ANALYTIC HIERARCHY PROCESS IN ACADEMIC STAFF SELECTION
AT FACULTY OF SCIENCE IN UNIVERSITY TECHNOLOGY MALAYSIA
VOON CHUI KHIM
A dissertation submitted in partial fulfillment of the
requirement for the award of the degree of
Master of Science (Mathematics)
Faculty of Science
Universiti Teknologi Malaysia
NOVEMBER 2009
iv
ACKNOWLEDGEMENT
First and foremost I offer my sincerest gratitude to my supervisor, Dr.
Rohanin Bt Ahmad, who has supported me throughout my thesis with her patience
and knowledge whilst allowing me the room to work in my own way. Above all and
the most needed, she provided me unflinching encouragement and support in various
ways. One simply could not wish for a better or friendlier supervisor. I am totally
grateful for her encouragement, guidance, critics and friendship. This project would
not be completed without her moral support and professional advice from time to
time. I would also like to express my sincere thanks to Mdm Rashidah Ahmad and
Dr. Zaitul Marlizawati Zainuddin for evaluating my work.
Their efforts have
certainly brought me to even greater height.
I convey special acknowledgement to Mr. Harun Bin Mohd, Mr Ramli
Ibrahim, Prof Dr. Mustaffa Shamsuddin, Prof Dr. Norsarahaida S. Amin and Prof Dr
Noriah Bidin, for their indispensable help in the knowledge acquisition process for
the collection of data in this study.
I am also indebted to Kementerian Pelajaran
Malaysia (KPM) for funding my Master study.
It is my sincere hope that the
knowledge and experience gained here can be put to good use.
I am especially grateful to all my family members for their continuous
supports and motivation. Finally, I would also like to extend my sincere appreciation
to all my colleagues, my course mates and everybody who have provided me with
valuable assistance at various occasions to the successful realization of thesis, as well
as expressing my apology that I could not mention personally one by one.
v
ABSTRACT
Academic staff selection is an important process for a university since the
decision affects the quality of education and the success of the university. Selection
committee always faces up to the uncertainty and vagueness in the conventional
decision making process. These subjective perception and the experiences of
decision maker can be effectively represented and reached to a more effective and
unbiased decision by introducing Analytic Hierarchy Process (AHP) to the normally
practiced selection process.
AHP permits pair-wise comparison judgments in
expressing the relative strength of impact for the criteria in the hierarchy which can
be translated to quantitative data to help in improving the traditional method and
simplifies the process of selecting a best academic staff by considering the criteria
that may influence the decision made. This is crucial in order to formalize a decision
process, reduce time commitments, create a process orientation, document the
strategy, and result in better decisions for the selection process. This study embarks
on finding out the potential criteria and developing an AHP aided decision making
model for the academic staff selection process in the Faculty of Science, UTM. The
selection criteria of Knowledge, Working Experience/Interpersonal Skill, General
Traits, Scholar/ Extra Curricular Activities and References used in the model
developed are determined based on the knowledge acquisition from experts’ in the
university and literature reviews done. The ranking of generated candidate profiles
shows that Candidate 2 possesses the highest priority, thus should be selected as the
academic staff, by using the AHP model developed. Microsoft Excel and Expert
Choice 11.5 are used to assist in accomplishing the tedious calculations involved, as
well as providing effective aids for discussions and analyzing of the results. A few
suggestions for future work and research direction in the area of academic selection
process is also being discussed.
vi
ABSTRAK
Pemilihan kakitangan akademik merupakan satu proses yang penting
disebabkan keputusan pemilihan akan mempengaruhi kualiti pendidikan dan
kejayaan suatu universiti.
Dalam proses pemilihan tradisional, jawatankuasa
pemilihan selalunya menghadapi ketakpastian dan kekaburan semasa membuat
keputusan.
Tanggapan subjektif dan pengalaman pembuat keputusan boleh
diwakilkan secara berkesan dan keputusan yang lebih efektif dan saksama dapat
dibuat dengan memperkenalkan Proses Hierarki Analitik (AHP) ke dalam proses
pemilihan tersebut. AHP membenarkan perbandingan penilaian secara berpasangan,
mempamerkan secara jelas kekuatan relatif kritera-kriteria dalam hierarki. Dengan
mempertimbangkan kriteria-kriteria yang dapat diterjemahkan kepada data
kuantitatif dan berkemungkinan mempengaruhi keputusan yang dibuat, keberkesanan
kaedah pemilihan dan permudahan proses pemilihan dapat dibuat. Ini bertujuan
memformalkan proses pemilihan, menjimatkan penggunaan masa, mewujudkan
proses yang berorientasi, mendokumenkan strategi dan menghasilkan keputusan
yang lebih memuaskan dalam proses pemilihan. Kajian bermula dengan penemuan
kriteria-kriteria yang berkaitan dan seterusnya membangunkan satu model pembuat
keputusan dengan bantuan AHP dalam proses pemilihan kakitangan akademik dalam
Fakulti Sains, UTM. Kriteria-kriteria pemilihan adalah Pengetahuan, Pengalaman
Kerja/ Kemahiran Diri, Perwatakan, Kegiatan luar dan Rujukan yang ditentukan
berdasarkan pengetahuan daripada pakar-pakar dalam universiti dan rujukan journal.
Kedudukan calon berdasarkan profil yang dijanakan menunjukkan bahawa Calon 2
memiliki tahap keutamaan tertinggi.
Microsoft Excel dan Expert Choice 11.5
digunakan dalam mempermudahkan pengiraan, perbincangan dan analisis keputusan.
Beberapa cadangan untuk kerja penyelidikan seterusnya berserta hala-tuju
penyelidikan dalam bidang yang sama juga turut dibincangkan.
