Nasr, 2003)
Supplier: “X”
Elements
Costs
Telephoning
$50
Visit to Supplier Plant
$200
Special Shipment (air freight)
$200
Miscellaneous Expense)
$50
Total Delivery Cost
$500
Total Value of Purchase
$25000
Delivery-Cost Ratio (total delivery cost/total purchase)
2%
Table 2.4: Illustration of the Cost-Ratio Method (Dahel and Nasr, 2003)
Quality
Delivery
Service
Total Cost Bid
Net Adjusted
Cost Ratio
Cost Ratio
Cost Ratio
Ratio
Price
Cost
X
1%
2%
2%
5%
$25000
$26230
Y
2%
1%
6%
9%
$24500
$26705
Z
2%
2%
4%
8%
$23000
$24840
Supplier
Because of its complexity, the cost-ratio method is not widely used in industry
(Dobler and Burt, 1990). It requires a comprehensive cost-accounting system to generate
the precise cost data needed. In addition, all performance measures (cost ratios) are
artificially expressed in the same units. Another disadvantage of this method is that it
ignores judgmental factors which are difficult to quantify and factors which cannot be
converted into a cost ratio.
20
2.4.1.2 Mathematical Methods
The mathematical method encompasses four major categories: fuzzy sets,
weightage indexing, multivariate and statistical process control (Aamer,2005). Studies
under mathematical methods do not incorporate performance, but they examine
supplier’s criteria and tradeoffs (Aamer, 2005).
a) Fuzzy Sets / Multiple Criteria Decision-Making (MCDM)
According to Kahraman et al. (2003), fuzzy sets or sometimes called as Multiple
Criteria Decision Making (MCDM), are oriented to the rationality of uncertainty due to
imprecision or vagueness. A major contribution of fuzzy set theory is its capability of
representing vague data. The theory also allows mathematical operators and
programming to apply to the fuzzy domain. A fuzzy set is a class of objects with a
continuum of grades of membership. Such a set is characterized by a membership
(characteristic) function, which assigns to each object a grade of membership ranging
between zero and one.
The analytic hierarchy process (AHP) is one of the extensively used multicriteria decision-making methods. One of the main advantages of this method is the
relative ease with which it handles multiple criteria. In addition to this, AHP is easier to
understand and it can effectively handle both qualitative and quantitative data. The use
of AHP does not involve cumbersome mathematics. AHP involves the principles of
decomposition, pair-wise comparisons, and priority vector generation and synthesis.
Though the purpose of AHP is to capture the expert’s knowledge, the conventional AHP
still cannot reflect the human thinking style. Therefore, fuzzy AHP, a fuzzy extension of
AHP, was developed to solve the hierarchical fuzzy problems.
The AHP was developed primarily by Saaty (1980). The process of AHP is to
compare pairs of similar factors in the hierarchy by setting priorities among factors.
21
AHP provides a systematic approach that makes it easier for purchasing managers to
quantify their subjective evaluations. However, this approach shows weakness in its
ability to prioritize the performance factors. Factor priorities are measured by Saaty's
scale, which uses numbers to represent the relative importance of criteria regardless of
the type of criteria. In addition, Saaty's scale is an estimation of impreciseness within
two factors; it may not truly reflect the evaluator's need to capture the impreciseness of
an individual factor. Moreover, setting priorities in AHP may result in logical
inconsistencies, leading to ineffective measurement.
The first step is to construct a hierarchy defining the problem. The first level in
the hierarchy represents the problem's objective. In the second level, specific factors
used to reach the objective are constructed. Usually these specific factors can be broken
into sub-factors, and their levels may vary. In the hierarchy, the alternatives always
follow the sub-factor and established at the lowest level. The second step is to make
pairwise comparison within each level against a given factor or sub-factor at the adjacent
higher level. Since the first level has only one factor (objective of the problem), no
pairwise comparison is made. Pairwise comparisons start from the second level and the
comparisons are measured by using Saaty's scales. The compared results represent
priorities of criteria and/or sub-criteria. Once the performance criteria are prioritized, the
final step is to synthesize priorities for each competitive alternative. The synthesized
value represents the final rating of the alternative and is used to compare with the other
synthesized values. The process of synthesizing priorities is carried out by multiplying
the priority of the alternative at the lowest level by the priorities of those criteria at a
higher level of the hierarchy. The process is repeated until the final ratings for different
suppliers are developed. The selection decision is based on the highest final ratings.
Figure 2.2 shows a three-level hierarchy for the supplier evaluation problem
using AHP approach. In this figure, four criteria are used to evaluate suppliers and four
suppliers are being considered. After the hierarchy has been constructed, the next step is
to develop a set of pair-wise comparisons to define the relative importance for the
factors within each level of the hierarchy. Pair-wise comparisons use Saaty's scale (from
22
1 to 9, representing equal importance of both factors to absolute importance of one
factor over another, respectively) to measure importance among factors in each level.
Table 2.5 shows the pair-wise comparison matrix for criteria used in this example and
illustrates the procedure for calculating priorities of criteria at the second level. Note that
the matrix uses reciprocal values of scaled comparisons for matrix elements below the
diagonal. In the matrix, the priorities of the four factors with respect to the decision
objective are obtained as follows:
1. the elements of each column are divided by the sum of that column;
2. the elements in each resulting row are added; and
3. each row sum is divided by all the row sums.
Figure 2.2
Supplier Evaluation Hierarchies, Liu and Wu (2005).
23
Table 2.5: Pair-wise Comparison Matrix: Evaluation Criteria (Second Level in the
Hierarchy) (Kahraman et al.,2003)
Quality
Price
Service
Delivery
Priority
Quality
1
2
4
3
.457
Price
1/2
1
3
3
.300
Service
1/4
1/3
1
2
.138
Delivery
1/3
1/3
1/2
1
.105
Following the steps described above, one can determine the priorities of criteria
at different levels. In this example, each supplier (at the lowest level) has four weights
with respect to the four criteria at the second level. These weights are then multiplied by
the corresponding weights associated with each criterion at the second level, and the
multiplied results are summed across the criteria to get the final score for the supplier.
Repeating this synthesizing process, one can obtain the other suppliers' final scores and
can rank the suppliers based on their final scores.
The next common approach to multiple criteria-decision making that had been
used is Multi-Attribute Decision Making (MADM) (Moutaz and Booth, 1995). MADM
refers to making decisions in the presence of multiple, usually conflicting criteria.
MCDM problems are commonly categorized as continuous or discrete, depending on the
domain of alternatives. In MADM method, the decision maker needs to select the best
alternative in consideration. At the same time, the decision maker needs to satisfy the
criterion/parameters required for the selection.
24
b) Linear Weighting Method
Linear weighting method utilizes simple weighting in algorithms to rank supplier
performance factors subjectively, such as quality and delivery. Linear weighting studies
are characterized in this research as mathematical because they do not evaluate factors
but weight factors. Many studies use linear weighting methods supplier selection process
because of their uncomplicated execution process and the use of both qualitative and
quantitative factors. Stern and El-Ansary (1982) state that this method is flexible since
any number of evaluation factors that are felt important can be included. Under this
method, supplier performance must be expressed in numerical terms for each factor.
Also, the performance comparisons of suppliers require the same factors, weights, and
performance measurement standard all the time. The weighted-point method has the
advantage of being less intuitive although there is a certain degree of subjectivity state
that it is difficult to effectively take qualitative evaluation criteria into consideration.
The first step in this method is to assign weights to each performance factors that
are judged to be important by the buyer. The first step in this method is to assign weights
to each performance factor selected. The weights of the factors are assigned based on the
relative importance of factors. Then, a specific procedure is developed to measure the
actual supplier performance on each evaluation factor. Next, supplier performance
ratings are multiplied by their respective importance weight to get weighted values.
Finally these weighted values are totaled to get the overall rating for each supplier. The
supplier who has the highest overall rating is the best selection. Table 2.6 illustrates the
method.
In this example, three performance factors are selected – quality, price and
delivery – and the weights for quality, price and delivery are 35%, 15% and 50%,
respectively. The weighted values of performance are calculated and the overall rating is
summed as shown in the last column in the table. According to the overall rating, the
supplier Y receives the highest rating in performance; therefore, supplier Y is picked.
25
Table 2.6: Illustration of the weight-point method for supplier performance
Evaluation (Stern and El-Ansary,1982)
Supplier “X” Performance Evaluation
Factor
Weight
Performance
Performance Rating
(Weight X Performance)
Quality
35%
95% Acceptable
33.25
Price
15%
(Lowest Price Offered) / (Actual Price)
13.50
= ($90) / ($100)
= 90%
Delivery
50%
90% on schedule
45.00
Overall Evaluation : 91.75%
Supplier “Y” Performance Evaluation
Factor
Weight
Performance
Performance Rating
(Weight X Performance)
Quality
35%
98% Acceptable
34.30
Price
15%
(Lowest Price Offered) / (Actual Price)
15.00
= ($100) / ($100)
= 100%
Delivery
50%
92% on schedule
45.00
Overall Evaluation : 95.30%
Youssef et al. (1998) suggested that weighted point models are considered
excellent tools for several reasons. For example, the mathematics underlying weighted
point models are simple, the models can be adapted to virtually any type of purchase
decision, the models are relatively inexpensive to carry out about other models, and they
simplify optimal decision making. However, weighted point models also have some
drawbacks. Linear weighting methods are still considered biased because of their
excessive subjectivity. The linear weighting methods process relies heavily on human
judgment to weight different supplier’s factors. The main advantages of this method are
that the performance measures for the various factors must be in uniform units; in the
26
example, percentages are used. Furthermore, Nydick and Hill (1992) stated that it is
difficult to effectively take qualitative evaluation criteria into consideration.
c) Multivariate analysis
The use of Multivariate analysis is quite recent in a decision making method. The
multivariate techniques currently been used are the principal component analysis, factor
analysis, discriminant analysis and cluster analysis (Ganesalingam and Ganesh S, 2001).
In their paper, Lasch and Janker (2005) applied factor analysis for their supplier
selection method. They showed that the new system has been run using computerized
system and can be used for pre-qualification, selection and controlling of suppliers. They
proved that it is easy to handle and very practical without using manual criteriaweighting. So, the result is not 'misjudged' because it does not rely on human to weight
different supplier's factors.
The designed supplier rating system using multivariate analysis represents an
appropriate rating system which fulfils all demands of supplier rating. The
representation resulting from multivariate analysis can be used for pre-qualification,
selection and for controlling of the suppliers. In the case of this study, a proposed
method based on multivariate analysis will be constructed to measure supplier
performance evaluation and performance assessment. Multivariate analysis and its
application will be discussed in detailed in the next chapter, Chapter 3.
d) Statistical process control.
Statistical Process Control (SPC) is a methodology and philosophy for
monitoring a process to identify special causes of variation and signal the need to take
corrective action when appropriate. A practical definition of statistical process control is
that both process averages and variances are constant over time (Evans and Lindsay,
2002). Montgomery (2001) explains that SPC philosophy integrates an array of quality
27
tools designed to solve the problems that result in process variation and states that, “SPC
can be applied to any process”
Process capability indices (PCIs), one of tool of SPC, is used to provide a
numerical measure of whether a production process is capable of producing items
satisfying the quality requirement preset in the factory that had received substantial
research attention. Quantifying process potential and performance is important for any
successful quality improvement activity and quality program implementation. The
relationships between the actual process performance and the specification limits may be
quantified using appropriate process capability indices. The two capability indices which
have been widely used in manufacturing industry are Cp and Cpk. These two indices
provide numerical measures of whether a manufacturing process meets the preset
capability requirement (Chen et al., 2003).
The overview of all categories of Supplier Selection and Performance
Evaluation explained above can be summarized and shown in Figure 2.3.
.
Figure 2.3: Overview of Supplier Selection and Performance Evaluation Categories
28
2.5
Criteria for Supplier Performance Assessment
After a list of potential suppliers have been considered, the buying firm should
thoroughly evaluate suppliers. Suppliers are evaluated by a number of criteria
determined by the buying firm. Supplier evaluation is complicated because various
criteria are involved in the process and different companies may have specific
requirements. The following sections will review the literature on relevant criteria used
in supplier evaluation. Criteria can include various aspects ranging from quality issue to
cost elements.
It is surprising that only a relatively small number of studies have attempted to
identify the information that purchasing professionals may considered valuable.
Historically, performance checks have focused mainly on quantifiable aspects of the
supplier evaluation decision – such as cost, quality, delivery of incoming materials and
other similar factors. In order to provide a comprehensive view of the important criteria,
Dickson (1966) sent questionnaire to 273 purchasing American and Canadian agents and
managers selected from the membership list of the National Association of Purchasing
Managers. A total of 170 (62.3%) of questionnaire were received. Table 2.7 summarizes
the findings of the study regarding the importance of the 23 criteria for supplier
selection. The results shown in Table 2.7 demonstrate the inherently multi-objective
nature of the supplier evaluation process. The study found that purchasing professionals
consider the four top criteria to be quality, delivery, performance history, and warranties
and after sales services.
To perform assessment to suppliers, various types of criteria can be used. The
criteria can contain all the tangible factors which are easy to measure (quantitative data),
or they can be the intangible factors which are difficult to quantify or measure
(qualitative data). A firm should select the criteria that best serve its purpose.
29
Table 2.7: Criteria proposed by Dickson (1966)
Rank
Factor
Mean
Evaluation
Rating
1
Quality
3.508
2
Delivery
3.417
3
Performance history
2.998
4
Warranties and claim policies
2.849
5
Production facilities and capacity
2.775
6
Price
2.758
7
Technical capability
2.545
8
Financial position
2.514
9
Procedural compliance
2.488
10
Communication system
2.426
11
Reputation and position in industry
2.412
12
Desire for business
2.256
13
Management and organization
2.216
14
Operating controls
2.211
15
Repair service
2.187
16
Anitude
2.120
17
Impression
2.054
18
Packaging ability
2.009
19
Labor relations record
2.003
20
Geographical location
1.872
21
Amount of past business
1.597
22
Training aids
1.537
23
Reciprocal arrangements
0.610
Extreme Importance
Considerable importance
Average importance
Slight importance
30
In general, the few most common performance criteria included in supplier
assessment are quality, delivery, cost, and customer service. These criteria are discussed
in some detail by Vaniman (1998).
a) Quality
Supplier quality improvement and defect-free performance are characteristics
that are valued by customers across all industries. Quality performance can be expressed
in ways such as percent of defective or nonconforming, number of rejected lots, and
number of lots requiring rework or repair. In the quality term, any area that needs
measurement for potential attention and corrective action could be an element to be
considered in the supplier assessment design.
b) Delivery
To decrease inventories and implement just-in-time (JIT) concept, delivery of
goods by supplier is very important. Both early and late delivery will give different
values to customer. Early delivery increased carrying costs on the customer, and the
usable life of a dated material is consumed unnecessarily. Late delivery can translate
into schedule delays and higher costs. Examples of delivery measurements used in the
performance determination include percent on-time deliveries, percent early deliveries,
percent late deliveries, percent overage, and percent shortage.
c) Cost
A major criterion of the supplier evaluation process is the cost of doing business
with a specific supplier. The purchased or contracted cost for materials and components
does not represent the total cost of the procurement goods. Nonconforming materials,
expediting, stocking, and manufacturing downtime are just a few examples of non-value
added costs that are incurred as a result of poor quality and service. The performance
indices of the total cost approach are an effective way to determine total product cost
and communicate the total cost picture to the supplier.
31
c) Customer Service
Customer service contributes to a strong customer-supplier partnership. For this
reason, it is often an element that is assessed by customers during a supplier
performance assessment period. Communication, responsiveness, and technical support
are crucial to achieving customer satisfaction.
In addition, Kahraman et al. (2003) had included certain criteria for suppliers
selection. These criteria are:
i.
Financial - Suppliers should have a sound financial position to ensure that
performance standards can be maintained and that products and services will
continue to be available.
ii.
Managerial - Companies need to have compatible approaches to management,
especially for integrated and strategic relationships. Thus, the firm should have
confidence in its supplier’s management’s ability to run the company.
iii.
Technical - Technical support from suppliers is needed to provide a consistently
high quality product or service, promote successful development efforts and
ensure future improvements.
iv.
Support resource - The supplier’s resources need to be adequate to support
product or service development production, and delivery. This criteria need to
consider the supplier’s facilities, information systems, and provisions for
education and training.
v.
Quality systems and process- The supplier’s quality systems and processes
selection criteria may consider the supplier’s quality assurance and control
32
procedures, complaint handling procedures, quality manuals, ISO 9000 standard
registration status, and internal rating and reporting systems.
vi.
Globalization and localization. A firm’s sourcing strategy may recognize definite
advantages or disadvantages associated with choosing suppliers in a particular
region or country. The firm’s risk assessment should have identified potential
risks, such as currency fluctuations, shifts in political policy, and the
accompanying domestic or international regulatory and market changes that
result.
2.6
Previous Research Studies on Criteria Assessment for Supplier’s
Performance Evaluation
This section reviews on criteria selection by previous researchers with respect to
supplier’s performance evaluation. The majority of the studies presented in the supplier
evaluation and selection literature focused on the selection and overlooked the
evaluation of suppliers. The presented studies occasionally focused on how to better
evaluate and measure the suppliers’ criteria. Table 2.8 references the chronological study
of literature review. It references the supplier evaluation and selection studies presented
in this chapter. Also, Table 2.8 references the studies that included applied research;
either just theoretically or applicable in industry, methods to evaluate suppliers and
criteria considered during supplier evaluation.
Prahinski et al. (2001) addressed suppliers’ quality by considering the concept of
concurrent engineering. The authors focused on developing a model to achieve quality
through integrating manufacturing cost, quality loss cost, assembly yield, and process
capability index. The study stressed the importance of considering tolerance design in
33
selecting suppliers. Nonetheless, the model was still limited to design tolerances to
achieve good quality. It did not present how to evaluate suppliers’ quality.
Dahel and Nasr (2003) evaluated supplier performance from a cost of quality
perspective. The authors used five costs of quality categories: prevention cost, appraisal
cost, internal failure cost, external failure cost, and consequential costs of failure. These
costs were tracked throughout a supplier’s entire organization: purchasing, production,
design engineering, production supporting, and sales. This model looked at evaluating
the cost of suppliers’ quality not the quality of suppliers.
Muralidharan and Anantharaman (2001) presented more detailed suppliers’
criteria evaluation than other presented studies. The study had touched on several value
stream elements but from a macro level. The authors developed a questionnaire to
evaluate suppliers based on engineering technical capability, project management
expertise, material planning and production scheduling, production technology and
capability, commitment to continuous improvement and cost reduction, use of quality
tools, business structure, and management commitment to quality and teamwork.
Clearly, the quality evaluation focused more on the management aspect of suppliers’
quality rather than conformance to quality.