vii
TABLE OF CONTENTS
CHAPTER
1
2
TITLE
PAGE
DECLARATION
ii
DEDICATION
iii
ACKNOWLEDGEMENTS
iv
ABSTRACT
v
ABSTRAK
vi
TABLE OF CONTENTS
vii
LIST OF TABLES
xi
LIST OF FIGURES
xiv
LIST OF ABBREVIATIONS
xvii
LIST OF SYMBOLS
xix
LIST OF APPENDICES
xx
INTRODUCTION
1
1.1
Introduction
1
1.2
Background of the Problem
3
1.3
Statement of the Problem
5
1.4
Objectives of the Study
6
1.5
Scope of the Study
6
1.6
Significance of the Study
8
1.7
Structure of Dissertation
8
LITERATURE REVIEW
11
2.1
11
Introduction
viii
2.2
Overview of Faculty Academic Staff Selection
12
Problem
2.2.1
3
15
2.3
Multiple Criteria Decision Making (MCDM)
18
2.4
Analytic Hierarchy Process (AHP)
19
2.5
Academic Staff Selection Model
22
2.6
Selection Criteria
22
2.7
Summary
23
RESEARCH METHODOLOGY
25
3.1
Introduction
25
3.2
Methodology
25
3.2.1
Knowledge Acquisition
26
3.2.1.1
27
3.3
Interviews
Analytic Hierarchy Process (AHP)
28
3.3.1 Research Design and Procedure
29
3.3.2
3.3.3
4
Flow Charts
Implementation of AHP – Development of
Model for Faculty Academic Staff Selection
36
Decision Making
37
3.4
Research Framework
38
3.5
Summary
40
KNOWLEDGE
ACQUISITION
FOR
THE
DEVELOPMENT OF MODEL - A CASE STUDY OF
UTM
41
4.1
Introduction
41
4.2
Background
42
4.2.1 Registrar’s Office
42
4.2.2 Faculty of Science
44
Selection Process
46
4.3.1 Registrar’s Office
46
4.3
4.3.2
Selection Process at Faculty Level (Faculty of
Science)
50
ix
4.3.3
4.4
5
Selection Criteria
55
Conclusion
58
THE AHP ACADEMIC STAFF SELECTION MODEL
60
5.1
Introduction
60
5.2
Modeling
61
5.3
The AHP Model for Academic Staff Selection
61
5.3.1 Development of AHP hierarchy
62
5.3.2 Development of a conceptual framework
63
5.3.3 Decision Hierarchy
63
5.3.3.1 Selection Criteria
64
5.3.3.2 Hierarchical Structure
68
5.3.3.3 Importance of the criteria
69
5.3.3.4 Discordance levels
70
5.3.3.5
70
Rating Scales
5.3.4 Information from Experts
71
5.3.5
Employment of pair-wise comparison
72
5.3.5.1
Pair-wise comparison matrix
72
5.3.5.2
Computational of Eigenvalue (
max),
Consistency
Index
(CI),
Consistency Ratio (CR) and weights
or priorities of the criteria or subcriteria.
5.4
6
Summary
76
85
RESULTS AND ANALYSIS
88
6.1
Introduction
88
6.2
Rating Each Candidate
89
6.3
Result and Analysis
96
6.3.1
Candidates Priority Weights
96
6.3.1.1
Priority Weights
97
6.3.1.2
Calculation of Priority Weights
106
6.3.2
Sensitivity Analysis
109
x
6.3.2.1
Performance Sensitivity
109
6.3.2.2
Dynamic Sensitivity
111
6.3.2.3
Gradient Sensitivity Graph
112
6.3.2.4
Head-to-Head Sensitivity Analysis
113
6.3.2.5
Two-Dimensional Sensitivity
115
6.3.2.6
Impact of Changes in Weighting of
Criteria Using Sensitivity Analysis
6.4
7
Summary
117
122
SUMMARY AND CONCLUSION
123
7.1
Introduction
123
7.2
Summary
124
7.3
Suggestion for Future Work
128
REFERENCES
129
Appendix A – C
133 – 163
xi
LIST OF TABLES
TABLE NO.
TITLE
PAGE
3.1
Scale of Relative Importance
31
3.2
Random Consistency Table
35
4.1
Mark distribution in the marking rubric
58
5.1
Sample Pair-wise Comparison Table
69
5.2
Pair-wise comparison matrix for the five selection
criteria
5.3
Pair-wise comparison matrix for the criterion of
Knowledge, KNW
5.4
74
Pair-wise comparison matrix for the sub-criterion of
General Traits, GNT
5.6
73
Pair-wise comparison matrix for the criterion of
Working experience/ interpersonal skill, EIT
5.5
73
74
Pair-wise comparison matrix for the sub-criterion of
Scholar / Extracurricular activities, ACT
75
5.7
Priority of the sub-criteria of Knowledge
81
6.1
Pair-wise comparison of candidate profile for the subcriterion of Knowledge: Professional Knowledge (PFS)
6.2
Pair-wise comparison of candidate profile for the subcriterion of Knowledge: Education Background (EDU)
6.3
90
Pair-wise comparison of candidate profile for the subcriterion of Knowledge: General Knowledge (GNL)
6.4
90
90
Pair-wise comparison of candidate profile for the subcriterion of Working Experience / Interpersonal Skill:
Working Duration (PRD)
91
xii
6.5
Pair-wise comparison of candidate profile for the subcriterion of Working Experience / Interpersonal Skill:
Working Field (FLD)
6.6
91
Pair-wise comparison of candidate profile for the subcriterion of Working Experience / Interpersonal Skill:
Teaching Experience/ Skill (TCH)
6.7
91
Pair-wise comparison of candidate profile for the subcriterion of Working Experience / Interpersonal Skill:
Independence (IDP)
6.8
92
Pair-wise comparison of candidate profile for the subcriterion of Working Experience / Interpersonal Skill:
Intelligence / Eloquence (ITL)
6.9
92
Pair-wise comparison of candidate profile for the subcriterion of Working Experience / Interpersonal Skill:
Ability to Communicate in English (ABL)
6.10
Pair-wise comparison of candidate profile for the subcriterion of General Traits: Manner/ Politeness (MNR)
6.11
93
Pair-wise comparison of candidate profile for the subcriterion of General Traits: Professional Interest (PIT)
6.14
93
Pair-wise comparison of candidate profile for the subcriterion of General Traits: Age (AGE)
6.13
93
Pair-wise comparison of candidate profile for the subcriterion of General Traits: Appearance (APR)
6.12
92
94
Pair-wise comparison of candidate profile for the subcriterion of Scholar/ Extra Curricular Activities:
94
Publication (PBL)
6.15
Pair-wise comparison of candidate profile for the subcriterion of Scholar/ Extra Curricular Activities:
94
Researches (RSH)
6.16
Pair-wise comparison of candidate profile for the subcriterion of Scholar/ Extra Curricular Activities:
Rewards (RWD)
95
xiii
6.17
Pair-wise comparison of candidate profile for the subcriterion of Scholar/ Extra Curricular Activities:
Experience with Diverse Population (PPL)
6.18
95
Pair-wise comparison of candidate profile for the
criterion of References (RFC)
95
6.19
Candidate priority weights synthesized with respect to
98
the criteria, sub-criteria and the Goal
6.20
Candidate priority weights synthesized with respect to
106
the KNW
6.21
Priority weights of the criteria weight and Candidate 3
107
and 5
6.22
Ranking of candidates after changes of criteria weight is
119
made
6.23
Calculation of global priority weight
121
xiv
LIST OF FIGURES
FIGURE NO.