Youssef et al. (1998) assumed known lead-time in their model based on
historical data. The study focused on determining when to place orders to hit the due
date. Hence, no delivery evaluation was presented. Sucky (2001) developed a supplier
selection model considering the risk of several parameters. One of the parameters in the
model was the suppliers’ logistics complexity, which was described as a poor
optimization of the suppliers’ logistic network. The study focused on the selection model
development rather than the suppliers’ delivery evaluation.
34
Table 2.8 : Chronological summary of the literature review
Applied Research
Categories
Rating
Industry
Selection Criteria
Mathematical Methods
Methods
X
X
X
X
Demirtas and Ustun
2008
X
X
X
X
X
Sung and Krishnan
2007
X
X
X
X
Carr et al.
2007
Araz and Ozkarahan
2006
X
Amid et al.
2006
X
Liao and Rittscher
2006
X
Millington et. All
2005
X
Enriquez
2005
Aamer
2005
Najmi et al.
2005
Chen et al.
2005
X
Liu and Wu
2005
X
Zhang and Wang
2005
Peng and Chen
2005
X
Lasch and Janker
2004
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Supplier Criteria
2008
(Financial, QS,
Wan
Mgt, etc)
Cost
Quality
Delivery
SPC
Multivariate
Weightage
Indexing
Fuzzy sets
(AHP,MADM, etc)
Factors Costing
Factors Ranking
Other
Industrial
Equipment
Electronics
Theoritically
Automotive
Supplier evaluation
Year
Supplier Selection
Author(s)
Methods
35
Table 2.8 : continued
Brennan
2004
X
X
Kahraman et al.
2003
Schmitz and Platts
2003
Chen et al.
2003
X
Cebi and Bayraktar
2003
X
Hongwei et al.
2003
X
Tsai et al.
2002
X
Muralidharan et al.
2001
X
Prahinski
2001
X
Sucky
2001
X
Lacey and Elliot
2000
X
Krause et al
2000
X
X
Tsai
1999
X
X
Forker et al.
1999
Larson and Kulchitsky
1999
Giunipero et al.
1998
Li et al.
1997
Barua et al
1997
X
Dobler, D. W
1997
X
Hahn, et al
1996
X
Youssef et al.
1992
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
36
More studies focused on both quality and delivery. Cebi and Bayraktar (2003b)
assigned cost to suppliers’ performance parameters. Quality and delivery were among
the parameters in the model. The authors used very simple evaluation measurements for
both quality and delivery. Quality was evaluated by number of scraped parts.
Meanwhile, delivery was evaluated by on-time delivery, where 5 days early and two
days beyond deadline were considered on-time. No specific details were presented on
how to evaluate each factor. The study focused more on how to select a supplier given
the quality and delivery values.
Carr and Pearson (1999) presented a cost evaluation to measure supplier’s
conformance. The study assigned dollar value to the activities associated with the
resolutions of non-conformances. In addition, delivery was evaluated based on the cost
associated with transportation and delivery expediting. The model was very limited in
the parameters that evaluated suppliers’ quality and delivery.
Li et al. (1997) presented supplier evaluation metrics. Quality was evaluated
based on 4 criteria: number of returned or waste units/ units supplied, physical or
performance measurements carried out when goods entered the plant, replacement
guarantees, and certification. Delivery was evaluated based on average delivery time,
schedule average delay, and average gap between goods ordered and goods delivered.
Barua et al. (1997) also presented limited suppliers evaluation metrics.
Suppliers’ quality evaluation was based on 4 parameters: rejection rate, lot certification,
sorting effort, and defective acceptance. Delivery was also evaluated using compliance
with quantity, compliance with due dates, and compliance with packaging standards.
Giunipero et al. (1998) presented an activity based costing evaluation of quality
and delivery. The authors associated cost with quality audit, quality problems setup and
defects cost, inventory holding cost, and transportation cost. Nothing very specific was
37
presented on how to evaluate rather than capturing the cost of both quality and
transportation.
Lacey and Elliot (2000) reported an actual case study of selecting suppliers in the
telecommunication industry. Thus, evaluation processes were geared toward
telecommunication practices not manufacturing practices. Peng and Chen (2005)
presented a series of articles in the area of Artificial Intelligence. The authors included
delivery and quality as input of the supplier selection criteria in the intelligent model.
However, very generic referral to quality and delivery was presented. For example,
quality was referred to as rejection from customers, rejection in production line, and
rejection in incoming quality. Also delivery was referred to as compliance with quantity
and compliance with due date. The data used in the model did not evaluate the suppliers.
Again, the study focused more on the model than the evaluation itself.
Kahraman et al. (2003) also focused on developing a mathematical model more
than evaluating criteria. The authors mentioned that quality and service criteria were an
important part of the model, but just like most of the models the focus was more on how
to use the value rather than how to evaluate current supplier. Very general statements
were made about quality being measured by ISO 9000 and end user criteria.
Forker at al. (1999) examines supplier improvement initiatives (TQM and
supplier development practices) from two perspectives: the buying firm and the
supplying firm. The objective was to determine if there was matched agreement between
the two firms. Regarding the supplier development practices, they found that:
"Buyers were more inclined than suppliers to believe in: (1) the utility of their
supplier rating system; (2) their reliance on a few dependable suppliers; (3) the
usefulness of the technical assistance they extended to their suppliers; and (4) the
importance of quality (versus price or schedule) in their supplier selection decisions. On
the other hand suppliers gave a significantly higher rating than the buying firms to (5)
the clarity of the customer firm’s specifications"
38
The main strength of their research was the comparisons between the perceptions
of the two supply chain participants. However, in relation to this thesis, the authors overemphasized total quality management and underemphasize supplier development
practices.
Hartley and Choi (1996) focus on the barriers that hinder effective buyersupplier relationships with respect to quality management. They provide descriptive
statistics and anecdotal evidence based on questionnaires from 300 automotive suppliers
and in-depth interviews with two of the three major buying firms. They found the
barriers that include: poor communication and feedback; supplier complacency-,
complacency by the buying firm; the buying firm's credibility as perceived by their
suppliers; and misconceptions regarding purchasing power.
Liao and Rittscher (2006) use descriptive statistics to analyze supplier's
perceptions on the effectiveness of the automobile manufacturer's process-based
evaluation. They evaluate the three stages of the process-based evaluation: the
preparation stage, where the supplier prepares for the audit by ensuring documents are in
order; the evaluation visit, where the evaluator from the buying firm reviews the
documents, clarifies the process, and attempts to detect both business problems and
manufacturing problems that need to be rectified; and the feedback stage, where the
evaluator provides feedback and suggestions to the supplier. Based on 65 supplier
respondents, the preparation stage was considered the most important stage (61 percent
of respondents), and the feedback stage was considered the second most important stage
(33 percent of the respondents). This study offers a thorough analysis of the processbased evaluation from the supplier's perspective.
Aamer’s (2005) assessment study was to address suppliers’ selection from a
value stream perspective. It was an overall evaluation to the suppliers and it was still
very limited in the parameters and the areas it assessed. The most common parameters
found in Aamer’s assessment and the reviewed literatures fall under the following
quality and delivery categories:
39

Quality management system [ISO 9000]

Quality planning and assurance processes

Quality performance [such as PPM, Cpk]

Quality reliability [warranty cost, failure frequency, customer
complaint and serviceability]
quality problem solving methods

Quality safety parts management

Logistics system evaluation

Delivery precision [on-time percentage]
However, the literature lack an effective and comprehensive model for suppliers’
quality and delivery evaluation that takes into account the entire supplier value stream
practices. Aamer presented a result of survey to find the different evaluation methods or
factors considered across industry. The survey proved the lack of predictive measures of
suppliers’ selection and evaluation process. The majority of the presented methods in the
literature fall short to adequately and comprehensively evaluate suppliers’ quantitative
factors such as quality and delivery. Only a few direct practices of the two factors
quality and delivery are considered in the evaluation models. Indeed, there are other
subtle direct and indirect activities that contribute to quality and delivery conformance.
This study attempted to define these latent activities in the entire supplier value stream
to ultimately minimize the risk associated with suppliers’ performance.
In more details, this study addressed the risk associated with underestimating the
most critical factors. To overcome this issue and minimize the associated risk, this study
considered a holistic approach to evaluating suppliers’ conformance. The study
addressed the possible practices in the entire organization to examine their impact on
products’ conformance.
Araz and Ozkarahan (2006) have proposed a supplier evaluation and
management methodology for the product development process, in which suppliers are
categorized and compared according to their performances on several design based
criteria, potential reasons for differences in suppliers’ performances are identified, and
40
performances of the suppliers are improved by applying supplier development programs.
The proposed methodology considers the strategic partnership and concurrent product
development concepts to identify the supplier selection criteria rather than the traditional
selection criteria. They also describes a new fuzzy method, called as Preference Ranking
Organization Method Sort (PROMSORT) and based on the Preference Ranking
Organization Method for Enrichment Evaluations (PROMETHEE) methodology. They
utilized it in sorting the suppliers based on their preference relations, which assist
management in selecting suppliers for strategic partnership (“perfect” suppliers),
supplier development programs (“good” suppliers), competitive partnership (“moderate”
suppliers) and pruning (“bad” suppliers).
2.7
Summary
This chapter had reviewed the relevant criteria important to firms in selecting
suppliers and several evaluation techniques which were found in the literature. To pick
the best supplier for the buying firm’s objective, the firm should develop a strategy; that
is, develop a supplier which is “keep the best, and improve the rest”.
The rationale of this study was that the preceding most common evaluation
practices overlooked the supplier’s performance evaluation activities, which could have
tremendous impact on a product conformance. For instance, how the purchasing
department selects suppliers has tremendously influence with their product quality and
delivery conformance. If defective materials are bought, scraps and defects are
produced. This study attempts to reveal the most overlooked value stream practices and
prove their implication on conformance, and apply the practices into a model to evaluate
suppliers’ performance to minimize the risk associated when selecting evaluating
suppliers. The next chapter will discuss on multivariate analysis theory and its
applications.
CHAPTER 3
MULTIVARIATE ANALYSIS THEORY AND ITS APPLICATION
This chapter will present basic definitions of multivariate analysis and the
original set of factor analysis theory, as well as the steps of operation. Only those
concepts that are used in this study are described.
3.1
Introduction
As the name indicates, multivariate analysis comprises a set of techniques
dedicated to the analysis of data sets with more than one variable. Several of these
techniques were developed recently in part because they require the computational
capabilities of modern computers. Also, because most of them are recent, these
techniques are not always unified in their presentation, and the choice of the proper
technique for a given problem is often difficult.
The analysis is based on the statistical principle of multivariate statistics,
which involves observation and analysis of more than one variable at a time. In
design and analysis, the technique is used to perform analysis on a multiple
dimensions while taking into account the effects of all variables on the responses of
interest. Uses for multivariate analysis include (Johnson and Wichern, 2001):
42
a. Design for capability (also known as capability-based design);
b. Data reduction or simplification;
c. Analysis of alternatives, the selection of concepts to fulfill a customer need;
d. Sorting and grouping;
e. Prediction and hypotheses testing;
f. Investigation of the dependence among variables.
Multivariate analysis can be complicated by the desire to include physicsbased analysis to calculate the effects of variables for a hierarchical "system-ofsystems." Often, studies that wish to use multivariate analysis are stalled by the
dimensionality of the problem. These concerns are often eased through the use of
surrogate models, highly accurate approximations of the physics-based code. Since
surrogate models take the form of an equation, they can be evaluated very quickly.
This becomes an enabler for large-scale multivariate studies: while a Monte Carlo
simulation across the design space is difficult with physics-based codes, it becomes
trivial when evaluating surrogate models, which often take the form of response
surface equations (Pett et al. 2003)
There are techniques and applications of Multivariate analysis such as
Multivariate Linear Regression Models, Principal Components, Factor Analysis,
Canonical Correlation Analysis, Cluster Analysis and Multidimensional Scaling. In
this study, only Factor Analysis is used as the main analytical tool.
3.2
Factor Analysis
Factor analysis is a generic term used to describe a number of methods
designed to analyze the interrelationships within a set of variables. Factor analysis
has one feature in common, which is to construct a few hypothetical variables, called
factors, that are supposed to contain the essential information in a larger set of
observed variables (Dillon and Goldstein, 1984).
43
For example, there are several variables to choose from in order to buy a car.
Let’s say, that the input variables are: low-cost repair, roomy interior, comes in a
variety of colors, good handling, modern looking, high resale value, comfortable,
large power engine, eye catching and large trunk space. All these variables actually
have interrelationships with each other based on certain factors. In practice, one sees
that roomy interior, comfortable and large trunk space; can be concluded in one
factor called “comfortability”. For low-cost repairs, high resale value and large
power engine; can be conclude in another factor called “cost efficiency”. For the
modern looking, comes from variety of colors and eye catching; can be concluded in
another factor called “stylishness”. Thus, the description of these factors is related to
“comfortability”, “cost efficiency” and “stylishness”.
Before examining the mechanics of a factor analysis solution, it is important
to consider the characteristic of a successful factor analysis. One goal is to represent
relationships among sets of variables parsimoniously. That is, one would like to
explain the observed correlation using as few factors as possible. If many factors are
needed, little simplification or summarization occurs. One would also like the factors
to be meaningful. A good factor solution is both simple and interpretable. When
factors can be interpreted, new insights are possible (Norušis, 1994).
The factors are constructed in a way that reduces the overall complexity of
the data by taking advantage of inherent interdependencies. As a result, a small
number of factors will usually account to approximately the same amount of
information as do the much larger set of original observations. Thus, factor analysis
is, in this one sense, a multivariate method of data reduction.
Factor analysis attempts to “explain” the correlation by an analysis, which,
when carried out successfully, yields a small number of underlying factor, which
contain all the essential information about the correlations among the test.
Interpretation of the factors has led to the theory of fundamental aspects of human
ability.
44
3.2.1 Steps in Factor Analysis
Factor analysis, like many other statistical techniques, can be employed in
several ways. But in this case, factor analysis is a method for investigating whether a
number of variables of interest Y1 , Y2 , .... Yl are linearly related to a smaller number of
unobservable factors F1 , F2 ,..., Fk (Harman, 1976)
The distinctions between the various techniques of factor analysis are largely
connected with differences in these assumptions. The distinction can be presented in
two different manners, which are:
(i)
At the level of variances and covariances of the observed variables.
(ii)
At the level of intercorrelations between observed variables.
Finding the number of factors that can adequately explain the observed
correlations or covariances among the observed variables. The typical approach at
this stage is to input the relevant matrix into the factor analysis program and choose
one of the many methods of obtaining the initial solution.
The third step is the rotation to a terminal solution and interpretation. This
method is customary to applying rotation in an effort to find another set of loadings
that fit the observations and the factor structure equally well which can be more
easily interpreted. Any rotated factor solution explains exactly as much covariation
in the data as the initial solution. What is attempted through the rotation is a possible
“simplication” where there exist different criteria of simplicity which lead to
different method of rotation.
The last step is constructing of factor scales and their use in further analysis.
It is important to note that whatever method is used, the factor scales created are not
the same as the underlying factors and are error-prone indicator of the underlying
factor.
45
3.2.2 Examining the Correlation Matrix
One of the goals of factor analysis is to obtain factors that help explain these
correlations; the variables must be related to each other for the factor model to be
appropriate. If the correlations between variables are small, it is unlikely that they
share common factors.
Bartlett’s test of sphericity can be used to test the hypothesis that the
correlation matrix is an identity matrix; that is, all diagonal terms are 1 and all offdiagonal term are 0. The test requires that the data be a sample from a multivariate
normal population. If the value of the test statistic for sphericity (based on the chisquare transformation of determinant of the correlation matrix) is large and the
associated significant level is small, so it appears unlikely that the population
correlation matrix is an identity. If the hypothesis that the population correlation
matrix is an identity cannot be rejected because the observed significance level is
large, we should reconsider the use of the factor model.
Another indicator of the strength of the relationship among variables is the
partial correlation coefficient. If variables share common factors, the partial
correlation coefficients between pairs of variables should be small when the linear
effects of the other variables are eliminated. The partial correlations are then
estimates of the correlation between the unique factors and should be closed to 0
when the factor analysis assumptions are met. (Recall that the unique factors are
assumed to be uncorrelated with each other).
The negative of the partial correlation coefficient is called the anti-image
correlation. If the proportion of large coefficients is high, we should reconsider the
used of the factor model.
The Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy is an index
for comparing the magnitudes of the observed correlation coefficient to the
magnitudes of the partial correlation coefficients (Pett et al. 2003). It is computed as
46
2
ij
 r
KMO 
 r   a
i j
2
ij
i j
2
ij
i j
[3.1]
Where rij is the simple correlation coefficient between variables i and j , and
aij
is the partial correlation coefficient between variables i and j . If the sum of the
squared partial correlation coefficients between all pairs of variables is small when
compared to the sum of the square of correlation coefficients, the KMO measure is
closed to 1. Small values for the KMO measure indicate that the factor analysis of
the variables may not be a good idea, since the correlation between pairs of variables
cannot be explained by the other variables. According to Pett et al. (2003), the
Kaiser had characterizes measures in the KMO 0.90’s as excellent, in the 0.80’s as
meritorious, in the 0.70’s as middling; in the 0.60’s as mediocre, in the 0.50’s as
miserable, and below 0.50 as unacceptable.
A measure of sampling adequacy (MSA) can be computed for each
individual variable in a similar manner. Instead of including all pair of variables n the
summations, only coefficient involving that variable are included. For the i th
variable, the measure of sampling adequacy is
2
ij
r
MSAi 
j i
2
ij
r
j i
  a ij2
j i
[3.2]
These measures of sampling adequacy on the diagonal correlation matrix.
Again, reasonably large value is needed for a good factor analysis. Thus, we might
consider eliminating variables with small values for the measure of sampling
adequacy.
The squared multiple correlation coefficient between a variables and
all other variables. This value was called the communality which is the proportion of
variance explain by the common factors. Communalities can arrange from 0 to 1,
with 0 indicating that the common factors explain none of the variance and 1
47
indicating that the entire variable is explain by the common factors. The variance that
is not explained by the common factors is attributed to the unique factor and is called
the uniqueness of the variable.
3.2.3 Factor Extraction
The goal of factor rotation is to determine the factors. One will obtain
estimates of the initial factors from principal component analysis. In principal
component analysis, linear combinations of the observed variables are formed. The
first principal component is that the combination that accounts for the largest amount
of variances in the sample. The second principal component accounts for the next
largest amount of variance and is uncorrelated with the first. Successive components
explain progressively smaller portions of the total sample variance, and all are
uncorrelated with each other.
It is possible to compute as many principal components as there are variables.