TITLE
PAGE
2.1
The Hiring Process
14
2.2
University-Wide Selection Process
16
2.3
Committee-Defined Selection Process
17
2.4
A schematic diagram of the AHP Process
20
2.5
Analytic Hierarchy Process
21
3.1
AHP steps used in this study
36
3.2
Research framework
39
4.1
Number of Staff and Technical Personnel in the
Faculty of Science
44
4.2
Detail of Academic Staff in the faculty
45
4.3
Intake and Appointment of Academic Staff flow chart
in UTM
4.4
The overall Process of Academic Staff Selection in
Faculty
5.1
51
54
Hierarchical Structure for the Selection of Academic
Staff
68
5.2
Data grid for the input of criteria weights
75
5.3
Priorities of criteria level 1
76
5.4
The priorities for the criteria in the selection of the best
candidate as the most appropriate academic staff in
Faculty of Science, UTM
5.5
82
The priorities for the sub-criterion of Knowledge
(KNW)
82
xv
5.6
The priorities for the sub-criteria of Working
experience / Interpersonal skill (EIT)
5.7
83
The priorities for the sub-criteria of General Traits
(GNT)
5.8
The
priorities
83
for
the
sub-criteria
of
Scholar/
Extracurricular Activities (ACT)
5.9
84
The academic staff selection criteria and sub-criteria
used in the AHP model and their local and global
priorities
5.10
85
Summary of the work flow in Chapter 5
85
6.1
Priority weights of candidates synthesized with respect
to Professional Knowledge
6.2
Normalized priority weights of candidates synthesized
with respect to Professional Knowledge
6.3
101
Priority weights of candidates synthesized with respect
to Ability to Communicate in English
6.11
101
Priority weights of candidates synthesized with respect
to Intelligence / Eloquence
6.10
101
Priority weights of candidates synthesized with respect
to Independence
6.9
100
Priority weights of candidates synthesized with respect
to Teaching Experience / Skill
6.8
100
Priority weights of candidates synthesized with respect
to Working Field
6.7
100
Priority weights of candidates synthesized with respect
to Working Duration
6.6
100
Priority weights of candidates synthesized with respect
to General Knowledge
6.5
99
Priority weights of candidates synthesized with respect
to Education Background
6.4
99
101
Priority weights of candidates synthesized with respect
to Manner / Politeness
102
xvi
6.12
Priority weights of candidates synthesized with respect
to Appearance
6.13
Priority weights of candidates synthesized with respect
to Age
6.14
104
Priority weights of candidates synthesized with respect
to Goal – Best candidates
6.21
103
Priority weights of candidates synthesized with respect
to References
6.20
103
Priority weights of candidates synthesized with respect
to Experience with Diverse Population
6.19
103
Priority weights of candidates synthesized with respect
to Rewards
6.18
103
Priority weights of candidates synthesized with respect
to Researches
6.17
102
Priority weights of candidates synthesized with respect
to Publication
6.16
102
Priority weights of candidates synthesized with respect
to Professional Interest
6.15
102
104
Priority weights of the sub-criterion, criterion RFC and
Goal in the selection process
105
6.22
Performance sensitivity graph with respect to Goal
110
6.23
Dynamic sensitivity graph with respect to the Goal
111
6.24
Gradient sensitivity graph with respect to Goal
113
6.25
Head-to-Head Sensitivity Analysis Graph for the
comparison between Candidate 1 and Candidate 2
6.26
Head-to-Head Sensitivity Analysis Graph for the
comparison between Candidate 4 and Candidate 1
6.27
116
Performance Sensitivity for nodes below Goal (after
changes of criteria weight made)
6.29
115
Two-Dimensional Sensitivity Graph synthesized with
respect to KNW and ACT
6.28
114
118
Dynamic Sensitivity for nodes below Goal (after
changes of criteria weight made)
118
xvii
6.30
Gradient Sensitivity Graph drawn with respect to ACT
(after changes of criteria weight made)
119
6.31
Hierarchy Tree with the Priority Weights
120
7.1
Implementation of AHP aided academic staff selection
model in the selection process in Faculty of Science,
UTM
125
xvii
LIST OF ABBREVIATIONS
AHP
-
Analytic Hierarchy Process
MCDM
-
Multiple Criteria Decision Making
UTM
-
University of Technology Malaysia
MADM
-
Multi-attribute decision making
KA
-
knowledge acquisition
CR
-
Consistency ratio
CI
-
consistency index
HCM
-
Human Capital Management
IIS
-
Ibnu Sina Institute for Fundamental Science Studies
APSI
-
Advance Photonic Science Institute
TP(P)
-
Deputy registrar (services)
PP(HRM)1
-
assistant of registrar (Human Resource Management)
PT(P/O)
-
administrative assistant of deputy registrar
JP
-
Selection Committee
JE
-
Executive Committee
JPA
-
Jabatan Perkhidmatan Awam
KNW
-
Knowledge
GNT
-
General Traits
ACT
-
Scholar/ Extracurricular Activities
EIT
-
Working Experience/ Interpersonal Skill
RFC
-
References
PFS
-
Professional Knowledge
EDU
-
Education Background
GNL
-
General Knowledge
ITL
-
Intelligence / Eloquence
xviii
ABL
-
Ability to Communicate in English
FLD
-
Working Field
TCH
-
Teaching Experience / Skill
PRD
-
Working Duration
IDP
-
Independence
APR
-
Appearance
MNR
-
Manner/ Politeness
PIT
-
Professional Interest
AGE
-
Age
PBL
-
Publication
RSH
-
Researches
RWD
-
Rewards
PPL
-
Experience with Diverse Population
xix
LIST OF SYMBOLS
Cn
-
n Criteria
aij
-
Element in a matrix, judgment on a pair of
criteria Ci, Cj
Wi
-
Weights
λ max
-
Largest eigenvalue
A
-
Matrix A
a, b, c, d, e, f -
scores
obtained
by
applicant
during
interview
xn
-
Relative importance by n experts
-
Geometric mean
I
-
Identity Matrix
<
-
Less than
W
-
Matrix W, Eigenvector
w1, w2, w3
-
Weights (elements of W)
wi =
-
Priority Weight of candidate
-
Weights of the j sub-criteria
-
Priority weights of the i candidates on the j
sub-criterion
xx
LIST OF APPENDICES
APPENDIX
A
TITLE
PAGE
Sample of documents obtained from knowledge
133
acquisition of case study
B
Pair-wise comparison table
148
C
Sensitivity analysis
152
1
CHAPTER 1
INTRODUCTION
1.1
Introduction
Personnel selection is a process of identifying, weighting, and evaluating
candidates against job requirements. The basic idea of personnel selection is to
choose the best applicant for a job. A very common problem in the personnel
selection is that the biases of those doing the rating have a tendency to creep into the
selection process (Arvey, 1982).
It is desirable to establish adequate selection
attributes (criteria) to discuss the attributes carefully to ensure that the right person is
chosen (Tung, 1981).