If all principal components are used, each variable can be exactly represented by
them, but nothing has been gained, since there are as many factors (principal
components) as variables. When all factors are included in the solution, all of the
variance of each variable is accounted for, and there is no need for a unique factor in
the model. The proportion of variance accounted for by the common factors, or the
communality of variable, is 1 for all the variables. In general, principal component
analysis is a separate technique from factor analysis. That is, it can be used whenever
uncorrelated linear combinations of the observed variables are desired. All it does is
to transform a set of correlated variables to a set of uncorrelated variables (principal
components) (Johnson and Wichern, 2001)
To help us decide how many factors we need to represent the data, it is
helpful to examine the percentage of total variance explained by each. The total
variance is the sum of the variance of each variable. For simplicity, all variables and
48
factors are expressed in standardized form, with a mean of 0 and a standard deviation
of 1.
Several procedures have been proposed for the determining the number of
factors to use is a model. One criterion suggests that only factors that account for
variances greater than 1 (eigenvalue greater than 1) should be included. Factor with
variances less than 1 are no better than a single variable, since each variable has a
variance of 1. Although this is the default criterion in the SPSS Factor Analysis
procedure, it is not always a good solution (Norušis, 1994)
3.2.3.1 Methods for Factor Extraction
Several different methods can be used to obtain estimates of the common
factors. These methods differ in the criterion used to define “good fit”. Johnson and
Wichern (2001) details pertaining on this:
(i)
Principal-axis factoring precedes much the same as principal
component analysis, except that the diagonals of the correlation
matrix are replaced by estimates of the communalities. At the first
step, square multiple correlation coefficients can be used as initial
estimates of the communalities. Based on these, the requisite
number of factors is extracted. The communalities are estimated
from the factor loadings, and factors are again extracted with the
new communality estimated replacing the old. This continues until
negligible change occurs in the communality estimates.
(ii)
The unweighted least-squares method produces, for a fixed
number of factors, a factor pattern matrix that minimizes the sum
of the squared differences between the observed and reproduced
correlation matrices (ignoring the diagonals).
49
(iii)
The generalized least-squares method minimized the same
criterion; however, correlations are weighted inversely by the
uniqueness of the variables. That is correlation involving variables
with high uniqueness are given less weight than correlations
involving variables with low uniqueness.
(iv)
The maximum-likelihood method produces parameter estimates
that are the most likely to have produced the observed correlation
matrix is the sample is from a multivariate normal distribution.
Again, the correlations are weighted by the inverse of the
uniqueness of the variables, and an iterative algorithm is employed.
(v)
The alpha method considers the variables in a particular analysis
to be a sample from the universe of potential variables. It
maximizes the alpha reliability of the factors. This differ from the
previously describe methods, which consider the cases to be a
sample from some population and the variables to be fixed. With
alpha factor extraction, the eigenvalues can no longer be obtained
as sum of the squared factor loadings, and the communalities for
each variable are not the sum of the square loadings on the
individual factors.
(vi)
In image factoring, the common part of the variable is defined as
its linear regression on remaining variables, rather than a function
of hypothetical factors. This common part is called a partial image.
The residual about the regression, which represents the unique part
of a variable, is called a partial anti-image.
50
3.2.4 The Rotation Phase
Although the factor matrix obtained in the extraction phase indicates the
relationship between the factors and individual variables, it is usually difficult to
identify meaningful factors based on this matrix. Most of the times, variables in any
factors seems do not correlated in any pattern. Found that most factors are correlated
with many variables. Since one of the goals of factor analysis is to identify factors
that are substantively meaningful (in the sense that they summarize sets of closely
related variables), the rotation phase of factor analysis attempt to transform the initial
matrix into one that is easier to interpret.
The purpose of rotation is to achieve a simple structure. This means that one
would like each factor to have non-zero loadings for only some of the variables. This
helps on interpreting the factors. One would also like each variable to have non-zero
loadings for only a few factors, preferably one. This permits the factors to be
differentiated from each other. If several factors have high loading on the same
variables, it is difficult to ascertain how the factors differ.
Rotation does not affect the goodness of fit of a factor solution. That is,
although the factor matrix changes, the communalities and the percentage of total
variance explained do not change. The percentage of variance accounted for by each
of the factors does change, however. Rotation redistributes the explained variance for
the individual factors. Different rotation methods may actually result in the
identification of somewhat different factors (Johnson and Wichern, 2001). When the
axes are maintained at the right angles, the rotation is called orthogonal. If the axes
are not maintained at right angles, the rotation is called oblique.
A variety of algorithms is used for orthogonal rotation to a simple structure,
such as;
51
(i)
The most commonly used method is the varimax method, which
attempts to minimize the number of variables that have high
loadings on a factor. This should enhance the interpretability of the
factors.
(ii)
The quartimax method emphasizes simple interpretation of
variables, since the solution minimizes the number of factors needed
to explain a variable. A quartimax rotation often results in general
factors with high-to-moderate loadings on most variables. This is
one of the main shortcomings of the quartimax method.
(iii)
The equamax method is a combination of the varimax method,
which simplifies the factors, and the quartimax method, which
simplifies variables.
A convenient means of examining the success of an orthogonal rotation to
plot the variables using the factor loadings as coordinates. When the factor solution
involves more than two dimensions, SPSS produces a three-dimensional plot of the
first three factors.
To identify the factors, it is necessary to group the variables that have large
loadings for the same factors. Plots of the loadings are one way of determining the
clusters of variables. Another convenient strategy is to sort the factor pattern matrix
so that variables with high loadings on the same factor appear together. Small factor
loadings can be omitted; considered that the no loading less than 0.5 in absolute
value are displayed.
Oblique rotation preserves the communalities of the variables, as does
orthogonal rotation. When oblique rotation is used, the factor loadings and factor
variable correlations are no longer identical. The factor loadings are still partial
regression coefficients, but since the factors are correlated, they are no longer equal
to the simple factor variables correlations. Therefore, separate factor loading and
factor structure matrices are displayed as part of the output. (Norušis, 1994)
52
3.2.4.1
Factor Scores
Since one of the goals of factor analysis is to reduce a large number of
variables to a smaller number of factors, it is often desirable to estimate factor scores
for each case. The factors score can be used in subsequent analysis to represent the
value of the factors. Plots of factor scores for pairs of factors are useful for detecting
unusual observations.
There are several methods for estimating factor score coefficient. Each has
different properties and results in different scores (Harman, 1976). The three
methods available in the SPSS Factor Analysis procedure (Anderson-Rubin,
regression and Barlett) all result in scores with a mean of 0. The Anderson-Rubin
method always produces uncorrelated scores with a standard deviation of 1, even
when the original factors are estimated to be correlated. The regression factors scores
(the default) have a variance equal to the square multiple correlation between the
estimates factor scores and the true factor values. Regression method factor scores
can be correlated even when factors are assumed to be orthogonal. If principal
components extraction is used, all three methods result in the same factor scores,
which are no longer estimated but are exact.
3.3
Summary
The supplier evaluation process is a multi-variate decision making problem.
Since multivariate analysis is used in this study, this chapter has reviewed some basic
concepts of multivariate analysis. This chapter also describes original set of factor
analysis theory, as well as the steps of operation. It explains about multivariate
analysis, its uses and application. Details of steps for conducting Factor Analysis, the
multivariate tool used in this research are explained briefly. The next chapter will
discuss on research methodology.
CHAPTER 4
RESEARCH METHODOLOGY
4.1
Introduction
This chapter describes the methodology used in this research in order to
achieve the objectives of the study. This chapter provides in detail the procedure to
develop the instrument for supplier performance evaluation and assessment in
automotive industry.
4.1.1
Overview of Research Methodology
The research was carried out in four stages which comprise critical of
literature review, survey methodology, data analysis and development of the
assessment tool. First stage, literature review consists on identification of current
automotive industry scenario, problem issues on supplier evaluation and
development. It also discussed on supplier evaluation methods by researchers and the
criteria upon it. Literature review has been discussed in detail in Chapter 2 and 3.
54
Second stage is the survey methodology. Data was collected through
questionnaires, visits and interviews from the relevant authorities. Questionnaires
have been developed, validated and pre-survey before they were sent to the concern
parties. On the third stage, data that have been collected was analyzed and tested for
validation. Initially, reliability test was conducted to validate the questionnaire before
it applied for multivariate analysis at later stage. Then, test such as on-response bias
hypotheses, t-test and standard deviation test were conducted on the data to conform
the validation of data entry.
After that, the data was analyzed using multivariate analysis. For this
research, multivariate analysis method that used is Factor Analysis by using SPSS
Software. Results for Kaiser-Meyer-Olkin (KMO) measurement, Bartlett’s test,
communalities, and Scree plot are incorporated in it. The factor analysis results are
then given interpretation and used to produce assessment tool for automotive
manufacturer. The flowchart in Figure 4.1 shows the overview on how the study was
conducted.
55
Start
Supplier Performance & Development
Issues (Section 2.2 &2.3)
Literature Review
Identification of current automotive
industries scenario
Data Collection from Supplier
Multivariate Analysis
(Factor analysis)
Design the Supplier
Performance Assessment
Evaluation
satisfactory
Method & Categories (Section 2.4)
-Designing Questionnaire (Section 4.2.2)
- Expert Validation & Pilot Survey
(Section 4.2.3)
- Data Collection (Section 4.3)
-Reliability Test (Section 5.4)
-Non-Response Bias Hypotheses
(Section 5.5.1)
-T-Test (Section 5.5.2)
- Data Validation (Section 5.6)
-Data Processing with SPSS
(KMO and Bartlett’s Test,
Communalities, Total Variance
Explained) (Section 5.7)
- Scree Plot, Rotated Component Matrix
- Interpreting Result (Sec 5.8)
-Design of Tool for Supplier
Performance Assessment (Sec 5.9)
Conclusions
Figure 4.1: Overview of the Research Methodology
56
4.2
Survey Methodology
4.2.1 Introduction
To determine the current practices in the Automotive Industry regarding
implementation of quality practices and procedures, it was decided that a mail survey
would provide the desired information in a timely manner. This method is chosen
due to the advantage that the designed questionnaire could be sent to a large number
of organizations in a limited time. The objective of this survey is to find out level of
quality practices implemented in the supplier firms. The methodology begins with
designing the questionnaire and then followed by data collection via pilot study and
finally full survey.
4.2.2 Survey Questionnaire Development
Survey questionnaire was decided to be developed as the instrument for
collecting data. At this point, mention should be made as to how this was determined
and research methodology in general. The researcher most decide the following: 1)
clearly define the goal, and 2) define exactly what the research is to accomplish
(Sanders, 1995). Since the former case was deemed to be appropriate for the purpose
of this study, it was determined that any samples gathered must meet the following
criteria:
i.
be representative of the studied population,
ii.
be of a sufficient number for statistical purposes, and
iii.
be gathered via a neutral or unbiased mechanism.
57
The survey was designed with the following criteria in mind:
1) simplicity (majority of questions were 1-5 Likert type scale),
2) readability (a large plain font was used),
3) conciseness (questions were kept to a minimum and were one line or less in
length), 4) visually appealing and
5) ease of return (self addressed stamped envelopes were included along with this
author's post office/e-mail addresses repeated on each side of the survey).
Data gathered via this manner would be accurate, up to date, and relevant to the
study being undertaken (Sanders, 1995).
The survey questions dealt with two prime areas which are background
information and implementation of quality practices and procedures. As explained in
earlier chapter, many studies have identified five factors of quality implementation;
which are Quality System, In Process Quality, Logistics and Management, Shipping
and Delivery and After Sales Service.
The developed questionnaire (attached in Appendix A) has been divided into
two parts. Part A is asking on general information of the organization and part B has
been divided into 5 sections that attempt to find out factors that are practiced for
quality implementation in the suppliers firm.
Part A: General profile of companies participating in this research
Part A has a total of 6 questions. Two types of questions have been used;
multiple choices and open-end questions. The selection of question types depended
on the aim or the objective that needs to be achieved by each question. Part A
focuses on background information of company which includes type of product and
activities, organization type and main customers. It also has questions which includes
customer’s criteria when assessing their performance. Questions about accreditation
and quality certification/awards are also included in this section.
58
Part B: Assessment Questions
Part B focuses on the respondent’s perception of quality implementation
practices in the suppliers firm. It is divided into 5 sections as follows:
Section 1 : Quality System
Section 2 : In Process Quality
Section 3 : Logistics and Management
Section 4 : Shipping and Delivery
Section 5 : After Sales Service
A five point Likert type scale was used for all items in different sections. This
scale was suitable to rate all relevant alternatives along the continuum for the
respondents to express their opinions (Ahmed and Hassan, 2003).
4.2.3 Expert Validation and Pilot Study of Survey
With self-administered questionnaires, it is important to minimize confusion
by being extremely clear in the instructions and questionnaire design. It has been
found that respondents with mail surveys rarely contact the researcher to ask
questions (Czaja et al., 1996). To minimize the opportunity for confusion, a pilot
survey was conducted. The purpose of the pilot survey is to refine the measurement
scales and validate the sample viability in meeting the objectives of the research.
Engineers and experienced researchers were selected to respond to the written
draft of the final field survey. Each was contacted to provide information regarding
their experiences and comments about the quality of the questionnaire, research
design, and other relevant points. Five engineers/assistant managers working at firsttier automotive suppliers provided an in-depth understanding of the actual buyersupplier relationship, the communication between the buying and selling firms, the
59
evaluation process, and their business performance. Feedback and responses from
them had been incorporated to the final questionnaire. Some of the initial questions
were eliminated or combined as per their suggestion. Initially, there were four
segments in the questionnaire, but one of the experts has suggested combining three
of the segment to one. In his opinion, this will give psychological effect that there’s
not so many question to answer; so that the rates of response will increase during the
full survey. Three researchers had assessed format issues, comprehension and clarity
of questions. The researchers have also proof read the questionnaire by checking the
spelling, sentence structure, and other grammatical issues.
The process resulted in modifying selected wording on the survey
questionnaire, revising the survey format, eliminating few questions for this research.
The final survey instrument was five pages in length and comprised of two parts A
and B. Part A is asking on general information of the organization and part B has 67
questions that attempt to find out factors that are practiced for quality
implementation in the suppliers firm. After the survey instrument was redesigned, the
survey was mailed to managers at automotive suppliers.
4.3
Data Collection
4.3.1 Mail surveys
Mail survey was conducted to collect data in this study. Mail surveys are used
because they are relatively inexpensive way to collect quantitative data. Another
value in using mail is that one can contact respondents who might otherwise be
inaccessible (Cooper and Schindler, 2003). Response rates vary but are usually
between 5 and 10 percent data (Gervitz, 1994). Higher response rates occur when the
participants have a special interest in participating in the research.
60
Due to the information necessary for this research, the individual most
involved in the quality management system and supplier management was selected
as the respondent because they are more aware of the quality system and problems
faced in the company. As for car manufacturers, the individual responding to the
survey was either the quality manager, head of department or supplier quality
assurance personnel/engineers to fill up the questionnaire. These supplier firms
represent some of the most critical suppliers for major Malaysia automobile
manufacturers: Proton, Perodua, Toyota and Honda.
4.3.2 Sampling Frames
Three sampling frames were used for this survey.
1. List of Suppliers (2007) from Proton Holdings Berhad. This consisted total of
228 suppliers which 201 of them are local suppliers.
2. List of UMW-TOYOTA Sdn Bhd Local Vendors (2006). It have total of 60
local suppliers.
3. List of Honda (M) Sdn Bhd Local Vendors (2007). It have total of 74 local
suppliers
Some of the companies are also supplier for not just one automotive manufacturer.
Some of them are even a supplier for all major automotive manufacturers in
Malaysia. From the list, total of 278 companies have been chosen for the full survey.
61
4.3.3 Population of the study
Population size used in as shown in Table 4.1.
Table 4.1 Population Size
No of
Questionnaire
No of
for each
Population
Main Business
Company
company
Population
Automotive
Electrical Components
71
5
356
Suppliers in
Interior Components
62
5
307
Malaysia
Chassis Components
11
5
178
Exterior Components
10
5
161
Accessories
8
5
129
Stampings
5
5
81
Others
11
5
178
Total
Grand Total
278
1390
The population in this study is from Automotive Suppliers in Malaysia. A
total of 278 suppliers were identified from list obtained from few automotive
manufacturers. Every company is mailed 5 questionnaires each. And grand total of
population of the study is 1390. The population of selected Automotive Suppliers is
including medium companies and larger companies (More than 50 employees).
Companies categorized as small companies are those with the number of employees
less than 50 while medium companies have employee between 50 and 300. The
companies with the approximate number of employees above 300 are categorized as
large companies (Hashim and Wafa, 2001)
62
4.3.4 Management for Non-response and Steps to Increase Response Rates
To easily manage the response of the questionnaires, each survey included an
identification number, printed in the upper right-hand corner of the front cover. This
identification number was used to track returned responses and reduced the total
number of contacts for those respondents who provided an early return. The
respondents were assured on the confidentiality of their responses, meaning that all
their responses, feedback and information would not be connected with their
company or themselves. This is to make sure their response and feedback are more
sincere and produce unbiased results.
In order to reduce the amount of non-responses, the following methods have
been implemented. One of the efforts is enclosing a stamped addressed reply
envelope to encourage the respondents to reply the survey. A short reminder letter,
have been sent about 15 days after the questionnaire was mailed to the companies
that have not replied. Second follow-ups consisting a short letter together with the
original letter, another copies of questionnaire and return envelope have been sent
about 10 days after the first reminder to those companies who did not respond to the
first reminder.
After 6 weeks from the initial mailing date, the companies that have not
replied will be considered as non-response. All the survey questionnaire feedbacks
were collected and then analysis of the data been proceed. The analysis of the data
shows in the next section.
63
4.4
Data Analysis
4.4.1 Reliability
The basic idea of reliability is consistency. The goal of reliability is to
minimize the errors and biases in a study (Yin, 1984). Reliability refers to the
demonstration that the operations and procedures of the research inquiry can be
repeated by other researchers which then achieve similar findings, that is, the extent
of findings can be replicated (Riege, 2003). Reliability is concerned with estimates of
the degree to which a measurement is free from random or unstable errors. Reliable
instruments can be used with confidence that transient and situational factors are not
interfering (Cooper and Schindler, 2000).
Each of the reliability technique leads to a single numerical index called a
reliability coefficient. Reliability coefficients of the data's consistency normally
assumes a value somewhere between 0.00 and +1.00, with these two "endpoints"
representing situations where consistency is either totally absent or totally present
(Huck and Cormier, 1995). Due to the distinction of time and condition, there are
generally three methods used to assess the reliability: test-retest reliability, alternate
forms reliability and internal consistency reliability (Huck and Cormier, 1995:
Cooper and Schindler. 2000).
The most commonly used measure of reliability is internal consistency, which
applies to the consistency or homogeneity among the variables in a summated scale.
Assessments of internal consistency focus on the degree to which the same
characteristic is being measured. For example, a Likert-type questionnaire where five
response options for each statement extend from -strongly agree" to "strongly
disagree': and are scored with the integers 5 through. The high correlation determines
the similarity among the items. Cronbach's alpha is expressed as a - correlation
coefficient, ranging in value from 0 to +1. Nunnally (1978) has suggested that the
generally accepted standard for reliability estimates is above 0.70.
64
The reliability of the test instrument in this research is tested by using
Cronbach's Alpha. Cronbach's alpha being the most widely used measure for multi
item scales at the interval level of measurement (Nunnaly, 1978: Huck and Cormier,
1995: Cooper and Schindler, 2000). Cronbach's value is more versatile because it can
be used with instruments made up of the items that can be scored with three or more
possible value. (Huck and Cormier, 1995).
4.4.2 Multivariate Analysis
4.4.2.1 Using Factor Analysis for Supplier Performance Assessment and
Evaluation
Factor analysis, one of a method for multivariate analysis will be used. Figure
4.2 shows that Factor Analysis can be used for Supplier Performance Assessment
and Evaluation. It can also be used for Supplier Control. The biggest advantage for
using it is it requires minimal manual interferences (Lasch and Janker, 2005).
65
Purchasing Situation
Ideal
supplier
Criteria
Factor Analysis
Supplier Evaluation
OBJECTIVE:
Use of a multivariate method
without manual interferences
Supplier Controlling
Figure 4.2: Using Factor Analysis as a Supplier Evaluation and Controlling
4.4.2.2 Steps in Using Factor Analysis in Supplier Performance Assessment and
Evaluation
As shown in Figure 4.3, Factor analysis usually proceeds in few steps:- (All
mathematical procedures will be realized by the statistical standard tool SPSS)
i.
In the first step, the correlation matrix for all variables is computed.
Variable that do not appear to be related to other variables can be
identified from the correlation matrix and associated statistics. The
appropriateness of the factor model can also be evaluated.
ii.
In the second step, goodness of fit test will be done to confirm that Factor
Analysis can be done to the data or not. One of the goals of factor
analysis is to obtain factors that help explain these correlations; the
66
variables must be related to each other for the factor model to be
appropriate. If the correlations between variables are small, it is unlikely
that they share common factors. The common Goodness of fit test that are
using are Bartlett’s test of sphericity and Kaiser-Meyer-Olkin (KMO)
test.
iii.
Communalities are measures of the amount of variance in each variable
that the factors explain. We are only interested in communalities after
extraction in principal Component Analysis. A low values (<0.1)
indicates that the variable should be omitted from the analysis.
iv.
Next, factor extraction, where the number of factors necessary to
represent the data and the method for calculating them must be
determined. The result will be represent by Scree Plot and Component
Matrix.
v.
The fifth step is rotations which focus on transforming the factor to make
them more interpretable.
vi.
At the fifth step, scores for each factor can be computed for each case.
These scores can then be used to build cluster and then we can have a
choice of the cluster containing the ideal supplier.
67
Data
Correlation
Matrix
Goodness of fit test
[Kaiser-Meyer-Olkin (KMO) and
Bartlett's Test of Sphericity]
analysis can't be
done; data not
appropriate for
Factor Analysis
no
yes
Communalities
Check
yes
no
Factors Extraction
(Principal Component
Analysis)
Results
(How many
factors)
Scree Plot
Component
Matrix
Cluster Building
Choice of cluster
containing ideal
supplier
Figure 4.3: Factor analysis steps
Omitted data
that have low
value (<0.1)
68
4.5
Summary
This chapter has described in detail the research methodology employed to
collect and analyze the necessary data in order to meet the objective of the study.
Postal questionnaire approach was adopted in this study. The aim of this study is to
show that factor analysis can be used for supplier performance assessment in
Malaysian automotive industry, and to proposed new instrument that can simplify the
process of supplier assessment by automotive manufacturers. In the next chapter,
results and analysis will be presented.
CHAPTER 5
SURVEY RESULTS AND ANALYSIS
5.1
Introduction
In chapter 4, the research design and methodology were discussed. The
characteristics of the sample and the overall assessment of the measurement model
will be described in this chapter.
The survey yielded 338 returned usable
questionnaires for a 24.3% response rate. It starts with presenting descriptive statistic
of respondents on the general information about company such as their main
products, certification, etc. in the first section.
In the second section, the data were examined for data entry errors, missing
data and outliers. In the third section, the measurement model is assessed to
determine the fit. Based on the correlation matrix, factor extraction and fit indices;
the constructed model was developed to be used in the case study. Based on the
measurement model fit, 41 items were retained and 22 were dropped. The
constructed model shows as the final result at the end of the chapter.
70
5.2
Response rate
The target population in this survey is all automotive industry suppliers in
Malaysia. From list obtained from few automotive manufacturers, a total of 278
suppliers were identified. A total of 1390 questionnaires were mailed to these
companies in which each of the company have been mailed 5 questionnaires. At the
end, a total of 346 were returned which represented respond rates at about 24.9%.
Found that the 346 surveys returned were actually come from a total of 86
companies. The number of returned questionnaires that were found to be usable in
this study was 338, which represented about 24.3%
Response rate is relatively high compared to previous authors. From review
of literature studies, this gave low response rate such as Ahmad and Hassan
(11.15%). Since the response rate is high, it shows that the respondents are interested
to this research and participated on it.
The author also believed that steps on
increasing response rates (as per discussed in Chapter 4) such as stamped address
reply enveloped and two times reminder to the suppliers that not yet reply have given
positive results. It is also believed that non-respondent bias does not exist. Table 5.1
provides a summary of the total response organized by frequency of responses.
Table 5.1: Summary of responses
No Frequency of responses
No. of Companies
Total
1
1
5
5
2
2
2
4
3
3
18
54
4
4
22
88
5
5
39
195
Total No. of Companies
86
Total Respondents
346
Not-usable (Rejected)
Total Usable
8
338
71
5.3
Descriptive Statistics of Respondents
The first aspects to be investigated were the general information of the
companies such as company’s main products, company’s certification, etc. Using
descriptive statistics, the results are summarized in tables and figures in the form of
percentage and frequency. From the results, information about characteristics of the
respondent can be identified such as classification of companies, experience and
maturity in implementing quality, etc.
From 86 companies, we found that two of the firms are actually a distributor
for automotive manufacturer and the rest are parts/products supplier. Survey result
from this two company has been isolated because of the result of their questionnaire
is actually from their overseas supplier and it is beyond the scope of this research.
From these 2 companies, found that 5 survey results that they responded need to be
opting out from the analysis because of it.
Table 5.2 Classification of the supplier
Classification
Frequency
Percent.
Cumulative
Percent.
Manufacturer
84
97.7
97.7
Distributor
2
2.3
100
Service Provider
0
0.0
100
Designer
0
0.0
100
Software Developer
0
0.0
100
Others
0
0.0
100
Total
86
100.0
The respondents represented a wide variety of business that supports the
automotive industry. The diversity of the respondents based on their main business as
suppliers to the automotive manufacturer is shown in Table 5.3. As can be seen, the
predominant main businesses of the respondents are to supply electrical components
72
(25.6%). Found that suppliers for interior components and chassis component gave
the largest responses for the surveys after electrical components supplier. Supplier
for accessories and stamping made the smallest portion, as only 13% from the total
number of responded companies.
Table 5.3 Supplier’s Main Business to the Automotive Manufacturer
Cumulative
Main Business
Frequency Percent Percent
Electrical Components
22
25.6
25.6
Interior Components
19
22.1
47.7
Chassis Components
11
12.8
60.5
Exterior Components
10
11.6
72.1
Accessories
8
9.3
81.4
Stampings
5
5.8
87.2
Others
11
12.8
100.0
Total
86
100
The length of time the supplier has been conducting business with the
automotive manufacturer indicates the understanding of the respondent with the
requirements demanded by the automotive company. Table 5.4 shows that 87% of
the respondent companies had conducted their business with the automotive
manufacturer for more than 5 years. It means that the data shows a high degree of
confidence because the longer of time in conducting business together, the suppliers
can fulfill more requirements demands by the automotive manufacturer.
73
Table 5.4 The length of time in conducting business
Years of business
relationship
Cumulative
Frequency Percent Percent
Less than 1 year
0
0.0
0.0
1-2 years
0
0.0
0.0
3-5 years
11
12.8
12.8
5-10 years
42
48.8
61.6
More than 10 years
33
38.4
100
Total
86
100.0
In terms of quality certification, almost 91.9% of the respondents have MS
ISO 9001:2000 certification as shown in Table 5.5. This is followed by TS 16949
certification with 65.1%, MS ISO 14000 with 60.5 and OHSAS certification with
43.0. Most of the companies have complied with more than one certification. Some
companies are in the process of certifying to MS ISO 14001 and TS 16949. The
survey result shows that automotive suppliers have high certification achievement
ratio especially in MS ISO 9001 and TS 16949 to ‘maintain’ their business
relationship with the automotive manufacturers. These certifications are likely part of
the requirements by most of the automotive suppliers nowadays.
Table 5.5 Quality Certification
ISO 9001 TS16949 OHSAS ISO14000 OTHERS
N (No. of
companies)
% of N
79
56
37
52
4
91.9
65.1
43.0
60.5
4.7
74
5.4
Reliability Test
For the survey questionnaire, a test to assess its reliability must be done
before it is applied for multivariate analysis. The Cronbach’s alpha coefficient was
used to test the reliability of the survey questionnaire. The Cronbach's alpha
measures how well a set of items (or variables) measures a single unidimensional
latent construct (Nunnally, 1978). It determines the extent to which items within a
factor are related to each other and assists to identify problem items that should be
excluded from the scale so as to improve reliability. Cronbach's alpha will generally
increase when the correlations between the items increase. According to Nunnally
(1978) he suggested that generally accepted standard for reliability estimates is above
0.70. Using SPSS software, reliability analysis procedure was performed separately
for the items of each section. The results are shown at Appendix C. The summary of
reliability is given in Table 5.6.
Table 5.6 Internal consistency of the questionnaire
Section
No. of items
Alpha Value
1
Quality
19
0.891
2
In Process Quality
8
0.873
3
Logistics and management
18
0.792
4
Shipping and Delivery
12
0.926
5
After Sales Service
5
0.884
As can be seen in the table, the lowest Cronbach’s alpha value is 0.792 and
the highest alpha value is 0.926. It indicates that all the scales are acceptable because
all the Cronbach’s alpha values are above 0.7. It means the questionnaire from the
study was reliable. The overall instrument reliability is 0.875.
75
5.5
Non-Response Bias
To minimize the opportunity for non-response bias, it is important to obtain
the highest possible response rates (Lessler and Kalsbeek, 1992). As mentioned in
section 4.3.4, to reduce the amount of non-responses few action has been taken. If
there’s no reply after the initial survey questionnaires, a reminder letter will be sent.
If still no reply after that, second follow-up consisting a short letter together with the
original letter and another copies of survey questionnaire to those companies who did
not respond to the first reminder. A follow-up phone call also had been done to those
companies who yet reply any responses.. In addition, Dillman (2000) suggestions to
personalize the cover letter, include a self-addressed postage-paid return envelope,
assure confidentiality, and create an attractive and easy-to-understand survey
Once the opportunity for non-response was minimized, evidence of nonresponse bias was assessed by comparing answers between questionnaires that were
returned early and those returned late to determine if there were statistical differences
(Lessler and Kalsbeek, 1992). The sample was split into two groups based on if they
were returned before or after the mailing of the second survey package. Responses
for first survey were 207 and second survey were 131.
5.5.1 Hypothesis
The sample was split into two groups based on if they were returned before or
after the mailing of the second survey package. Responses for first phase survey
were 207 and second phase survey were 131. Twelve of the sixty two survey items
(20%) were randomly selected and t-tests were performed on each item (n1=207,
n2=131), as shown in Table 5.7.
76
Hypotheses for test are:
H0: There are no significant differences between early and late responses
H1: There are significant differences between early and late responses.
5.5.2 T-test for Equality of Mean
Results from independent t-test for equality of means (refer Table 5.7), gave
all the p-value above 0.05. If the confidence interval did not include zero, the null
hypothesis that there is no statistical difference would be rejected. The t-tests yield
that there are no significant differences between early and late responses. It shows
that early/late respondent bias does not exist. So, both of the results from Early and
Late responses can be used together for data analysis. Therefore, the data analysis
preceded with scale purification and testing of the measurement model as described
in the next section.
77
Table 5.7
Comparing Early to late Respondents
Elements
Q1B
Q2A
Q4A
Definition of responsibility make sense for the
process involved
All quality methods are documented and used
as a basis for establishing quality
Suppliers encouraged using SPC.
P-
µ
σ
µ
σ
Early
Early
Late
Early value Result
4.135
0.738
4.191 0.745
0.251 Not Sig
4.271
0.797
4.359 0.785
0.160 Not Sig
3.667
1.047
3.756 1.060
0.225 Not Sig
4.077
0.699
4.130 0.684
0.249 Not Sig
4.043
0.699
4.115 0.675
0.179 Not Sig
4.082
0.749
4.107 0.736
0.383 Not Sig
4.077
0.656
4.115 0.675
0.308 Not Sig
3.995
0.873
3.992 0.941
0.489 Not Sig
4.063
0.764
4.107 0.787
0.305 Not Sig
3.845
1.026
3.878 1.067
0.390 Not Sig
4.155
0.734
4.206 0.731
0.265 Not Sig
3.952
0.742
4.000 0.765
0.282 Not Sig
A definite program to bring about continual
Q7A
improvement in quality and productivity is
planned and being carried out.
Q8A
Process/product
auditing
functions
and
responsibilities are clearly defined
Gages and test equipment and personnel
Q10B
appropriately are located throughout the
producer's operations.
Effective controls in place to provide accurate
Q11B
part
number
identification
throughout
processing, storage, packaging and shipping.
Q13A
Q14A
Sound storage function with good materials
location system exists
The handling, storage and packaging are
adequate to preserve product quality
A documented and implemented process
exists to ensure that returnable container
Q17C
inventory and their availability in quantity and
quality is
adequate
to cover
customer
requirements.
Q19A
Customer satisfaction levels are measured and
monitored.
Information on quality, customer, operational
Q19C
and financial performances are collected and
analyzed.
Level of significance = 0.05
78
5.6
Examination of Data
Examining the data is important to gain several critical insights into the
characteristics of the data. The ability to interpret the results of the multivariate
techniques may be eased by an understanding of the attributes of the data. In
addition,
hidden problems within the data may have significant implications on analyses of the
applied multivariate techniques that are conducted at a later stage in the analysis. It is
better to be aware of the data characteristics at the onset of the analysis rather than be
blind-sided by the implications, or lack of implications, of the analytical results.
5.6.1 Validation of Data Entry
The completed surveys data were entered into Microsoft Excel for ease in
handling both numeric and character data within the same column field. The survey
responses were entered into the software exactly as the key informant had provided
them. At no time was the supplier's company name associated with the key
informant's responses. Confidentiality was maintained. The keyed data were later
transferred into SPSS software, where each data field was tested for frequency,
mean, and standard deviation to weed out any obvious data entry errors. Table 5.8
contains the means, µ and standard deviations, σ for the items (elements). Found that
means from ever elements are quite high (more than 3.60) out of 5.0 Likert-scale.
This shows that most of the automotive supplier’s perceptions are high on the quality
implementation in their respected companies. This is obviously due to most of them
are already competent because of high certification achievement (i.e ISO 9001 and
TS1649) with more than 5 years as a supplier to automotive manufacturer. Small
value of standard deviation (less than 1.0) shows that there’s no data that fall far
outside the range of standard deviation and considered statistically significant. Result
shows that there’s no errors were found. This indicates that data entered were highly
reliable and ready for further analysis.
79
Table 5.8 Means and Standard Deviation of responses
ELEMENTS
µ
σ
SECTION 1 : QUALITY SYSTEM
Q1A
The responsibility for quality planning on new products are clearly defined
4.243
0.735
Q1B
Definition of responsibility clear for the process involved
4.157
0.740
Q1C
The responsible department for quality planning is defined
4.213
0.728
Q1D
The reporting relationships between departments are defined.
3.885
0.794
Q1E
The key contact personnel/department for quality planning and quality concern resolution
4.021
0.873
for new(and specifically-identified existing) products.
4.305
0.792
Q2B
The company perform feasibility analysis on potential new products
3.959
0.853
Q2C
The adequacy of the company's quality planning effort is assessed.
3.893
0.694
Q3C
FMEAs and Control Plans reviewed and updates as part of the procedure
3.973
0.783
Q3C
Customer approval obtained prior to implementing change.
4.003
1.026
Q3C
A procedure for updating operator instructions and visual aids for process and
3.929
0.719
product changes implemented
3.701
1.052
Q4A
Suppliers encouraged to use Statistical Process Control (SPC).
3.796
0.776
Q4B
Suppliers are selected on the basis of quality aspects.
3.639
0.974
Q4C
Evidence of statistical control and capability required from producers?
3.941
0.772
Q4D
The adequacy of the incoming material quality system is assessed
4.186
1.024
Q5A
Procedures defining the significant quality-related functions available (i.e; a quality
3.834
0.976
an adequate procedure for reacting to Engineering Sample (ES) test failures?)
4.127
0.904
Q5C
The procedures are written.
3.864
1.059
Q5D
There is a formal review system to verify implementation
4.243
0.735
are clearly defined.
Q2A
All quality methods are documented and used as a basis for establishing quality programs
manual) are written
Q5B
The procedures are appropriate to and adequate for the producer's operations. (e.g. is there
SECTION 2: IN PROCESS QUALITY
Q6A
SPC utilized for significant and critical product characteristics and process parameters.
3.964
0.818
Q6B
Control charts are being used effectively to monitor the processes.
3.820
0.815
Q6C
Production's reaction to out-of-control conditions is as specified in the Control Plan.
3.911
0.792
Q6D
The chart indicated that statistical control has been achieved and that process capability has
3.716
0.686
planned and being carried out.
4.098
0.693
Q7B
The statistical methods and other tool used to promote continual improvement indicated.
3.817
0.802
Q7C
The improvements priorities identified and projects teams established
3.846
0.751
Q7D
There is a quality improvement coordinating body (e.g quality steering committee)
3.612
0.969
been demonstrated
Q7A
A definite program to bring about continual improvement in quality and productivity is
80
Table 5.8 : continued
SECTION 3: LOGISTICS AND MANAGEMENT
Q8A
Process/product auditing functions and responsibilities are clearly defined
4.071
0.690
Q8B
The plant activities conduct process/product auditing (e.g; quality inspectors, production
4.021
0.673
4.186
0.721
4.210
0.747
special symbol (such as  )
4.204
0.806
Q9D
Sample sizes and frequencies are adequate
4.038
0.767
Q10A
Appropriate gages, measuring facilities, laboratory equipment and test equipment are
4.189
0.727
4.092
0.743
before being used
4.379
0.648
Q10D
Records indicated that gages and test equipment are periodically inspected and calibrated
4.482
0.622
Q10E
The producer used statistical methods to determine stability and capability of gages,
4.172
0.731
4.044
0.672
4.092
0.663
4.195
0.708
4.133
0.828
quality improvement
4.047
0.784
The working conditions are such that, it could not be detrimental to quality improvement.