Academic member selection is an important process for the universities as
this decision affects the quality of education and the success of the university. The
decision is to select the best candidate for the faculty. Decision committee faces up
to the uncertainty and vagueness in the decision-making process (Ertugrul and
Karakasoglu, 2007). Analytic Hierarchy Process (AHP) can be applied to decision
making in this areas. By using AHP, uncertainty and vagueness from subjective
perception and the experiences of decision maker can be effectively represented and
reached to a more effective decision. As academic members are related to the
2
success and failure of higher education institutions, well developed selection criteria
can signify the essential element of the position, attract a high quality pool of
applicants and provide a reliable standard that applicants can be considered against.
It is crucial that everyone in the selection committee understand the list of
selection criteria and use it as the focal point throughout candidate assessment.
Essential criteria are those teaching skills, past experiences, qualifications, abilities
and publications and researches that are relevant to the performance of the functions
of a person’s duties. Ensuring the selection criteria assists in laying the foundations
for future conversations around probation, performance and promotion in a more
objective, fair and effective manner. The selection criteria provide structure to assist
the selection committee in developing effective interview questions and in
identifying the applicants to measure their own suitability.
Additional information
extraneous to the decision criteria is excluded from deliberation in an effort to limit
rater bias. We believe the personnel selection process can be aided by some formal
decision making techniques, particularly the AHP with its emphasis on decision
making with intangible criteria (Gibney and Shang, 2007).
Academic member selection is a multi-criteria decision making problem and
selecting the best personnel among many alternatives is also a multi-criteria decision
making (MCDM) problem (Dagdeviren, 2008).
The selection procedures may vary
from university to university.
Multiple criteria decision making (MCDM) refers to making decisions in the
presence of multiple, usually conflicting criteria (Grandzol, 2005). It is one of the
well-known topics of decision making. Multicriteria is a term that comes in many
variants. The terms “multicriteria”, “multi-criteria” or “multiple criteria” in either
title or keywords yield different, but overlapping, returns, as do “multiattribute:,
“multi-attribute”, “multiple objective”, “multiobjective” and multi-objective”.
Multiple objectives tend to be associated with mathematical programming, while
3
multicriteria and multiattribute tend to be focused on the selection among a given set
of discrete alternatives (Olson, 2008). However, the terms are used as synonyms in
this study. Selection decisions are challenging because the balancing of multiple,
often conflicting attributes, criteria, or objectives are required (Olson, 2008). The
overall goal of any MCDA process is not to make a decision, but rather to support a
decision making process (Linkov et al, 2007).
1.2
Background of the problem
In this study, the decision is to aid human’s ability to select the best candidate
for academic staff of faculty using Analytical Hierarchy Process (AHP).
Identification of best candidates is vital to ensure the quality of education,
effectiveness of program and activities attended in a university. Traditional methods
choose people for specified jobs, not for an organization as a whole or for subunits,
work groups, or teams (Guion, 1998).
Hiring the wrong person may lead to
dysfunctional of departments, dissatisfied students and eventually repeat efforts. The
selection process is an important focus since it is a success factor to meeting a
university’s objectives of productivity, industrial harmony and growth.
A good selection system is a practical tool for users, based on sound theory
and research, allows for flexibility in the environment, treats candidates equitably,
achieves its purpose cost-effectively and facilitates organizational effectiveness. In
addition, the evaluation process in a personnel selection process, by its very nature is
subjective, leaving many areas open to bias and error. Traditional methods and
criteria of personnel selection are ineffective and it is necessary to find some modern
techniques since there will be arguments arose. Although evaluating applicants with
written and oral exams is essential when employing the personnel needed, but it is
not sufficient alone.
4
In personnel selection, criteria or factors that are to be the basis of assessment
and evaluation must be specified and the weights of these criteria must be determined.
The selection criteria provide structure to assist the selection committee in
developing effective interview questions and in identifying the applicant(s) best
suited to perform effectively, as well as for applicants to measure their own
suitability.
Each criterion has a different importance or weight in personnel
assessment.
Therefore, unsatisfactory selection may occur with assessment and
evaluation tools, such as written or oral exams and tests which are not based upon
certain criteria and weights (Dagdeviren and Yuksel, 2007). There are a couple of
issues over this practice. Often, the evaluations are individual assessments where
each candidate is evaluated based on the predetermined selection criteria.
Normally, the best applicant would be the one with the highest total score, but
in real, a candidate who does not have the highest academic qualification may be
better in terms of his working leadership and ethics while a candidate who is great in
research work may not be good in teaching.
Besides, the selection committee
members from the different professional backgrounds may weigh the selection
criteria differently. A member from the academic background may weigh academic
qualification and scientific works more than the member with the management
background whose priorities may be for working experience and some general traits.
The same candidate may receive uneven evaluations from the different committee
members (Mamat and Daniel, 2004). Generally, during the process of selection,
there are individual biases and stereotypes.
As in many decision problems, academic member selection problem is too
complicated.
Human generally fail to make a good prediction for quantitative
problems, whereas comparatively having a good guess in qualitative forecasting.
The nature of the faculty member selection enables the use of well-known multicriteria decision-making method called the analytic hierarchy process (AHP) to deal
with the issues mentioned above. AHP is a systematic approach for representing the
elements of this problem, hierarchically.
AHP permits pair-wise comparison
5
judgments (which are documented and can be re-examined) to express the relative
strength or intensity of impact of the elements (criteria) in the hierarchy. These
judgments are then translated to numbers. This help to improve the traditional
method and simplify the process of selecting a best candidate to become the
academic member in this study by considering the criteria that may influence the
decision made. AHP is a planning methodology which incorporates the best of
formal, incremental, and systemic paradigms (Saaty and Kearns, 1985).
It is important to select an appropriate academic staff for a university to
ensure the standard and quality and success of a university. Therefore, there is a
need to develop a suitable and effective model to improve the existing academic staff
selection model in the university. The intent of this study is to show the application
of a model that is not overly complex and that does legitimately aggregate across
scales can serve to formalize a decision process, reduce time commitments, create a
process orientation, document the strategy, and result in better decisions.
1.3
Statement of the problem
In order to develop the model of academic staff selection, this study will
embark on the problem of finding out the potential criteria which affect the selection
of academic staff particularly the efficiency of selection process with a mechanism
which can be free from biasness.
Selection of academic staffs in university, particularly UTM, has been done
using a guideline outlined by the Registrar Office and adapted by the various
departments in various faculties in the university. Even though the selection process
adopted by the departments of the faculties do include elements of quantitative
6
evaluations as well as some qualitative assessments on the potential candidates, the
various departments does not use any standardized or established selection model
customized for the university.
1.4
Objectives of the study
The main aim of the study is to develop a decision making model to aid the
personnel selection committee to select the most appropriate candidate to be the
faculty academic staff in university. The specific objectives are:
1.4.1
To investigate and propose the potential criteria which affect the selection
process of academic staff in university.