4.000
0.735
operators, laboratory technicians) are clearly defined.
Q9A
Written process control instructions available
Q9B
All Critical and Significant characteristics are included, especially those affecting function,
durability and appearance
Q9C
Control Items, especially Critical Characteristics and related operations identified with the
available to facilitate process control?
Q10B
Gages and test equipment and personnel appropriately are located throughout the producer's
operations.
Q10C
New gauges/test equipment are inspected to design specifications, calibrated and approved
measuring and test equipment.
Q11A
Controls that the producer use to indicate the processing and inspection status of products
throughout the producer's system is efficient.
Q11B
Effective controls in place to provide accurate part number identification throughout
processing, storage, packaging and shipping.
Q11C
Controls adequate to prevent movement of rejected incoming materials into the production
system
Q11D
There are effective controls to prevent their movement of non conforming products into
production.
Q12A
Q12B
Plant cleanliness, housekeeping, environmental and working conditions are conducive to
SECTION 4: SHIPPING AND DELIVERY
Q13A
Sound storage function with good materials location system exists
3.994
0.898
Q13B
Inventory records are updated periodically
4.210
0.710
Q13C
A good information system exist
3.988
0.676
Q13D
Engineering problems are considered in packaging (cushioning for fragile products, shelf
3.683
1.157
life for sensitive products, packaging operations configurations)
81
Table 5.8 : continued
Q14A
The handling, storage and packaging are adequate to preserve product quality
4.080
0.772
Q14B
The producer meet applicable packaging specifications for production and service parts.
4.062
0.591
Q14C
Effective controls are in place to assure correct service part identification
3.935
0.641
Q16A
The materials are consolidated and containerized appropriately.
4.328
0.646
Q16B
A standardized procedure exist to ensure what is being shipped is what customer ordered
4.195
0.683
Q17A
Good transportation procedures are exists and a documented control system exists for the
3.976
0.833
3.917
0.885
3.858
1.041
procurement, allocation and monitoring of all packing material
Q17B
Documented procedures for the follow-up of transportation issues relating to quality
(damages), cost (normal, premium freight and detention/demurrage costs) and delivery
(ordering and on-time performance) exists.
Q17C
A documented and implemented process exists to ensure that returnable container inventory
and their availability in quantity and quality is adequate to cover customer requirements.
SECTION 5: AFTER SALES SERVICE
Q18A
Nonconforming parts returned by customer are analyzed.
4.189
0.581
Q18B
The root cause of failure determined, verified and corrective action taken.
4.231
0.672
Q19A
Customer satisfaction levels are measured and monitored.
4.175
0.732
Q19B
A feedback system for customer is provided.
4.018
0.815
Q19C
Information on quality, customer, operational and financial performances are collected and
3.970
0.750
analysed.
82
5.7
Data Processing With Multivariate Analysis
The multivariate analysis tool used in this research was factor analysis.
Factor analysis is a statistical technique used to identify a relatively small number of
factors that can be used to represent the relationships among sets of many interrelated
variables. A huge number of variables can be used to describe a community.
However, descriptions of what is meant by the term community might be greatly
simplified if it was possible to identify underlying dimensions, or factors, of
communities.
Before examining the mechanics of a factor analysis solution, it is important
to consider the characteristics of a successful factor analysis. One goal is to represent
relationships among sets of variables parsimoniously. That is, we would like to
explain the observed correlation using as few factors as possible. If many factors are
needed, little simplification or summarization occurs. One would also like the factors
to be meaningful. A good factor solution is both simple and interpretable. When
factors can be interpreted, new insights are possible (Norusis, 1994).
Using SPSS v.16 Software, factor analysis was performed to construct the
factors that can be used for automotive manufacturers as their tool for supplier
evaluation and performance evaluation. During running the analysis, they are a few
criterias applied to determine the final result of the model.
a) For determining the number of factors to be use in the model, only factors
that account for variances greater than 1 (Eigen value greater than 1) should
be included. Those factors Eigen value lower than 1.0, were eliminated. This
is because of if the eigenvalue is less than 1.0, then the variable is explaining
less variance than a single item. This described best by Dillon ang Goldstein
(1984)
“The rationale for only retaining factors with eigenvalues larger than one
holds for a higher order factor analysis just as for a lower order one. The
(normalized) eigenvectors of a matrix give the direction cosines determining
the rotation, while the eigenvalues give the variance associated with each
83
new axis. When the matrix is a correlation matrix, the variance of each
variable is already standardized to 1, so things are particularly simple. An
eigenvalue less than one represents a shrinking of an axis's importance in the
new universe.”
b) Principal component analysis is used to obtain estimates of the common
factors extracted. The communalities are estimated from the factor loadings,
and factors are again extracted with the new communality estimated replacing
the old. This continues until negligible change occurs in the communality
estimates.
c) Algorithms are used for orthogonal rotation to a simple structure. The most
commonly used method is the varimax method, which attempts to minimize
the number of variables that have high loadings on a factor. This should
enhance the interpretability of the factors.
There are total of 62 elements in the research. The output from the factor analysis
method is obtained by running it with the entire above criterion. Table 5.9 shows the
entire element and their nomenclatures.
84
Table 5.9 Total numbers of element and their Nomenclatures
Q1A
The responsibility for quality planning on new products are clearly defined
Q1B
Definition of responsibility clear for the process involved
Q1C
The responsible department for quality planning is defined
Q1D
The reporting relationships between departments are defined.
Q1E
The key contact personnel/department for quality planning and quality concern resolution are
clearly defined.
Q2A
All quality methods are documented and used as a basis for establishing quality programs for
new(and specifically-identified existing) products.
Q2B
The company perform feasibility analysis on potential new products
Q2C
The adequacy of the company's quality planning effort is assessed.
Q3A
FMEAs and Control Plans reviewed and updates as part of the procedure
Q3B
Customer approval obtained prior to implementing change.
Q3C
A procedure for updating operator instructions and visual aids for process and product
changes implemented
Q4A
Suppliers encouraged to use Statistical Process Control (SPC).
Q4B
Suppliers are selected on the basis of quality aspects.
Q4C
Evidence of statistical control and capability required from producers?
Q4D
The adequacy of the incoming material quality system is assessed
Q5A
Procedures defining the significant quality-related functions available (i.e; a quality manual)
are written
Q5B
The procedures are appropriate to and adequate for the producer's operations. (e.g. is there an
adequate procedure for reacting to Engineering Sample (ES) test failures?)
Q5C
The procedures are written.
Q5D
There is a formal review system to verify implementation
Q6A
SPC utilized for significant and critical product characteristics and process parameters.
Q6B
Control charts are being used effectively to monitor the processes.
Q6C
Production’s reaction to out-of-control conditions is as specified in the control plan.
Q6D
The chart indicated that statistical control has been achieved and that process capability has
been demonstrated
Q7A
A definite program to bring about continual improvement in quality and productivity is
planned and being carried out.
Q7B
The statistical methods and other tool used to promote continual improvement indicated.
85
Table 5.9 Continued
Q7C
The improvements priorities identified and projects teams established
Q7D
There is a quality improvement coordinating body (e.g quality steering committee)
Q8A
Process/product auditing functions and responsibilities are clearly defined
Q8B
The plant activities conduct process/product auditing (e.g; quality inspectors, production
operators, laboratory technicians) are clearly defined.
Q9A
Written process control instructions available
Q9B
All Critical and Significant characteristics are included, especially those affecting
function, durability and appearance
Q9C
Control Items, especially Critical Characteristics and related operations identified with
the special symbol (such as )
Q9D
Sample sizes and frequencies are adequate
Q10A
Appropriate gages, measuring facilities, laboratory equipment and test equipment are
available to facilitate process control?
Q10B
Gages and test equipment and personnel appropriately are located throughout the
producer's operations.
Q10C
New gauges/test equipment are inspected to design specifications, calibrated and
approved before being used
Q10C
Records indicated that gages and test equipment are periodically inspected and
calibrated
Q10D
The producer used statistical methods to determine stability and capability of gages,
measuring and test equipment.
Q10E
Controls that the producer use to indicate the processing and inspection status of
products throughout the producer's system is efficient.
Q11A
Effective controls in place to provide accurate part number identification throughout
processing, storage, packaging and shipping.
Q11B
Controls adequate to prevent movement of rejected incoming materials into the
production system
Q11C
There are effective controls to prevent their movement of non conforming products into
production.
Q11D
Plant cleanliness, housekeeping, environmental and working conditions are conducive to
quality improvement
Q12A
The working conditions are such that, it could not be detrimental to quality
improvement.
Q12B
Sound storage function with good materials location system exists
Q12C
Inventory records are updated periodically
Q13A
A good information system exist
86
Table 5.9 Continued
Q13B
Engineering problems are considered in packaging (cushioning for fragile products, shelf
life for sensitive products, packaging operations configurations)
Q13C
The handling, storage and packaging are adequate to preserve product quality
Q13D
The producer meet applicable packaging specifications for production and service parts.
Q14A
Effective controls are in place to assure correct service part identification
Q14B
The materials are consolidated and containerized appropriately.
Q14C
A standardized procedure exist to ensure what is being shipped is what customer ordered
Q16A
Good transportation procedures are exists and a documented control system exists for the
procurement, allocation and monitoring of all packing material
Q16B
Documented procedures for the follow-up of transportation issues relating to quality
(damages), cost (normal, premium freight and detention/demurrage costs) and delivery
(ordering and on-time performance) exists.
Q17A
A documented and implemented process exists to ensure that returnable container inventory
and their availability in quantity and quality is adequate to cover customer requirements.
Q17B
Nonconforming parts returned by customer are analyzed.
Q18B
The root cause of failure determined, verified and corrective action taken.
Q19A
Customer satisfaction levels are measured and monitored.
Q19B
A feedback system for customer is provided.
Q19C
Information on quality, customer, operational and financial performances are collected and
analysed.
87
5.7.1
Initial Output layout
5.7.1.1
KMO and Bartlett’s test
The Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy is an index
for comparing the magnitudes of the observed correlation coefficients to the
magnitudes of the partial correlation coefficients (Norusis, 1994) ranges between 0
and 1. A value close to zero indicates the diffusion in the pattern of correlations, and
therefore that factor analysis may not be appropriate. Values less than 0.5 are poor,
0.7-0.8 is good, and anything over 0.9 is excellent.
Another indicator of the strength of the relationship among variables is
Bartlett's test of sphericity. Bartlett's test of sphericity is used to test the null
hypothesis that the variables in the population correlation matrix are uncorrelated
(Norusis, 1994). The observed significance level is .0000. It is small enough to reject
the hypothesis. It is concluded that the strength of the relationship among variables is
strong.
From Table 5.10, it shows that result for KMO test is just good (0.631). It is
not good enough for Factor Analysis to be carried out. But the large value for
Bartlett’s Test of Sphericity shows that it is a good idea to proceed with a factor
analysis for the data.
Table 5.10: KMO and Bartlett’s Test (a)
Kaiser-Meyer-Olkin Measure of Sampling Adequacy.
Approx. Chi-Square
Bartlett's Test of Sphericity
.631
34958.54
Df
1891
Sig.
.000
88
5.7.1.2 Communalities
Communalities are measures on the amount of variance in each variable that
the factors explain. We are only interested in communalities after extraction in
principal Component Analysis. A low values (<0.1) indicates that the variable should
be omitted from the analysis by Hishamuddin Md. Som (2005). From the Table 5.11,
found that few elements (Q3A, Q8B, Q18B and Q19C) have low extraction
compared to other elements. Even though element Q18B had value of 0.267, it
smaller than other data that mostly more than 0.70 values. Thus, we should reject the
elements Q3A, Q8B, Q18B and Q19C because of low communalities compared to
other data. Next, the data were reanalyze by deleting or dropping these low
communalities variables.
89
Table 5.11 Communalities (a)
Element
Extraction
Element
Extraction
1
Q1A
0.769
32
Q9C
0.719
2
Q1B
0.792
33
Q9D
0.797
3
Q1C
0.810
34
Q10A
0.770
4
Q1D
0.764
35
Q10B
0.868
5
Q1E
0.738
36
Q10C
0.774
6
Q2A
0.745
37
Q10D
0.687
7
Q2B
0.808
38
Q10E
0.686
8
Q2C
0.828
39
Q11A
0.762
9
Q3A
0.102
40
Q11B
0.830
10
Q3B
0.822
41
Q11C
0.829
11
Q3C
0.756
42
Q11D
0.729
12
Q4A
0.781
43
Q12A
0.838
13
Q4B
0.781
44
Q12B
0.818
14
Q4C
0.807
45
Q12C
0.882
15
Q4D
0.812
46
Q13A
0.733
16
Q5A
0.827
47
Q13B
0.850
17
Q5B
0.854
48
Q13C
0.750
18
Q5C
0.822
49
Q13D
0.755
19
Q5D
0.835
50
Q14A
0.859
20
Q6A
0.873
51
Q14B
0.815
21
Q6B
0.824
52
Q14C
0.834
22
Q6C
0.793
53
Q16A
0.852
23
Q6D
0.907
54
Q16B
0.879
24
Q7A
0.773
55
Q17A
0.870
25
Q7B
0.839
56
Q17B
0.858
26
Q7C
0.641
57
Q17C
0.882
27
Q7D
0.572
58
Q18A
0.793
28
Q8A
0.819
59
Q18B
0.267
29
Q8B
0.134
60
Q19A
0.848
30
Q9A
0.873
61
Q19B
0.890
31
Q9B
0.856
62
Q19C
0.077
Extraction Method: Principal Component Analysis.
90
5.7.2
Output for rerun
5.7.2.1
KMO and Bartlett’s Test with communalities
After taken elements Q3A, Q8B, Q18B and Q19C out, the data must be
reanalyze again for KMO test, Bartlett’s test and communalities. This is due to
amount of correlation values between elements have changed without those rejected
variables. After running the factor analysis by excluding the rejected variables, it was
found that Kaiser-Meyer-Olkin (KMO) measure of Sampling Adequacy has
increased from 0.631 to 0.828. There’s also no sign of low communalities among all
the elements. This shows that factor analysis is much more appropriate now (see
Table 5.12 for KMO and Bartlett’s test and Table 5.13 for communalities (b)).
Table 5.12
KMO and Bartlett's Test (b)
Kaiser-Meyer-Olkin Measure of Sampling Adequacy.
Bartlett's Test of Sphericity
.828
Approx. Chi-Square
30118.89
Df
1653
Sig.
.000
91
Table 5.13 Communalities (b)
Element
Extraction
Element
Extraction
1
Q1A
0.749
30
Q9C
0.711
2
Q1B
0.807
31
Q9D
0.752
3
Q1C
0.777
32
Q10A
0.707
4
Q1D
0.763
33
Q10B
0.870
5
Q1E
0.730
34
Q10C
0.782
6
Q2A
0.734
35
Q10D
0.718
7
Q2B
0.781
36
Q10E
0.770
8
Q2C
0.828
37
Q11A
0.747
9
Q3B
0.817
38
Q11B
0.806
10
Q3C
0.781
39
Q11C
0.827
11
Q4A
0.749
40
Q11D
0.708
12
Q4B
0.787
41
Q12A
0.842
13
Q4C
0.802
42
Q12B
0.828
14
Q4D
0.800
43
Q12C
0.889
15
Q5A
0.846
44
Q13A
0.765
16
Q5B
0.879
45
Q13B
0.774
17
Q5C
0.823
46
Q13C
0.700
18
Q5D
0.785
47
Q13D
0.755
19
Q6A
0.875
48
Q14A
0.837
20
Q6B
0.818
49
Q14B
0.793
21
Q6C
0.796
50
Q14C
0.800
22
Q6D
0.911
51
Q16A
0.825
23
Q7A
0.781
52
Q16B
0.886
24
Q7B
0.823
53
Q17A
0.871
25
Q7C
0.681
54
Q17B
0.848
26
Q7D
0.707
55
Q17C
0.871
27
Q8A
0.731
56
Q18A
0.733
28
Q9A
0.812
57
Q19A
0.742
29
Q9B
0.836
58
Q19B
0.817
92
5.7.3 Total Variance Explained
Principal Component Analysis extracts as many factors as there are variables.
The relative importance of these variables declines (as measured by the % of
variance explained and their eigenvalues). Variance explained is the variance of the
model's predictions with the total variance (of the data). Variance Explained is part
of the variance of any residual that can be attributed to a specific condition (cause)
(Norusis, 1994). The higher the explained variance relative to the total variance, the
stronger the statistical measure used.
Table 5.14 show the first 13 component of the total variance explained. The
eigenvalues associated with each factor represent the variance explained by that
particular linear component and SPSS also displays the eigenvalue in terms of the
percentage of variance explained (so, factor 1 explains 42.26% of total variance). It
should be clear that the first few factors explain relatively large amount of variance
(especially factor 1) whereas subsequent factors shows only small amount of
variance. SPSS then extracts all factors with an Eigenvalue greater than 1.0, which
leaves with 11 factors. The Eigenvalues associated with these factors are again
displayed (and the percentage of variance explained) in the columns labeled
Extraction sums of Squared loadings. The values in this part of the table are the same
as the values before extraction, except that the values for the discarded factors are
ignored because they are not significant. As in Table 5.14, after the 11th factor, the
data for 12 onwards are blank. This shows that only 11 factors can be extracted from
the analysis.
93
Table 5.14 Total variance explained
Extraction Sums of Squared
Rotation Sums of Squared
Loadings
Loadings
Initial Eigenvalues
% of
Total Variance
Cumulative
% of
Cumulative
% of
Cumulative
%
Total
Variance
%
Total
Variance
%
1 24.511 42.260
42.260
24.511
42.260
42.260
8.498
14.652
14.652
2
5.053
8.712
50.972
5.053
8.712
50.972
5.411
9.330
23.982
3
2.823
4.867
55.839
2.823
4.867
55.839
5.362
9.244
33.227
4
2.609
4.498
60.337
2.609
4.498
60.337
4.506
7.769
40.995
5
2.291
3.950
64.287
2.291
3.950
64.287
4.414
7.610
48.605
6
2.122
3.659
67.946
2.122
3.659
67.946
3.847
6.633
55.238
7
1.721
2.967
70.913
1.721
2.967
70.913
3.781
6.519
61.757
8
1.514
2.610
73.522
1.514
2.610
73.522
3.151
5.432
67.189
9
1.252
2.159
75.682
1.252
2.159
75.682
3.030
5.224
72.413
10 1.057
1.823
77.505
1.057
1.823
77.505
2.709
4.670
77.083
11 1.034
1.783
79.287
1.034
1.783
79.287
1.278
2.204
79.287
12
.959
1.654
80.941
13
.937
1.615
82.556
5.7.4 Scree Plot and Component plot in rotated space
Eigenvalues are not the only method that can used to judge how many factors
should be extracted; it also be done using the scree plot. The graph is useful for
determining how many factors to retain. The point of interest is where the curve
starts to flatten. It can be seen that the curve begins to flatten between factors 5 and
11. (Refer Figure 5.1). So, level of confident that 11 factors solution is appropriate.