1.4.3
To develop a decision making model incorporating AHP to assist in the
selection of most appropriate academic staff in university.
1.5
Scope of the study
This research on academic staff selection problem is to identify the criteria
for selection of academic staff in university using knowledge acquisition phases and
to assess the information which involves multi criteria decision making ability by
Analytic Hierarchy Process. Faculty members are those who research, those who do
teach and those who do both (Grandzol, 2005). The following are the scopes of the
study:
7
(a) The study is focused on the application of AHP as one of the tools for
implementing multiple criteria academic staff selection in university, not on
the best methodology for the problem.
(b) Several criteria considered to be relevant to the selection decision are used in
the ranking of the academic staff selection problem.
(c) A decision hierarchy, which is the most important aspect of the study, will be
constructed.
In line with this, the criteria or exact requirement of the
selection of faculty academic staff will be carefully studied and considered,
as inadequate and inappropriate information will affect utilization and
efficiency of the decision making tool developed.
(d) The profiles of candidates used in simulation of the selection model
developed are simulated data. Collection of actual profiles of those already
holding the posts, as well as those of the potential candidates to the posts, is
not done due to the sensitivity of the data involved and also the time
constraint of this project.
(e) An in-depth study on the selection process adopted by the Registrar Office of
University of Technology Malaysia (UTM) is carried out in this research in
order to identify mainly the more relevant selection criteria used in the
selection process of academic staffs in university. This is not a case study of
whether the selection of academic staffs of UTM has been done efficiently
and unbiasly.
8
1.6
Significance of the study
This research will contribute toward the enhancement of a decision making
tool to effectively aid the selection process of academic staffs in university. The
enhancement is achieved through the introduction of AHP to the normally practiced
selection process. This tool will give support to decision making in problems that are
too complex to be solved by the intuitive use of common sense alone and in this case,
the selection process of academic staffs in university.
In addition, the developed AHP selection model will aid the personnel
selection committee who are not extensively trained in AHP quickly understood the
process to select the most appropriate faculty academic staff in university. The study
will contribute an availability of results and solutions for comparison and validation
purposes in related researches or industries.
In this research, the background information regarding the selection process
is first presented. Next, the AHP as a personnel recruiting tool is described, and the
methodology is applied within the context of the faculty academic staff selection
process.
1.7 Structure of Dissertation
This structure of the dissertation is framed into 7 major chapters and a brief
of description and summary of the content of each chapter is presented as following:
9
Chapter 1 gave the introduction of this study, discussed on the research
project undertaken and the importance of the research. It then justifies the need for
the research, aims, scope, objective and limitation of the research.
In Chapter 2, the general idea on the selection of academic staff in Faculty of
Science, UTM is discussed. Literature reviews for Analytic Hierarchy Process (AHP)
as one of the Multi-Criteria Decision Making (MCDM) method, process and model
of the selection of academic staff as well as the brief discussion on the application of
AHP in the selection of academic staff in a faculty are done.
Chapter 3 described the methodology adopted in this study. The knowledge
acquisition from experts and the overview of application of AHP in the academic
staff selection process in Faculty of Science, UTM are discussed in detail.
Whereas in Chapter 4, a case study on the existing academic staff selection
process of Faculty of Science, UTM is done. The whole selection process which
includes the flow of the process, selection committee members involved and the
selection criteria are explained in detail in the chapter.
Chapter 5 presented the AHP aided academic staff selection model developed
in this study. The selection criteria and their sub-criteria determined and the priority
weights of the selection criteria and the sub-criteria synthesized are discussed in this
chapter. The calculations involved are shown in detail.
Chapter 6 detailed the implementation of AHP model in the academic staff
selection process in the Faculty of Science. Ranking of candidates are done in order
to select the best candidate for the faculty and the discussion as well as the analysis is
10
done and presented using appropriate graphs and charts by the help of Expert Choice
11.5.
Lastly, Chapter 7 draws the conclusion and summary of the research done in
this study. The finding of the research is concluded and some suggestions are listed
for future works in the same area.
13
CHAPTER 2
LITERATURE REVIEW
2.1
Introduction
This chapter discusses general idea on the academic staff selection, selection
methodologies and the selection criteria. It begins with the overview of the academic
staff selection in a faculty.
Analytic Hierarchy Process as a Multiple Criteria
Decision Making (MCDM) tool is studied.
Literature reviews is also done on the
application of AHP in the selection of academic staff.
A closer look on academic staff selection as proposed by Grandzol (2005) as
one of the approaches to improve faculty selection process is one of the literature
reviews done as detailed in section 2.3 in this chapter. Other related works done
previously by other researches also discussed in the later section.
12
2.2
Overview of Faculty Academic Staff Selection Problem
The faculty academic staff selection process requires a structured problem
solving methodology that considers the university’s context. Candidates of faculty
member need to go through a number of procedures such as interviews, written tests,
presentations, video presentations and so on before the final decision is made
(Mamat and Daniel, 2007). The traditional selection process uses an experimental
and statistical techniques approach which generally has individual biases and
stereotypes (Golec and Kahya, 2007). In any evaluation process, there is always a
possibility for error to occur. Some of the typical rating errors influencing the
selection decision are halo effect, horn effect, error of central tendency and
stereotyping or initial impression (Grensing, 1986).
Three main conflicts are identified in group decision-making. First, there may
be a group leader, or leaders, who play a dominant or more important role in the
selection of solutions. The group members thus do not have equal ‘importance
(weight)’ in a decision activity. Second, group members often have different ideas of
selection criteria, such as for the goals or priorities of the decision objectives. Third,
preferences of group members for alternative solutions would be expected to vary
from one to another. Hence, determining a ‘best’ satisfactory solution in a group
requires the aggregation of individual roles, preferences and judgments. (Zhang and
Lu, 2003). Therefore, there exists multiple criteria decision making model to ensure
an acceptable and efficient selection process.
In the planning stage of recruitment, selection committee members are
nominated. The selection committee responsible in making the decision is often a
mix of academic and human resource personals. Often, they are the President, the
Deans and / or the Associate Deans, the head of human resource and/ or the assistant
(Mamat and Daniel, 2004). Selection can be defined as choosing the most suitable or
qualified applicant who matches an organization’s criteria and has the potential for
13
successful job performance. This process is a critical gateway to ensure that the right
candidates are chosen and unsuitable ones filtered (Hishamuddin, 2000).
Its
activities usually follow a standard pattern, beginning with an initial screening
interview and concluding with the final employment as shown in Figure 2.1 (Taylor
III et al., 1998).
In a case study done by Grandzol (2005) at a mid-sized, state-related school
located in a semi-rural area of the northeastern United States, AHP was used to
improve the faculty selection process. There are existing detailed procedures and
guidelines for faculty searches at the university.
In the university selection,
affirmative action and bargaining unit issues made compliance mandatory and a stepby-step sequence of recruitment activities was well defined.
The selection process starts with preparing job description which including
the qualifications. This is followed by establishing committee using departmental
procedures and briefing committee by the Director of Social Equity.
Next,
development of criteria or rating sheet and advertisement of position will be done,
followed by the review of applications. The selection committee selects candidates
for further consideration after the application has been received. Then, campus visits
will be conducted by the selection committee. Finally, the department of faculty will
do discussion or recommendation of candidate for the academic staff in a university.
Recognizing that the absence of any clearly defined assessment process leads
to undue redundancy, inefficient use of time, needless deliberations, and insufficient
audit trails, the faculty selection committee discussed and developed another more
detail procedure for the selection process.
14
,ƵŵĂŶƌĞƐŽƵƌĐĞƉůĂŶ
/ĚĞŶƚŝĨŝĐĂƚŝŽŶŽĨŶĞĞĚƐĨŽƌ
ƉĞŽƉůĞ
ZĞĐƌƵŝƚŵĞŶƚ
/ŶƚĞƌŶĂůĂŶĚdžƚĞƌŶĂů^ŽƵƌĐĞƐ
WŽŽůŽĨĐĂŶĚŝĚĂƚĞƐ
:ŽďƌĞƋƵŝƌĞŵĞŶƚƐ
:ŽďĚĞƐĐƌŝƉƚŝŽŶƐĂŶĚ
ƐƉĞĐŝĨŝĐĂƚŝŽŶƐ
^ĞůĞĐƚŝŽŶ
ŽŵƉĂƌŝƐŽŶŽĨĐĂŶĚŝĚĂƚĞƐ
ǁŝƚŚũŽďƌĞƋƵŝƌĞŵĞŶƚǀŝĂ
ŝŶƚĞƌǀŝĞǁƐ͕ďĂĐŬŐƌŽƵŶĚ
ĐŚĞĐŬƐ͕ƚĞƐƚƐ͕ĞƚĐ͘
ĐĐĞƉƚĞĚĐĂŶĚŝĚĂƚĞƐ
ZĞũĞĐƚĞĚĐĂŶĚŝĚĂƚĞƐ
:ŽďƉůĂĐĞŵĞŶƚ
Figure 2.1. The Hiring Process.
First, the selection committee formulates preference structure to identify the
assessment criteria for the candidates. This will decide on the selection criteria most
important to the department of management for the position at hand. The importance
of the criteria is then differentiated. At this step, the relative importance of the
criteria is differentiated by completing pair-wise comparison for each set of criteria
and sub-criteria.
questionnaires.
Each committee member completed software-generated
15
Next, a minimum discordance level of which establishing minimum
acceptable scores for each criterion is set. Then, the rating scale for each criterion is
created. This is to determine the appropriate scales to measure or assess each
candidate. Objective or quantitative measures will be applied to differentiate the
ratings when possible. This is followed by rating each of the candidates on each
criterion individually. To reach group consensus on candidate ratings, the committee
members review each accepted application individually and complete a rating
scorecard for each applicant.
The following step is to determine the initial ranking of candidates. This step
involved application of the AHP algorithm, which was accomplished using Expert
Choice (1995) software. By using the software, a pair-wise comparison for upper
echelon candidates can be performed. Finally, the committee recommends the best
candidate(s) based on the pair-wise comparison performed.
2.2.1 Flow Charts
Application of AHP starts at the stage of developing criteria for the selection
process of candidates in the university. The developed criteria are used to review the
application sent by candidates. Selection of candidates for further consideration and
discussion or recommendation of candidates will be based on the AHP aided
decision making tool. The detailed procedures and guidelines for faculty member
selection committee to select the appropriate candidates can be summarized as shown
in Figure 2.2 and Figure 2.3 shows the process of selection defined by the selection
committee. AHP is applied to determine the appropriate candidates based on the
criteria and the ranking obtained (Grandzol, 2005).
16
Prepare job description, including qualifications
Establish committee using departmental
Brief committee (Director of Social Equity)
input
Develop criteria or rating sheet
Advertise position
Review applications
Selection Model
using
input
Analytic
Hierarchy
Process
output
(AHP)
Select candidates for further consideration
Conduct campus visits
Discuss/ recommend candidate (department faculty)
Flow of Information
input
output
Descriptions
The input of relevant information into the selection model
developed
The processed information (output) being used as a critical
aid in the decision making process
Figure 2.2. University-Wide Selection Process
17
Figure 2.3. Committee-Defined Selection Process
18
2.3
Multiple criteria Decision Making (MCDM)
Decision making is intimately associated with every sphere of human life
(Rafikul, 2003). Multiple criteria are very important in judgmental decision making
and there have been many applications of multi-criteria models in decision support
system (Olson, 2008).
Multiple Criteria Decision Making (MCDM) is one of the
most widely used methods among the numerous approaches available for conflict
management. Practical problems are often characterized by several noncommensurable and competing (conflicting) criteria and there is no solution
satisfying all the criteria simultaneously in this approach. Therefore, a compromise
solution for problems with conflicting criteria should be determined to help decision
makers reach a final decision (Tzeng et al, 2002).
One of the important aspects of multiple criteria decision making problem
(MCDM) is the determination of criteria weights.
The criteria in a MCDM problem
might not be equally important. Some criterion may be strongly more important than
the others (Rafikul, 2003).
MCDM techniques can help decision makers distinguish the kernel of a
complicated problem by identifying different criteria on a categorized basis, thus
achieving a multi-dimensional decision.
Since the problem is thus scrutinized,
MCDM can provide better decision than from that of a single criterion, and
compromises to group conflicts can be achieved (Shih et al., 2005).
Multi-attribute decision making (MADM) problems involve the design of a
“best” alternative by considering the tradeoffs within a set of interacting design
constraints. It refers to making selections among some courses of action in the
presence of multiple, usually conflicting, attributes (Kahraman, 2008). MADM has
the purpose of selection or evaluation and is characterized by a finite number of
19
prescribed alternatives, criteria defined by attributes (into which constraints are
incorporated), implicit or ill-defined objectives and not much interaction with the
decision makers (Hwang and Yoon, 1981).
2.4
Analytic Hierarchy Process (AHP)
Saaty's Analytic Hierarchy Process (AHP) has been developed to solve
decision problems in various fields by prioritizing the alternatives using eigenvectors
and manipulations in matrix algebra over the past two decades (Tung and Tang,
1998). AHP has been used by decision makers all over the world to model problems
in more than 30 diverse areas including resource allocation, strategic planning and
public policy. It has been used to rank, select, evaluate, and benchmark a wide
variety of decision alternatives (Wasil and Golden, 2003).