94
Based on the component plot in rotated space, the pattern of the grouping
variables into their loading factors can be observed (Figure 5.2). However, this
output comes with five-dimension and it’s hard to visualize.
Flatten Curve
Figure 5.1 Scree Plot
Figure 5.2 Component plot in rotated space
95
5.7.5
Rotated Component Matrix
The rotation phase of factor analysis attempt to transform the initial matrix
into one that is easier to interpret. The purpose of rotation is to achieve a simple
structure. This helps on interpreting the factors. This permits the factors to be
differentiated from each other.
Based on Rotated Component Matrix (Table 5.15), each element can be
identified in their group factors. They are 11 factors altogether. But, from the result
of rotated component matrix, found out that Factor ‘G’, Factor ‘H’ and Factor ‘K’
have only one or two element. Basically, these 3 factors can be eliminated as it didn’t
have significant value to the research. The elements that should be eliminated are
elements Q7A, Q3C, Q11A and Q7D. So, finally there are 8 factors left. Factor ‘A’
have 10 elements, Factor ‘B’ have 4 elements, Factor ‘C’ have 6 elements, Factor
‘D’ have 5 elements, Factor ‘E’ have 4 elements. Factor ‘F’ have 3 elements, Factor
‘I’ have 3 elements and Factor ‘J’ have 3 elements. In the next section, all these
factors will be given interpretation based on their elements.
96
Table 5.15
Rotated Component Matrix
Rotated Component Matrix
Component (Factor)
Element
A
B
Q14A
0.735
0.185
0.221
0.164
0.169
Q13C
0.718
0.009
0.072
0.165
Q13B
0.716
0.206
0.241
Q13A
0.711
0.288
Q14C
0.709
Q12B
C
D
E
F
G
H
0.234
-0.027
0.145
0.095
0.163
0.217
0.117
0.228
0.189
0.090
0.110
0.112
-0.136
0.169
-0.012
0.128
0.231
-0.035
0.084
0.230
0.035
0.227
0.154
-0.004
0.008
-0.035
0.073
0.138
0.012
0.275
0.177
0.011
0.104
0.129
0.315
0.056
0.358
-0.041
0.075
0.032
0.674
0.028
0.361
0.101
0.192
0.001
0.208
-0.108
0.215
0.304
0.038
Q12A
0.667
0.016
0.313
0.214
0.281
-0.057
0.241
-0.052
0.252
0.204
0.076
Q14B
0.667
0.135
-0.036
0.142
0.155
0.231
0.262
0.383
-0.041
0.110
0.042
Q12C
0.635
-0.223
0.119
0.185
0.232
0.140
0.159
0.240
0.445
0.176
-0.031
Q19B
0.634
0.157
0.039
0.322
0.135
0.349
0.126
0.256
0.096
0.181
-0.151
Q6D
0.043
0.861
0.076
-0.014
0.149
0.153
0.041
0.309
-0.023
-0.120
0.070
Q6A
0.082
0.809
0.211
-0.004
0.252
0.125
0.204
-0.032
0.104
0.181
-0.057
Q6B
0.146
0.800
0.143
0.206
0.058
0.110
0.148
0.043
0.194
0.071
0.110
Q6C
0.206
0.745
0.219
0.074
0.151
0.153
0.268
-0.009
0.097
0.083
0.100
Q9B
0.196
0.387
0.648
0.205
0.262
0.211
0.154
0.159
-0.068
0.107
-0.090
Q10B
0.315
-0.064
0.642
0.136
-0.022
0.134
0.249
0.370
0.207
0.207
0.182
Q1B
0.157
0.309
0.634
0.091
0.097
0.415
-0.041
0.014
0.290
0.093
-0.029
Q9C
0.139
0.340
0.626
0.087
0.179
0.047
0.306
0.049
-0.179
0.119
-0.002
Q9D
0.035
0.196
0.616
0.158
0.252
0.261
0.105
0.244
0.219
0.219
0.103
Q1A
0.275
0.321
0.604
0.177
0.038
0.390
0.104
-0.042
0.053
0.027
0.070
Q16B
0.063
-0.066
0.135
0.872
0.093
0.084
0.171
0.114
-0.184
0.082
-0.018
Q16A
0.246
0.094
0.261
0.793
0.044
0.106
0.149
-0.075
0.053
0.043
0.115
Q17B
0.327
0.116
0.100
0.761
0.138
0.024
-0.088
0.206
0.241
0.101
-0.028
Q17A
0.369
0.132
0.088
0.715
0.338
0.149
-0.130
0.089
0.185
0.005
0.050
Q17C
0.469
0.050
0.076
0.625
0.126
0.184
-0.109
0.231
0.347
0.102
-0.087
Q5B
0.132
0.192
0.274
0.092
0.842
0.117
0.031
0.077
0.094
-0.026
-0.053
Q5D
0.178
0.045
-0.082
0.185
0.792
0.121
0.219
0.052
-0.029
0.089
0.092
Q5A
0.107
0.222
0.302
0.075
0.772
0.026
0.137
0.054
0.243
-0.018
0.109
Q5C
0.107
0.366
0.172
0.151
0.646
0.037
0.145
0.231
0.350
0.102
-0.022
I
J
K
97
Table 5.15 Continued
Q2B
0.138
0.184
0.089
0.116
0.067
0.782
-0.034
0.000
-0.036
-0.122
0.271
Q2C
0.218
0.269
0.268
0.045
0.160
0.731
0.153
0.072
0.194
0.066
-0.056
Q1E
0.319
-0.020
0.275
0.122
0.038
0.660
0.268
0.112
0.125
-0.019
-0.026
Q7A
0.132
0.320
0.302
0.043
0.165
0.002
0.692
0.067
0.133
0.075
0.183
Q3C
0.144
0.129
0.129
0.114
0.366
0.277
0.605
0.020
0.279
0.086
-0.225
Q11A
0.343
0.098
0.246
0.080
0.041
0.048
-0.035
0.720
-0.021
0.133
0.102
Q3B
0.172
0.048
0.169
0.200
0.282
0.130
0.299
0.058
0.715
0.120
0.028
Q4A
0.240
0.338
0.130
0.028
0.235
0.157
0.159
0.025
0.654
0.081
0.140
Q4C
0.260
0.253
0.036
-0.080
0.057
0.105
0.473
-0.125
0.604
0.208
-0.016
Q10E
0.294
0.170
0.017
0.043
0.060
0.019
-0.026
0.071
0.100
0.795
0.031
Q10D
0.348
-0.105
0.277
0.167
0.052
-0.057
0.103
0.204
0.135
0.636
0.012
Q10C
0.452
-0.034
0.215
0.148
0.096
-0.098
0.088
0.256
0.098
0.634
0.072
Q7D
0.202
0.220
0.025
0.018
0.273
0.159
0.243
0.141
0.112
0.122
0.641
5.8
Interpreting Result
The final step in factor analysis is an attempt to give interpretations to the
rotated factors. From the output revealed from the Rotated Component Matrix,
interpretation can be done. Interpretation of the rotated factors required skills to
identify the common themes among elements in the grouping factors.
The questions that load highly on Factor ‘A’ seem to all related on workplace
and materials handling. There for Factor ‘A is labeled as Workplace Environment
and Materials Handling. The questions that load highly on Factor ‘B’ all seem
related, which is labeled as Statistical Process Control (SPC).
The six questions that load highly on factor ‘C’ have the same theme on
Supplier’s production operation system. This factor can be labeled as Logistics and
Management.
Factor ‘D’ is related on procedures and shipping documentation
regarding finished products. It can be labeled as Shipping and Delivery.
98
The common theme on factor ‘E’ is In-house Quality Related matters. It
deals with quality procedures especially on quality implementation and
documentation.
The 3 questions that load highly in Factor ‘F’ are based on
responsibility on new products and quality planning. It might be labeled as Quality
Planning. Incoming and outgoing quality control can be concluded to Factor I as it is
related for incoming parts and outgoing products quality matters.
Final grouping of factor ‘J’ is Gauge and test equipments.
It is about
procedures and handling gauges and test equipments. All the factors and their
elements have been summarized in Table 5.16. This new form of Questionnaire
(constructed model of the instrument tool) is the final result and can be used for
supplier performance assessment and evaluation.
99
Table 5.16 Factors and their elements
Factor A
Workplace Environment and materials handling
Q14A
The handling, storage and packaging are adequate to preserve product quality
Q13C
A good information system exist
Q13B
Inventory records are updated periodically
Q13A
Sound storage function with good materials location system exists
Q14C
Effective controls are in place to assure correct service part identification
Q12B
The working conditions are such that, it could not be detrimental to quality improvement.
Plant cleanliness, housekeeping, environmental and working conditions are conducive to quality
Q12A
improvement
Q14B
The producer meet applicable packaging specifications for production and service parts.
Q12C
Actions have been taken to mitigate factors that can depreciate work environment and culture.
Q19B
A feedback system for customer is provided.
Factor B
Statistical Process Control (SPC)
The chart indicated that statistical control has been achieved and that process capability has been
Q6D
demonstrated
Q6A
SPC utilized for significant and critical product characteristics and process parameters.
Q6B
Control charts are being used effectively to monitor the processes.
Q6C
Production's reaction to out-of-control conditions is as specified in the Control Plan..
Factor C
Logistics and management
All Critical and Significant characteristics are included, especially those affecting function,
Q9B
durability and appearance
Gages and test equipment and personnel appropriately are located throughout the producer's
Q10B
operations.
Q1B
Definition of responsibility clear for the process involved
Control Items, especially Critical Characteristics and related operations identified with the special
)
Q9C
symbol (such as
Q9D
Sample sizes and frequencies are adequate
Q1A
The responsibility for quality planning on new products are clearly defined
100
Factor D
Shipping and delivery
Q16B
A standardized procedure exist to ensure what is being shipped is what customer ordered
Q16A
The materials are consolidated and containerized appropriately
Documented procedures for the follow-up of transportation issues relating to quality (damages),
cost (normal, premium freight and detention/demurrage costs) and delivery (ordering and on-time
Q17B
performance) exists.
Good transportation procedures are exists and a documented control system exists for the
Q17A
procurement, allocation and monitoring of all packing material
A documented and implemented process exists to ensure that returnable container inventory and
Q17C
Factor E
their availability in quantity and quality is adequate to cover customer requirements.
In-House Quality-related matters
The procedures are appropriate to and adequate for the producer's operations. (e.g. is there an
Q5B
adequate procedure for reacting to Engineering Sample (ES) test failures?)
Q5D
There is a formal review system to verify implementation
Procedures defining the significant quality-related functions available (i.e; a quality manual) are
Q5A
written
Q5C
The procedures are written.
Factor F
Quality planning
Q2B
The company perform feasibility analysis on potential new products
Q2C
The adequacy of the company's quality planning effort is assessed.
The key contact personnel/department for quality planning and quality concern resolution are
Q1E
Factor I
clearly defined.
Incoming/Outgoing Quality Control
Q3B
Customer approval obtained prior to implementing change.
Q4A
Suppliers encouraged to use Statistical Process Control (SPC).
Q4C
Evidence of statistical control and capability required from producers?
Factor J
Gages and test equipments
The producer used statistical methods to determine stability and capability of gages, measuring
Q10E
and test equipment
Q10D
Records indicated that gages and test equipment are periodically inspected and calibrated
New gauges/test equipment are inspected to design specifications, calibrated and approved before
Q10C
being used
101
5.9
Constructed Supplier Performance Assessment Tool for Automotive
Industry
Based on result of interpreting the final outcome, Supplier Performance
Assessment tool for Automotive Industry had been constructed. It have 8 Factors
with the total number of elements are 38. The final 8 factors are interpreted as
Workplace Environment and materials handling, Statistical Process Control,
Logistics and management, Shipping and delivery, In-House Quality matters, Quality
Planning, Incoming/Outgoing Quality Control and Gauges and Test equipments (see
Table 5.17r the constructed instrument).
Table 5.17 : Constructed Assessment tool for Automotive Industry
Workplace Environment and materials handling
1
The handling, storage and packaging are adequate to preserve product quality
NA
1 2 3 4 5
2
A good information system exist
NA
1 2 3 4 5
3
Inventory records are updated periodically
NA
1 2 3 4 5
4
Sound storage function with good materials location system exists
NA
1 2 3 4 5
5
Effective controls are in place to assure correct service part identification
NA
1 2 3 4 5
6
The working conditions are such that, it could not be detrimental to quality
NA
1 2 3 4 5
NA
1 2 3 4 5
NA
1 2 3 4 5
NA
1 2 3 4 5
NA
1 2 3 4 5
NA
1 2 3 4 5
NA
1 2 3 4 5
improvement.
7
Plant cleanliness, housekeeping, environmental and working conditions are
conducive to quality improvement
8
The producer meet applicable packaging specifications for production and service
parts.
9
Actions have been taken to mitigate factors that can depreciate work environment
and culture.
10
A feedback system for customer is provided.
Statistical Process Control (SPC)
11
The chart indicated that statistical control has been achieved and that process
capability has been demonstrated
12
SPC utilized for significant and critical product characteristics and process
parameters.
13
Control charts are being used effectively to monitor the processes.
NA
1 2 3 4 5
14
Production's reaction to out-of-control conditions is as in the Control Plan.
NA
1 2 3 4 5
102
Logistics and management
15
All Critical and Significant characteristics are included, especially those affecting
NA
1 2 3 4 5
NA
1 2 3 4 5
function, durability and appearance
16
Gages and test equipment and personnel appropriately are located throughout the
producer's operations.
17
Definition of responsibility clear for the process involved
NA
1 2 3 4 5
18
Control Items, especially Critical Characteristics and related operations identified
NA
1 2 3 4 5
with the special symbol (such as
)
19
Sample sizes and frequencies are adequate
NA
1 2 3 4 5
20
The responsibility for quality planning on new products are clearly defined
NA
1 2 3 4 5
Shipping and delivery
21
A standardized procedure exist to ensure what is being shipped is what ordered
NA
1 2 3 4 5
22
The materials are consolidated and containerized appropriately
NA
1 2 3 4 5
23
Documented procedures for the follow-up of transportation issues relating to
NA
1 2 3 4 5
NA
1 2 3 4 5
NA
1 2 3 4 5
NA
1 2 3 4 5
quality (damages), cost (normal, premium freight and detention/demurrage costs)
and delivery (ordering and on-time performance) exists.
24
Good transportation procedures are exists and a documented control system exists
for the procurement, allocation and monitoring of all packing material
25
A documented and implemented process exists to ensure that returnable container
inventory and their availability in quantity and quality is adequate to cover
customer requirements.
In-House Quality-related matters
26
The procedures are appropriate to and adequate for the producer's operations. (e.g.
is there an adequate procedure for reacting to Engineering Sample (ES) test
failures?)
27
There is a formal review system to verify implementation
NA
1 2 3 4 5
28
Procedures defining the significant quality-related functions available (i.e; a
NA
1 2 3 4 5
NA
1 2 3 4 5
quality manual) are written
29
The procedures are written.
Quality planning
30
The company perform feasibility analysis on potential new products
NA
1 2 3 4 5
31
The adequacy of the company's quality planning effort is assessed.
NA
1 2 3 4 5
32
The key contact personnel/department for quality planning and quality concern
NA
1 2 3 4 5
resolution are clearly defined.
103
Incoming/Outgoing Quality Control
33
Customer approval obtained prior to implementing change.
NA
1 2 3 4 5
34
Suppliers encouraged to use Statistical Process Control (SPC).
NA
1 2 3 4 5
35
Evidence of statistical control and capability required from producers?
NA
1 2 3 4 5
NA
1 2 3 4 5
NA
1 2 3 4 5
NA
1 2 3 4 5
Gages and test equipments
36
The producer used statistical methods to determine stability and capability of
gages, measuring and test equipment
37
Records indicated that gages and test equipment are periodically inspected and
calibrated
38
New gauges/test equipment are inspected to design specifications, calibrated and
approved before being used
This constructed assessment tool now can be used for automotive
manufacturer to assess their suppliers. It can be combined or can assimilate with their
current practices for assessing their suppliers. The Likert-scale (1-5) as shown as
Table 5.17 is just for example purpose only. Automotive manufacturer can use any of
their current means for keeping the score of their assessment based what suits best to
them. This proposed assessment tool hopefully can be such guidance for automotive
manufacturer to improve their assessment capability and strengthen its competitive
advantages of to survive in the market.
104
5.10
Discussion and Summary
This study was conducted in order to answer the problem statement of this
research. As mentioned before, the study objective is to develop instrument tool for
supplier performance evaluation and assessment in automotive industry.
This chapter has presented the results for the survey. The data were
examined, tested for validation and processed with multivariate analysis using SPSS
software and the developed instrument tool was supported. The sample includes 338
respondents from suppliers of automotive industry, representing 23.4 percent
response rate. The businesses are predominantly manufacturers of components,
including electrical, interior, chassis and exterior.
The survey result shows that most of the companies have complied with more
than one certification. The survey result shows that automotive suppliers have high
certification achievement ratio especially in MS ISO 9001 and TS 16949 to
‘maintain’ their business relationship with the automotive manufacturers. These
certifications are likely part of the requirements by most of the automotive suppliers
nowadays.
The data collection was run in two phases. In the first phase, the
questionnaires were send to all the suppliers. The second phase was the
questionnaires were re-sent after few weeks on no reply on the first attempt.
Responses were split into two groups based on if they were returned before or after
the mailing of the second survey package. The analysis found that there are no
significant differences in between early and late responses.
The Cronbach’s alpha coefficient used to test the reliability of the
questionnairel. It indicates that all the scales are acceptable because all the
Cronbach’s alpha values are above 0.7. This shows that the questionnaire is reliable.
Factor analysis has been used to process the data in this chapter. There were
total of 62 elements in the research. The output from the factor analysis method by
105
running it with the criterion such as factors for eigen value must lower that 1.0 and
algorithms used is varimax method, which attempts to minimize the number of
variables that have high loadings on a factor to enhance the interpretability of the
factors.
From the initial output, found that the Kaiser-Meyer-Olkin (KMO) test was
quite low which is 0.631. The communalities after extraction in Principal Component
Analysis shows for all the elements with low values. These variables were omitted in
the analysis (Elements Q3A, Q8B, Q18B and Q19C). After reanalyzing the data
without these variables, it was found that the KMO results increased to 0.828. It
shows that factor analysis is now more suitable compared with before.