AHP is a systematic procedure to represent hierarchically the elements of any
problem. It organizes the basic rationality by breaking down a problem into its
smaller and smaller constituent parts and then guides decision makers through a
series of pair-wise comparison judgments to express the relative strength or intensity
of impact of the elements in the hierarchy. The judgments are then translated to
numbers. The AHP includes procedures and principles used to synthesize the many
judgments to derive priorities among criteria and subsequently for alternative
solutions (Saaty and Kearns, 1985).
AHP is easy, comprehensive and logical. It can be used in both quantitative
and qualitative multi-criteria decision making problems and it is widely accepted by
the decision making community, be they the academics or the practitioners (Mamat
and Daniel, 2007).
It is a practical, versatile and powerful tool that explicitly
20
identifies the factors that matter and provides a consistent structure and process for
evaluating candidates (Gibney and Shang, 2007).
AHP shows how to use judgment and experience to analyze a complex
decision problem by combining both qualitative and quantitative aspects in a single
framework and generating a set of priorities for alternatives (Faggini and Lux, 2009).
It elicits opinions from experts or decision makers. The two advantages of AHP:
First, it adopts a pair-wise comparison process by comparing two objects at one time
to formulate a judgment as to their relative weight; Second, with an adequate
measurement, this method is more accurate (with less experimental error) to achieve
a higher level of consistency since it requires the respondents to think precisely
before giving their answers (Cheng and Heng, 2001).
Figure 2.4 shows the AHP, which follows an approach of pair-wise
comparison, provides a way for calibrating a numerical scale, particularly in new
areas where measurements and quantitative comparisons do not exist (Brent et al,
2007). The overall procedure of the AHP is shown in Figure 2.5.
Figure 2.4. A schematic diagram of the AHP Process.
21
ĞǀĞůŽƉŚŝĞƌĂƌĐŚLJŽĨ
ƉƌŽďůĞŵŝŶŐƌĂƉŚŝĐĂů
ƌĞƉƌĞƐĞŶƚĂƚŝŽŶ
KǀĞƌĂůůŐŽĂů͕ĐƌŝƚĞƌŝĂ͕ĂŶĚ
ĂƚƚƌŝďƵƚĞƐĂƌĞŝŶĚŝĨĨĞƌĞŶƚůĞǀĞůŽĨ
ŚŝĞƌĂƌĐŚLJ
ŽŶƐƚƌƵĐƚĂƉĂŝƌͲǁŝƐĞ
ĐŽŵƉĂƌŝƐŽŶŵĂƚƌŝdž
dǁŽĐƌŝƚĞƌŝĂĂƌĞĐŽŵƉĂƌĞĚĂƚ
ĞĂĐŚƚŝŵĞƚŽĨŝŶĚŽƵƚǁŚŝĐŚŽŶĞ
ŝƐŵŽƌĞŝŵƉŽƌƚĂŶƚ
^LJŶƚŚĞƐŝnjĂƚŝŽŶ
dŽĐĂůĐƵůĂƚĞƉƌŝŽƌŝƚLJŽĨĞĂĐŚ
ĐƌŝƚĞƌŝŽŶ
hŶĚĞƌŐŽĐŽŶƐŝƐƚĞŶĐLJƚĞƐƚ
dŽĐŚĞĐŬǁŚĞƚŚĞƌũƵĚŐŵĞŶƚ
ŽĨĚĞĐŝƐŝŽŶŵĂŬĞƌƐŝƐ
ĐŽŶƐŝƐƚĞŶƚ
ůůũƵĚŐŵĞŶƚƐĂƌĞ
ĐŽŶƐŝƐƚĞŶƚ͍
ŽŶƐŝƐƚĞŶĐLJŽĨĂůůũƵĚŐŵĞŶƚƐ
ŝŶĞĂĐŚůĞǀĞůŵƵƐƚďĞƚĞƐƚĞĚ
EŽ
zĞƐ
EŽ
ůůůĞǀĞůƐĂƌĞ
ĐŽŵƉĂƌĞĚ͍
ůůĐƌŝƚĞƌŝĂĂŶĚĂƚƚƌŝďƵƚĞƐŝŶ
ĞĂĐŚĐƌŝƚĞƌŝŽŶŵƵƐƚďĞƚĞƐƚĞĚ
zĞƐ
ĞǀĞůŽƉŽǀĞƌĂůůƉƌŝŽƌŝƚLJ
ƌĂŶŬŝŶŐ
Figure 2.5.
ĂƐĞĚŽŶĞĂĐŚĂƚƚƌŝďƵƚĞ͛Ɛ
ƉƌŝŽƌŝƚLJĂŶĚŝƚƐ
ĐŽƌƌĞƐƉŽŶĚŝŶŐĐƌŝƚĞƌŝŽŶ
ƉƌŝŽƌŝƚLJ
Analytic Hierarchy Process (Ho et al., 2006)
22
2.5
Academic Staff Selection Model
Multiple criteria methods, both qualitative and qualitative, were developed to
better model decision scenarios. The evaluation data of academic member suitability
for various subjective criteria and the weights of the criteria are usually expressed in
linguistic terms (Ertugrul and Karakasoglu, 2007). The most important criterion in
selecting the new faculty (academic) member is academic qualification followed by
working experience, leadership qualities and lastly general traits (Mamat and Daniel,
2004).
2.6
Selection Criteria
To determine hiring criteria, there is a need to examine experience, education,
intelligence and personality requirements. Some common criteria are intelligence,
communication skills, self-confidence, sociability, ambition and motivation,
leadership, adaptability and cooperativeness (Grensing, 1986).
The proposed
selection criteria by Mamat and Daniel (2007) to select an academic member
consisting of academic qualification, working experience, leadership qualities and
some general traits.
As described by Leigh et al (2007), assessments of knowledge, experience,
performance, practice-based skills and tasks in terms of competency are other critical
stages of professional development.
Formann (1992) has described that using some criteria being available per
applicant such adequacy of field of work, age, and number of applications, this
23
scaling procedure results in weights for each of the categories of the criteria
indicating the relative importance of each criterion, and scores for all applicants
pointing at their aptness. To avoid the decoy effects, not only must decision makers
provide explicit weights for the assessments to be used in making choices among job
finalists, but they must also use the pre-assigned weights when making a final
decision (Slaughter et al., 2006).
Avery and Renz (1992) in their paper accounted that an obvious feature
contributing to fairness in the practice of selecting employees is that each applicant
should be examined against a common set of criteria and standards.
According to Hishamuddin (2000), statistical strategy where all of the
information collected is combined according to a mathematical formula and the
candidate with the highest score will be selected has been proven to be more reliable
and valid than a clinical strategy where decision maker subjectively evaluates all of
the information collected and makes an overall judgment.