Based on the results, all the elements can be identified in 11 factors. But,
from the analysis of the rotated component matrix, three factors can be eliminated as
they have only one or two element in each factor. The final 8 factors are interpreted
as Workplace Environment and materials handling, Statistical Process Control,
Logistics and management, Shipping and delivery, In-House Quality matters, Quality
Planning, Incoming/Outgoing Quality Control and Gauges and Test equipments (see
Table 5.17 for the constructed instrument). It is hoped that instrument can assist
automotive instrument to improve their suppliers and to achieve the vision and
mission meeting the customer’s expectation. It is hoped that the constructed
instrument will benefit automotive manufacturer and strengthen its competitive
advantages of to survive in the market.
CHAPTER 6
CONCLUSION AND RECOMENDATION
6.1
Introduction
This chapter presents the conclusion of this study. The first section provides
the need of an instrument for supplier performance assessment and conclusion of
research findings in an attempt to answer the research objectives. The next section
addresses the limitation of the study and the last section presents the potential areas
for future research.
6.2
Summary and Conclusions of the Research
The supplier evaluation process is complicated because a variety of criteria
must be simultaneously considered. In some approaches to supplier evaluation, only
quantitative factors are allowed in the model, or qualitative factors can be used in the
model but the data are replaced by the assigned numbers. However, the assigned
numbers may not directly reflect the impreciseness of the performance data. In order
to obtain an effective evaluation, the impreciseness of data should be accurately
reflected.
107
More often than not, a supplier assessment is based on the lowest bid, and in
some cases on unsystematic and incomprehensive subjective evaluation and
interviews. Therefore, it becomes too late to proactively avoid supplier issues or
divest production flow of their symptoms. If causes of the suppliers’ issues (i.e
quality, delivery, etc) are accounted for early in the supplier assessment process, the
associated risk could be minimized.
The main objective of this study was to develop an assessment tool that can
be used as generic approach to measure supplier performance in automotive industry,
which can assist automotive manufacturer to improve their productivity and increase
customer satisfaction level. This study also intended to prove that multivariate
analysis can be used for development of supplier performance assessment tool. In
addition, this research means to give a benchmark from which to measure
improvement. By this research, all three objectives has been fulfilled. At the end of
the research, constructed assessment tool that have been developed fulfilled all the 3
objectives. The assessment tool can be used as generic approach to measure supplier
performance in automotive industry and was developed by multivariate analysis.
While doing data collection, 3rd objective been fulfilled when the data shows
condition on practices on quality implementation on their respected company. This
can give true benchmark to each supplier to measure their improvement to the next
level.
A survey through mail questionnaire was the approach adopted in this study
in order to find out the perception and application of TQM and quality practices
implementation status. A questionnaire was prepared and sent to automotive
suppliers across Malaysia, giving a 24.3 per cent rate. The respondents represented a
wide variety of business that supports the automotive industry. The predominant
main business of the respondents was to supply electrical components, giving 25.6
percent rate of total respondents.
The length of time the supplier has conducting business with the automotive
manufacturer indicates the understanding of the respondent with the requirements
demand by the automotive company. It was found that 87% of the respondent
108
companies are conducting business more than 5 years with the automotive
manufacturer. It means that the data have a high degree of confidence because the
more length of time in conducting business together, the more suppliers can fulfill
requirements demands by the automotive manufacturer. In terms of quality
certification, almost 91.9% of the respondents have MS ISO 9001:2000 certification.
Most of the companies have complied with more than one certification. The survey
result shows that automotive suppliers have high certification achievement ratio
especially in MS ISO 9001 and TS 16949 to ‘maintain’ their business relationship
with the automotive manufacturers.
The sample was split into two groups based on if they were returned before or
after the mailing of the second survey package. T-test for equality of mean has been
conducted to test whether there are significant responses between early and late
responses. The t-tests yield that there are no significant differences between early and
late responses. Therefore, the data analysis proceeded with scale purification and
testing of the measurement model.
There are total of 62 elements in the research. The output from the factor
analysis method by running it with the criterion of eigen value greater than 1, using
principal component analysis and varimax method for the algorithms. Initial output
shows that Kaiser-Meyer-Olkin is 0.631 (barely enough for factor analysis). From
the initial output, elements Q3A, Q8B, Q18B and Q19C were rejected because of
low communalities compared to other data.
After rerun factor analysis by excluding the rejected variables, found that
Kaiser-Meyer-Olkin measure of Sampling Adequacy has increased from 0.631 to
0.828. It shows that factor analysis is much more appropriate now. Based on Rotated
Component Matrix , each element can be identified in their group factors which
results 11 factors altogether. But, from the result of rotated component matrix, found
out that Factor 7, Factor 8 and Factor 11 have only one or two element. These 3
factors can be eliminated as it didn’t have significant value to the research. These
resulting 8 factors left. The 8 factors has been interpreted as Workplace Environment
and Materials Handling, Statistical Process Control (SPC), Logistics and
Management, Shipping and Delivery, In-house Quality Related matters, Quality
109
Planning. Incoming and outgoing quality control and finally Gauge and test
equipments construct the assessment tool for automotive industry.
Hopefully, the proposed tool can assist automotive manufacturer’s
improvement effort to achieve the organization vision and mission focused on
meeting customer satisfaction. It is believed that the proposed tool is applicable in
other industry. The other benefit of the tool is to strengthen the competitive
advantage of an organization to survive in the market.
6.3
Limitation of the study
As with any other research or study there will always be limitation and
weakness. There were some limitations to this study. One limitation was the scope of
the study. The study focused mainly on the automotive industry. However, it is
believed that proposed model can be applied in other industry such as electronics
industry or heavy manufacturers industry with some modifications. Other limitation
is the suppliers that were chosen are only Malaysian suppliers, because of resource
limitation, both time and money. If oversea suppliers were included in this research,
the result may be different as they may have different in values and perception as
compared in Malaysian automotive suppliers.
110
6.4
Future Research Recommendations
Problems faced by local suppliers preventing them to produce quality
products should be eliminated at its root. To be competent with foreign car suppliers,
all parties involved in determining the status of the local automotive suppliers should
put their heads and hands together and improve their quality practice. Future research
on this area is necessary to reveal the situation in the local automotive industry in
terms of local components, degree of quality system implementation, the link
between quality of product and certification and capabilities of vendors in producing
quality products.
In this study, the research focus was on the supplier’s perspective. The
buyer’s perspective has been incorporated into many studies and its relevance to this
study was nonessential. Future studies can compare both perspectives to assess the
level of similarity and potential bias.
Case study can be done to few automotive manufacturers. The results could
be categorized by the type of quality system they are accredited to or their
performance (Good, Average, Not good) according to automotive manufacturers.
Analysis could be conducted on the opinion of several car manufacturers on one
vendor. Car manufacturers might have different views on a particular vendor. Study
could be conducted on the difference of opinions, the reasons opinions differ whether
it is due to car manufacturers' requirements, vendor attitude or some quality problems
previously encountered by car manufacturers with the vendor. The data could be
linked to the quality practice in the vendor's organization. It is also hoped that in
future, the proposed tool be expandable to other sectors.
111
REFERENCES
Aamer, A. M. (2005). Supplier Evaluation and Selection: A Comprehensive model to minimize
the risk associated with quality and delivery. Industrial Engineering, The University of
Tennessee. Ph. D Dissertation
Ahmed, S. and M. Hassan (2003). "Survey and case investigation on application of quality
management tools and technique in SMI." International Journal of Quality and Reliability
Management 20(7): 795-826.
Amid, A., S. H. Ghodsypour and C. O’Brien (2006). "Fuzzy multiobjective linear model for
supplier selection in a supply chain." Int. J. Production Economics (104): 394-407.
Araz, C. and Ozkarahan, I. (2006). “Supplier evaluation and management system for strategic
sourcing based on a new multicriteria sorting procedure”. International Journal of
Production Economics. 106(2):585-606
Barua, A., S. Ravindran and A. Whinston (1997). "Efficient Selection of Supplier over the
Internet." Journal of Management Information Systems 13(4): 117-137.
Carr, Amelia S. and John N. Pearson (1999), “Strategically Managed Buyer-Supplier
Relationships and Performances Outcomes, “Journal of Operations Managements. 17:
497-519
Carr, Amelia, S. Kaynak and Hale (2007). “Communication methods, information sharing,
supplier development and performance: An empirical study of their relationships”.
International Journal of Operations and Product Management. 27(4): 346-370
Cebi, F. and D. Bayraktar (2003a). "A new measure for supplier performance evaluation." IEE
Transactions 29: 753-758.
112
Cebi, F. and D. Bayraktar (2003b). "An Integrated Approach For Supplier Selection " Logistics
Information Management 12(6): 395-400.
Chen, C. C., T. M. Yeh, et al. (2004). "Customer-focused rating System of supplier quality
performance." Journal of Manufacturing Technology Management 15(7): 599–606.
Chen, C. T., L. C. Torng, et al. (2005). "A fuzzy approach for supplier evaluation and selection
in supply chain management." Int. J. Production Economics 102(289–301).
Chen, T. W., J. Y. Lin, et al. (2003). "Selecting a supplier by fuzzy evaluation of capability
indices, Cpm " International Journal of Advance Manufacturing Technology 22: 534–540.
Chicago Press.
Czaja, Ronald and Johny Blair (1996), Designing Surveys: A Guide to decisions and Procedures,
Thousand Oaks, California: Pine Forge Press, Sage Publications.
Cooper, Donald R. and Schindler P.S (2001). Business Research Method. New York. McGrawHill International Edition
Dahel and Nasr-Eddine (2003). “Vendor selection and order quantity allocation in volume
discount environments”. Supply Chain Management: An International Journal
8(4):335-342
Demirtas E.A and Ustun O. (2008). “An integrated multiobjective decision making process for
supplier selection and order allocation” Omega 36(1): 76-90
Dickson, Gary W. (1966), “An analysis of vendor Selection Systems and Decisions,” Journal of
Purchasing: 5-17
Dillman , D. A. (2000). Mail and Internet Surveys: The Tailored Design Method. New York:
John Wiley & Sons, Inc.
113
Dillon, W. R., & Goldstein, M. (1984). Multivariate Analysis: Methods and Applications. New
York. John Wiley & Sons, Inc:
Dobler, D. W. and D. N. Burt (1990). Purchasing and Materials Management: Text and cases.
New York, McGraw-Hill.
Evans, J. R. and W. M. Lindsay (2002). The Management and Control of Quality. Cincinnati,
Ohio, Thomas Learning.
Enriquez J., Chuda and Jenny M.Y Leung (2005). “Inventory lot-sizing with supplier selection”.
Computers and Operations Research 32: 1-14
Forker, Laura B, William A. Ruch and James Hershauer, (1999) “Examining Suppier
improvement Efforts from Both Sides,” The journal of Supply Chain Management
35(3):40-50
Ganesalingam, S. and Ganesh, S.. (2002) Detection of financial distress via neural networks and
traditional multivariate statistical methods. The 6th Meeting of Saudi Association for
Mathematical Sciences, 9-10 April 2002.
Giunipero, L. C., et al. (1998). "Performance Based Evaluation Systems Under Total Quality
Management." International Journal of Purchasing and Materials Management 29(1): 3541.
Giunipero, Larry C and Daniel J. Brewer, (1998) “ Performance Based Evaluation Systems
Under Total Quality Management,” International Journal of Purchasing and Materials
Management 29(1):35-41
Hahn, C. K., K. H. Kim and J. S. Kim. (1996). "Costs of Competition: Implications for
Purchasing Strategy." Journal of Purchasing and Material Management 26(2): 35-41.
114
Handfield, R. b. (2003). "A Resource Dependence Perspective of Just-in-Time Purchasing."
Journal of Operations Management 11(3).
Harman, Harry H. (1976). Modern Factor Analysis. The Third Edition Revised. London.
Chapman and Hall/CRC.
Hartley, J. and T. Y. Choi (1996). "Supplier Development: Customer as a Catalyst of Process
change." Business Horizons 39(4): 37-44.
Hashim, Mohd Khairudin and Wafa, Syed Azizi (2002). Small and Medium Sized Enterprises in
Malaysia. Malaysia. Prentice Hall
Hishamuddin Md. Som (2005). Panduan Mudah Analisis Data Menggunakan SPSS Windows.
Hongwei, D., L. Benyouce and Xiaolan Xie. (2003). A simulation-optimization approach using
genetic search for supplier selection. Winter Simulation Conference, France.
Huck, Schuyler W. and Cormier, William H. (1995). Reading Statistic and Research. Second
Edition. New York. Harer Collins Publisher Inc.
Johnson R.A and Wichern D.W (2001). Applied Multivariate Statistical Analysis. 5th Edition.
New Jersey: Prentice Hall
Kahraman, C., U. Cebeci and Z. Ulukan (2003). "Multi-criteria supplier selection using fuzzy
AHP. Logistics Information Management." Logistics Information Management 16(2):
382-394.
Krause, D. R., T. Scannel and Calantone, R. (2000). "A Structural Analysis of the Effectiveness
of Buying Firm's Strategies to Improve Supplier Performance." Decision Sciences 31(1):
33-55.
115
Lacey G. and Elliot S. (2000). “Supplier qualification in the telecommunications industry”
Journal of Production Innovation 17(2):110-127
Larson, P. D. and J. Kulchitsky, (1999). "Single Sourcing and Supplier Certification." Industrial
Marketing Management 27(1): 73-81.
Lasch, R. and C. G.Janker (2005). "Supplier selection and controlling using multivariate
analysis." International Journal of Physical Distribution & Logistics Management 35(6):
409-425.
Lessler, J. and W. D. Kalsbeek (1992). Nonsampling Error in Surveys. New York: Wiley.
Li, C. C., Fun Y. P., Hung J.S (1997). " A new measure for supplier performance evaluation."
IIE Transactions 29: 753-758.
Liao, Z. and Rittscher J. (2006). "A multi-objective supplier selection model under stochastic
demand conditions." International Journal of Production Economics 105(1): 150-159.
Liu J. and Wu C. (2005). An Integrated Method for Supplier Selection in SCM. ICSSSM '05,
Services Systems and Services Management.
Millington, A., M. Eberhardt and Barry Wilkinson (2005). "Supplier performance and selection
in China." International Journal of Operations & Production Management 26(2): 185201.
Monczka, R.M and S.J Trecha, 1988. “Cost-based supplier performance evaluation”. Journal of
Purchasing and Material Management 24(2):2-7
Montgomery, D. C. (2001). Introduction to Statistical Process Control. New York, John Wiley
and Sons, Inc.
116
Moutaz, K. and M. Booth (1995). "A Robust Multivariate Statistical Procedure for Evaluation
and Selection of Industrial Robots." International Journal of Operations & Production
Management 12(2): 10-15
Muralidharan, C. and N. Anantharaman (2001). "Vendor rating in purchasing scenario : a
confidence interval approach." International Journal of Operations & Production
Management 21(10): 1305-1325.
Najmi, M., J. Rigas and Ip-Shing Fan (2005). "A framework to review performance
measurement systems." Business Process Management Journal 11(2): 109-122.
Norušis M.J (1994). SPSS Professional Statistics 11.0. London:Prentice Hall
Nunnally, J. C. (1978). Psychometric Theory. New York, McGraw-Hill.
Nydick, R. L. and R. P. Hill (1992). "Using the Analytic Hierarchy Process to Structure the
Supplier Selection Procedure." Journal of Purchasing and Material Management: 31-36.
Penerbit; Universiti Teknologi Malaysia Skudai Johor Darul Ta’zim
Peng K.C and. Chen B.D (2005) “Web based cost estimation and supplier selection based on
parametric cost estimation”. Journal of Purchasing and Materials Management 19(2):110127
Pett M.A, Lacky N. R and Sullivan J.J (2003). Making Sense of Factor Analysis: The Use of
Factor Analysis for Instrument Development in Health Care Research. 1st Edition. Ohio,
Sage Publications, Inc
Porter, A. M. (1997). "Supplier Evaluation Revisited." Purchasing 111(6): 58-68.
Prahinsky, C. (2001). Communication strategies and supplier performance evaluations in an
industrial supply chain, The Ohio State University. Ph.D.Dissertation
117
Ridge, Andreas M. (2003). Validity and reliability testin case study research: a literature review
with “hands-on” applications for each research phase. Qualitative Market Research: An
International Journal. 6(2): 75-86
Saaty, T. L. (1980). The Analytic Hierarchy Process. 2nd Edition. New York, McGraw Hill.
Sanders, Donald H. (1995) Statistics: A First Course. Fifth Edition NEW York, McGraw Hill
Saunders, A. (1994). "Supplier Audits as Part of a Supplier Partnership." The TQM Magazine
6(2): 41-42.
Schimtz, J. and K. W. Platts (2003). "Roles of supplier performance measurement: indication
from a study in the automotive industry." Management Decision 41(8): 711-721.
Stern, L. W. and A. I. El-Ansary (1982). Marketing Channels. Englewood Cliffs, New Jersey,
Princess Hall.
Sucky, E. (2001). A dynamic model for strategic supplier selection Operations Research
Proceedings.
Sucky, E. (2001). A dynamic model for strategic supplier selection Operations Research
Proceedings. September 15-18. Dearbon, Michigan: IEEE. 360-365
The University of Alabama. Ph.D Dissertation
Sung H. H and Krishnan R. (2007)“A hybrid approach to supplier selection for the maintenance
of a competitive supply chain”. Expert Systems with Applications. 34(2): 1303-1311
Tsai, W. C. (1999). A procedure for the supplier evaluation problem under risk using fuzzy sets,
The University of Alabama. Ph.Dissertation
118
Vaniman, J. (1998). "Supplier Performance Ratings for Enhanced Effectiveness." Logistics
Information Management 16: 22-27.
Wan L.N (2008)“An efficient and simple model for multiple criteria supplier selection problem”.
European Journal of Operational Research. 186(3): 1059-1067
Williams, A.J and B.A Oumlil (1987), “Supplier Selection and Development,” Journal of
Purchasing and Material Management, Vol. 23 (3):24-28
Yin, Robert K. (1997). The Abridged Version of Case Study Research: Design and Method.
England. Woodhead Publishing Ltd.
Youssef, M. A., M. Zairi, B. Mohanty.(1998). "Supplier selection in an advanced manufacturing
technology environment: an optimization model." Benchmarking for Quality
Management & Technology 3(4): 60-72.
Zhang, Q. and Z. Wang (2005). Fuzzy Integration Method of Synthetic Evaluation for Supplier.
ICSSSM'05. Services Systems and Services Management
119
APPENDIX A
QUESTIONNAIRE
120
Dept. of Mfg and Industrial Engineering
Faculty of Mechanical Engineering
81310 UTM Skudai, Johor.
Record No:
SUPPLIER PERFORMANCE ASSESSMENT QUESTIONNAIRE
Introduction:
This is a questionnaire to investigate on Supplier Performance Assessment of automotive vendors
and suppliers. Our main objective is to develop a tool that can simplify the process of supplier
assessment by automotive manufacturers.
All responses will be treated with the utmost confidence and the result obtained will be used for
research purposes only and no attempt will be made to identify any individual or organizations in
any of our publications.
Instruction:
The questionnaire comprises of two parts (A and B). Please answer both parts. Please write on the
space allocated, tick ( ) in the appropriate boxes or circle the responses. In some questions, you
may tick in more than one box. If the options you wish to tick on are not stated, please write them in
the “others” column.
The questionnaire will take no longer than 20 minutes to complete.
Thank you again. Your kindness is much appreciated.
PART A: GENERAL INFORMATION
In this section, we would like to know about your organisation in general.
Company Name:
Address:
Web Site:
1. My company is a
__________
for automotive manufacturer.
Distributor
Service Provider
Software Developer
Manufacturer
Designer
Others (Please specify)
2. What is the main product or services does your company provide to the automotive
manufacturer?
121
3. How many years have your company serve automotive manufacturer’s need?
Less than 1 year
1-2 years
3-5 ears
5-10 years
More than 10 years
4. Which characteristics does car manufacturer formally evaluate regarding your
company’s performance? Check all that are applicable.
Cost/Price
Product/ service quality
Conformance quality
High Quality design
3rd party quality awards
Delivery speed
Delivery reliability (on-time)
Development speed
Process capability
Volume flexibility
Product line breadth
Managerial capability
Financial health
Others:
5. Which of the following is your company certified to?
MS ISO 9001
TS16949
OHSAS 18001
MS ISO 14000
Others: _________________
6. Which of the following awards have your company won?
Prime Minister Quality Award
Quality Management Excellence Award
Industry Excellence Award
State Quality Award
Others: ______________
None
PART B: ASSESSMENT QUESTIONS
This part attempts to find out your perception of the factors that are practiced for quality
implementation in your company. It consists of 5 parts of questions. Please circle your perception
of the practice for each statement listed below in your organization. Please use the following
scales.
Practice – The extent or degree of practice in your organization.
NA= Not Available 1= very low, 2= low, 3= moderate, 4= High, 5= very high
SECTION 1 : QUALITY SYSTEM
Low
High
Degree of Practices
Q1
1
Responsibility for quality planning
The responsibility for quality planning on new products are clearly defined
NA
1 2 3 4 5
2
3
Definition of responsibility clear for the process involved
The responsible department for quality planning is defined
NA
NA
1 2 3 4 5
1 2 3 4 5
4
5
The reporting relationships between departments are defined.
The key contact personnel/department for quality planning and quality
concern resolution are clearly defined.
Quality Documentation (Control Plans, Process Failure Mode and Effect
Analyses, etc.)
All quality methods are documented and used as a basis for establishing
quality programs for new(and specifically-identified existing) products.
The company perform feasibility analysis on potential new products
The adequacy of the company's quality planning effort is assessed.
NA
NA
1 2 3 4 5
1 2 3 4 5
NA
1 2 3 4 5
NA
NA
1 2 3 4 5
1 2 3 4 5
Q2
1
2
3
Low
Q3
High122
1
Procedure for reviewing design and process changes prior to
implementation
FMEAs and Control Plans reviewed and updates as part of the procedure
NA
1 2 3 4 5
2
Customer approval obtained prior to implementing change.
NA
1 2 3 4 5
3
A procedure for updating operator instructions and visual aids for process and
NA
1 2 3 4 5
1
Effective system for assuring the quality of incoming products and
services
(e.g. ; plating, heat treating)
Suppliers encouraged to use Statistical Process Control (SPC).
NA
1 2 3 4 5
2
3
Suppliers are selected on the basis of quality aspects.
Evidence of statistical control and capability required from producers?
NA
NA
1 2 3 4 5
1 2 3 4 5
4
The adequacy of the incoming material quality system is assessed
NA
1 2 3 4 5
Q5
Quality-related functions procedures
1
NA
1 2 3 4 5
NA
1 2 3 4 5
3
Procedures defining the significant quality-related functions available (i.e; a
quality manual) are written
The procedures are appropriate to and adequate for the producer's operations.
(e.g. is there an adequate procedure for reacting to Engineering Sample (ES)
test failures?)
The procedures are written.
NA
1 2 3 4 5
4
There is a formal review system to verify implementation
NA
1 2 3 4 5
SECTION 2: IN PROCESS QUALITY
Degree of Practices
product changes implemented
Q4
2
Q6
Statistical Process Control (SPC)
1
SPC utilized for significant and critical product characteristics and process
parameters.
Control charts are being used effectively to monitor the processes.
Production's reaction to out-of-control conditions is as specified in the Control
Plan..
The chart indicated that statistical control has been achieved and that process
capability has been demonstrated
There’s a plan to improve the process in all cases where process capability has
not yet been demonstrated.
2
3
4
5
NA
1 2 3 4 5
NA
NA
1 2 3 4 5
1 2 3 4 5
NA
1 2 3 4 5
NA
1 2 3 4 5
Q7
Continuous improvement
1
A definite program to bring about continual improvement in quality and
productivity is planned and being carried out.
The statistical methods and other tool used to promote continual improvement
indicated.
The improvements priorities identified and projects teams established
NA
1 2 3 4 5
NA
1 2 3 4 5
NA
1 2 3 4 5
There is a quality improvement coordinating body (e.g quality steering
committee)
NA
1 2 3 4 5
SECTION 3: LOGISTICS AND MANAGEMENT
Degree of Practices
2
3
4
Q8
Process/product auditing functions
1
Process/product auditing functions and responsibilities are clearly defined
NA
1 2 3 4 5
2
The plant activities conduct process/product auditing (e.g; quality inspectors,
production operators, laboratory technicians) are clearly defined.
NA
1 2 3 4 5
Low
Q9
1
2
Written process for incoming, in-process, laboratory, layout inspection
and outgoing auditing
Written process control instructions available
4
All Critical and Significant characteristics are included, especially those
affecting function, durability and appearance
Control Items, especially Critical Characteristics and related operations
identified with the special symbol (such as  )
Sample sizes and frequencies are adequate
Q10
Gauges, measuring facilities, laboratory equipment and test equipment
1
Appropriate gages, measuring facilities, laboratory equipment and test
equipment are available to facilitate process control?
Gages and test equipment and personnel appropriately are located throughout
the producer's operations.
New gauges/test equipment are inspected to design specifications, calibrated
and approved before being used
Records indicated that gages and test equipment are periodically inspected and
calibrated
The producer used statistical methods to determine stability and capability of
gages, measuring and test equipment.
3
2
3
4
5
Q11
1
2
3
4
Q12
1
2
3
Processing and inspection indication
Controls that the producer use to indicate the processing and inspection status
of products throughout the producer's system is efficient.
Effective controls in place to provide accurate part number identification
throughout processing, storage, packaging and shipping.
Controls adequate to prevent movement of rejected incoming materials into
the production system
There are effective controls to prevent their movement of non conforming
products into production.
Work environment and culture
Plant cleanliness, housekeeping, environmental and working conditions are
conducive to quality improvement
The working conditions are such that, it could not be detrimental to quality
improvement.
Actions have been taken to mitigate factors that can depreciate work
environment and culture.
High 123
NA
1 2 3 4 5
NA
1 2 3 4 5
NA
1 2 3 4 5
NA
1 2 3 4 5
NA
1 2 3 4 5
NA
1 2 3 4 5
NA
1 2 3 4 5
NA
1 2 3 4 5
NA
1 2 3 4 5
NA
1 2 3 4 5
NA
1 2 3 4 5
NA
1 2 3 4 5
NA
1 2 3 4 5
NA
1 2 3 4 5
NA
1 2 3 4 5
NA
1 2 3 4 5
SECTION 4: SHIPPING AND DELIVERY
Degree of Practices
Q13
1
Storage and inventory
Sound storage function with good materials location system exists
NA
1 2 3 4 5
2
Inventory records are updated periodically
NA
1 2 3 4 5
3
A good information system exist
NA
1 2 3 4 5
4
Engineering problems are considered in packaging (cushioning for fragile
products, shelf life for sensitive products, packaging operations
configurations)
Handling and packing
The handling, storage and packaging are adequate to preserve product quality
NA
1 2 3 4 5
NA
1 2 3 4 5
The producer meet applicable packaging specifications for production and
service parts.
Effective controls are in place to assure correct service part identification
NA
1 2 3 4 5
NA
1 2 3 4 5
Q14
1
2
3
124
Low
High
Q15
Processing and inspection status of products
1
There are controls to indicate the processing and inspection status of products
throughout the company's system.
Effective controls are in place to provide accurate part number identification
throughout processing, storage, packaging and shipping.
Controls are adequate to prevent movement of rejected incoming materials
into the production system.
Nonconforming products are separated from the stream of production. There
are effective controls to prevent their movement.
Completed products
The materials are consolidated and containerized appropriately.
NA
1 2 3 4 5
NA
1 2 3 4 5
NA
1 2 3 4 5
NA
1 2 3 4 5
NA
1 2 3 4 5
A standardized procedure exist to ensure what is being shipped is what
customer ordered
Transportation Documentation
Good transportation procedures are exists and a documented control system
exists for the procurement, allocation and monitoring of all packing material
Documented procedures for the follow-up of transportation issues relating to
quality (damages), cost (normal, premium freight and detention/demurrage
costs) and delivery (ordering and on-time performance) exists.
A documented and implemented process exists to ensure that returnable
container inventory and their availability in quantity and quality is adequate to
cover customer requirements.
NA
1 2 3 4 5
NA
1 2 3 4 5
NA
1 2 3 4 5
NA
1 2 3 4 5
SECTION 5: AFTER SALES SERVICE
Degree of Practices
Q18
1
Reaction to customer concern
Nonconforming parts returned by customer are analyzed.
NA
1 2 3 4 5
2
The root cause of failure determined, verified and corrective action taken.
NA
1 2 3 4 5
Q19
1
2
3
Customer feedback
Customer satisfaction levels are measured and monitored.
A feedback system for customer is provided.
Information on quality, customer, operational and financial performances are
collected and analysed.
NA
NA
NA
1 2 3 4 5
1 2 3 4 5
1 2 3 4 5
2
3
4
Q16
1
2
Q17
1
2
3
Is your company interested to participate in the next phase of the project for a detailed case study?
Yes
No
Will consider
If yes, please attach your business card so that we can contact you.
Thank you for participating in this study. We assure you that all responses will be treated with the
utmost confidence and no single set of response will be identifiable.
Any queries please direct to:
Prof. Dr. Sha’ri M. Yusof,
Project Supervisor,
Dept. of Mfg and Industrial Engineering,
81310 UTM Skudai.
07-5534694, e-mail: shari@fkm.utm.my
Or
Mohd Azril bin Amil
email: azril.amil@gmail.com
Telephone: 012-3305260
125
APPENDIX B
LETTERS
126
APPENDIX B-1
SAMPLE LETTER FOR PRE-PILOT STUDY
(QUALITY EXPERTISE)
<DATE>
Dean <NAME>
I am currently research into Supplier Performance Assessment of automotive
vendors and suppliers. The main objective of my study is to develop a tool that can
simplify the process of supplier assessment by automotive manufacturers. Towards this
end, I have developed a questionnaire for this purpose.
Prior to conducting the full survey, I believe that it is important and necessary
that designed questionnaire meets its purpose and easily understood. As an expert,, I
would greatly appreciate your completing and commenting upon it.
I would like be grateful for your response before <date> to enable me to continue
with the next stage of the survey.
Thank you in anticipating.
Your sincerely,
Mohd Azril bin Amil
email: azril.amil@gmail.com
H/P: 012-3305260
CC: Prof. Dr.Shari Mohd. Yusof
Project Supervisor,
Faculty of Mechanical,
University Teknologi Malaysia.
Tel- 07-5534680 Fax:07-5566159 / Email: shari@fkm.utm.my
127
APPENDIX B1a
LIST OF QUALITY EXPERTS
1. En. Jafri Mohamed Rohani
Lecturer,
Faculty of Mechanical Engineering
UTM Skudai
Johor Bahru, Malaysia
2. Assoc. Prof. Dr. Adnan bin Hassan
Lecturer,
Faculty of Mechanical Engineering
UTM Skudai
Johor Bahru, Malaysia
3. Assoc. Prof. Noordin bin Mohd Yusof
Lecturer,
Faculty of Mechanical Engineering
UTM Skudai
Johor Bahru, Malaysia
4. Dr. Muhamad Zameri bin Mat Saman
Lecturer,
Faculty of Mechanical Engineering
UTM Skudai
Johor Bahru, Malaysia
5. En. Nasir bin Kadikon
PQA Assistant Manager,
J.K. Wire Harness Sdn. Bhd.
No.7,9 & 11, Jalan Firma 2/2,
Kawasan Perindustrian Tebrau 1,
81100 Johor Bahru,
128
6. En. Edly Ferdin Ramly,
Principal Consultant,
EFR Management Consultant
Jalan Merbau 3,
Bandar Putra,
81000 Kulai, Johor
7. En. Ramli Md Salleh
PQA Senior Engineer
GE-SHEN CORPORATION BERHAD
NO.11, Jalan Riang 23, Taman Gembira,
81200, Johor Bahru, West Malaysia.
8. En. Khairul Hanafiah Mahadi
QA Engineer,
Wonderful Wire & Cable Berhad,
No. 5 Jalan Firma 2/3
Kawasan Perindustrian Tebrau I
Johor Bahru, Johor.
9. Ms. Nimalee Sri Ellaiah
QA Engineer
Lot 153, PTD 64046, Jalan Angkasa Mas 6,
Kawasan Perindustrian Tebrau II,
Mukim Tebrau, 81100 Johor Bahru, Johor, Malaysia.
129
APPENDIX B-2
SAMPLE LETTER FOR FULL SURVEY
To: Quality Assurance Manager,
<COMPANY NAME>
<COMPANY ADDRESS>
<DATE>
Record No:
Dear Sir,
Survey on Supplier Performance Assessment in Automotive Industry.
I am currently research into Supplier Performance Assessment of automotive vendors and
suppliers. The main objective of my study is to develop a tool that can simplify the process
of supplier assessment by automotive manufacturers.
2.
I have developed questionnaire for that purpose. I would appreciate if you could
spare some time, which I know is very limited to respond to my questionnaire. The
questionnaire has been designed in such a way that the questions are fairly easy to answer.
(I.e. requiring a tick or a circle).
3
Attached are 5 copies of the questionnaires. I hope you can distribute it among your
colleagues and respond to it. Your responses to the questionnaire are crucial for the success of
my research, which hopefully will be of great assistance to accomplish the objective of the
study.
4.
I assure you that all responses will be kept confidential and only used for research
purpose. If you have further enquiries about the survey or the research in general, please feel
free to contact me or Prof. Dr.Sha'ri Yusof. I appreciate if you could return to my supervisor at
the address below before the <DATE> with attached envelope.
Thank you in advance,
Yours sincerely,
Mohd Azril bin Amil
email: azril.amil@gmail.com
H/P: 012-3305260
CC: Prof. Dr.Shari Mohd. Yusof
Project Supervisor,
Faculty of Mechanical,
University Teknologi Malaysia.
Tel- 07-5534680 Fax:07-5566159 / Email: shari@fkm.utm.my
130
APPENDIX B-3
SAMPLE LETTER FOR FOLLOW-UP SURVEY
To: Quality Assurance Manager,
<COMPANY NAME>
<COMPANY ADDRESS>
<DATE>
Record No:
Dear Sir,
Re: Survey on Supplier Performance Assessment in Automotive Industry.
I am currently research into Supplier Performance Assessment of automotive vendors and
suppliers. The main objective of my study is to develop a tool that can simplify the process
of supplier assessment by automotive manufacturers.
2.
I have developed questionnaire for that purpose. I would appreciate if you could
spare some time, which I know is very limited to respond to my questionnaire. The
questionnaire has been designed in such a way that the questions are fairly easy to answer.
(I.e. requiring a tick or a circle).
3
Attached are 3 copies of the questionnaires. I hope you can distribute it among your
colleagues and respond to it. Your responses to the questionnaire are crucial for the success of
my research, which hopefully will be of great assistance to accomplish the objective of the
study.
4.
I assure you that all responses will be kept confidential and only used for research
purpose. If you have further enquiries about the survey or the research in general, please feel
free to contact me or Prof. Dr.Sha'ri Yusof. I appreciate if you could return to my supervisor at
the address below before the <DATE> with attached envelope.
Thank you in advance,
Yours sincerely,
Mohd Azril bin Amil
email: azril.amil@gmail.com
H/P: 012-3305260
CC: Prof. Dr.Shari Mohd. Yusof
Project Supervisor,
Faculty of Mechanical,
University Teknologi Malaysia.
Tel- 07-5534680 Fax:07-5566159 / Email: shari@fkm.utm.my
131
APPENDIX C
RELIABILITY TEST
132
Reliability Section 1
**********Method 1 (space saver) will be used for this analysis **********
RELIABILITY ANALYSIS – SCALE (ALPHA)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Q1A
Q1B
Q1C
Q1D
Q1E
Q2A
Q2B
Q2C
Q3A
Q3B
Q3C
Q4A
Q4B
Q4C
Q4D
Q5A
Q5B
Q5C
Q5D
Reliability Coefficients
N of Cases = 338.0
Alpha =
.891
N of Items = 19
133
Reliability Section 2
**********Method 1 (space saver) will be used for this analysis **********
RELIABILITY ANALYSIS – SCALE (ALPHA)
1
2
3
4
5
6
7
8
Q6A
Q6B
Q6C
Q6D
Q7A
Q7B
Q7C
Q7D
Reliability Coefficients
N of Cases = 338.0
Alpha =
.873
N of Items = 8
134
Reliability Section 3
**********Method 1 (space saver) will be used for this analysis **********
RELIABILITY ANALYSIS – SCALE (ALPHA)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Q8A
Q8B
Q9A
Q9B
Q9C
Q9D
Q10A
Q10B
Q10C
Q10D
Q10E
Q11A
Q11B
Q11C
Q11D
Q12A
Q12B
Q12C
Reliability Coefficients
N of Cases = 338.0
Alpha =
.792
N of Items = 18
135
Reliability Section 4
**********Method 1 (space saver) will be used for this analysis **********
RELIABILITY ANALYSIS – SCALE (ALPHA)
1
2
3
4
5
6
7
8
9
10
11
12
Q13A
Q13B
Q13C
Q13D
Q14A
Q14B
Q14C
Q16A
Q16B
Q17A
Q17B
Q17C
Reliability Coefficients
N of Cases = 338.0
Alpha =
.926
N of Items = 12
136
Reliability Section 5
**********Method 1 (space saver) will be used for this analysis **********
RELIABILITY ANALYSIS – SCALE (ALPHA)
1
2
3
4
5
Q13A
Q13B
Q13C
Q13D
Q14A
Reliability Coefficients
N of Cases = 338.0
Alpha =
.884
N of Items = 5