2.7
Summary
This chapter provides an important background for the study of application of
AHP in the selection of academic staff in a faculty.
The idea of selection of
academic staff as well as its primary problem related to this study is presented.
Other related works done previously by other researches in this area are also being
discussed.
24
The methodology adopted in this research, knowledge acquisition process
which used to elicit expert’s knowledge in order to develop an AHP aided decision
making tool, principles of AHP and implementation of AHP in the selection of
faculty academic staff will be discussed in detail in the next chapter.
25
CHAPTER 3
RESEARCH METHODOLOGY
3.1
Introduction
This chapter describes the methodology adopted in order to realize the aim
and the objective of the research. There are two main sections in this chapter, i.e.,
details of knowledge acquisition adopted in this research and the overview on
Analytic Hierarchy Process as a decision making aiding tool.
3.2
Methodology
In this study, an attempt is made to select the most appropriate academic staff
in Faculty of Science, UTM. Analytic Hierarchy Process, AHP which is a multiple
criteria decision making technique has been applied in this study to establish a model
to select the most appropriate academic staff in Faculty of Science, UTM.
26
3.2.1
Knowledge Acquisition
Knowledge elicitation or knowledge acquisition (KA) techniques are
techniques developed to help elicit knowledge from an expert.
Knowledge
acquisition used to capture the experts’ knowledge in pre-qualification and selection
process and then transform the appropriate knowledge into a theoretical model.
Knowledge elicitation comprises a set of techniques and methods that attempt to
elicit an expert’s knowledge through some form of direct interaction with that expert
(Wilson and Corlett, 1995). Knowledge acquisition tools can be associated with
knowledge-based application problems and problem-solving methods.
This
descriptive approach provides a framework for analyzing and comparing tools and
techniques, and focuses the task of building knowledge-based system on the
knowledge acquisition process (Boose, 1989).
The knowledge that needs to be captured from the experts in this study are
the criteria and the importance of each criterion or sub-criterion compared to another
criterion or sub-criterion that affect the selection of faculty academic staff in a
university. The technique that has been used in this study to capture the experts’
knowledge is the interviews with experts.
Meanwhile, references from various
sources including analysis of existing documents and analysis of existing system are
used to obtain the useful information and to form the simulation model of the faculty
academic staff selection.
Although opinions and knowledge elicited from exerts may only provide a
very rough picture, it is still appropriate in this study. It is noteworthy that AHP does
not necessarily involve a large sample (Lam and Zhao, 1998).
27
3.2.1.1 Interviews
Interview is a process of collection of information through direct
communication between the interviewer and interviewee where question are asked
by the interviewer to obtain information from interviewee. In the interview, the
interviewer works directly with the respondent. This technique is one of the most
popular methods in getting information from experts. A semi-structured interview
was chosen to achieve the objective of the study. In a semi-structured interview, a
set of major questions are predetermined in advance and some follow up questions
may also be prepared. This type of interview is more flexible than the structured
approach since the interviewer has the opportunity to probe into areas that justify
further investigation (Hishamuddin, 2000) in the area of study.
The experts that have been interviewed in order to obtain useful information
in this study associate with academic staff selection process were the deputy registrar
from Registrars’ Office, UTM and the Heads of Departments from Faculty of
Science, UTM who are directly involved in the selection of academic staff.
Interviews done in this study is mainly to achieve the research objectives by
obtaining the information as follow:
(a) To verify the important criteria for the selection of faculty academic staff
(b) To obtain the sub-criteria information which used to access the main criteria
(c) To find out the relative importance of each criterion and sub-criterion
(d) The existing academic staff selection process
28
3.3
Analytic Hierarchy Process (AHP)
It is well known that the Analytic Hierarchy Process (AHP) developed by
Saaty in mid of 1970 is one of the most powerful approaches for decision aid in
solving of a multi criteria decision making (MCDM) problem.
The AHP is based
on sets pair wise comparison of decision maker that is represented on a human
being’s intrinsic ability to structure the perceptions hierarchically, compare pairs of
similar things against a given criteria or a common property and judge the intensity
of the importance of one thing over the other (Ciptomulyono, 2008).
AHP is used in both individual and group decision-making by bussiness and
industry and is particularly applicable to complex large-scale multiparty multicriteria decision problems. AHP has been applied to a variety of decisions involving
benefits, costs, opportunities, and risks and is particularly useful in problem solving.
In order to obtain an AHP ranking (i.e. overall relative weighting of the elements) the
AHP synthesizes all the judgments using the framework given by the hierarchy. To
do so, it is important to define a decision problems being in single decision elements
between which certain relationship exist. Hence, it does need to estimate relative
priority weights for the single decision elements on each hierarchy level of decision
problems.
In order to solve a MCDM problem, the AHP method could be utilized to
derive priorities based on judgment of decision maker. To achieve this objective, a
pair wise comparison technique needs to be constructed. The most common
techniques for an estimating relative priority weights is originally proposed
eigenvector method (Ciptomulyono, 2008).
29
3.3.1 Research Design and Procedure
AHP is proposed as a tool for implementing multiple criteria faculty
academic staff selection problem in this study. Identification of hierarchy and the
selection criteria are the key factors in using AHP. In AHP, a complex multicriteria
decision making problem can be handled by determining pertinent factors and then
structuring them into simple and comprehensible hierarchical structure. This hierarchical
structure descends in successive levels from an overall objective (goal) to various
dimensions and criteria, with numerical values (weights) assigned to each variable.
This method aids decision makers to maintain coherent thought patterns as they
reach conclusions.
The main objective or the goal of the problem structured as the first level or
top of the hierarchy. The problem is broken down into all possible related criteria
contributing to the decision process to ease the decision making process. These form
the intermediate level of the hierarchy and a list of the alternatives will form the
lowest level of the hierarchy.
The steps used in the implementation of AHP as multi-criteria decision
making tool in academic staff selection process in this study is explained in detail as
follow:
1. Define the decision problem. Researcher defines the decision problem clearly
since it drives the whole study. Researcher need to explain clearly what is the
problem and why AHP has to be used.
2. Develop a conceptual framework. Researcher decomposes the complexity of the
problem into different levels or components and synthesizes the relations of the
components.
30
3. Set up the decision hierarchy. The hierarchy formed consists of several levels
and different groups of related elements (criteria, sub-criteria, alternatives, etc.).
Each level consists of a finite number of decision elements. The formation of
hierarchy is based upon two assumptions (Cheng and Heng, 2001):
a. It is expected that each element of a level in the hierarchy would be
related to the elements at the adjacent levels.
AHP recognizes the
interaction between elements of two adjacent levels.
b. There is no hypothesized relationship between the elements of different
groups at the same level.
4. Obtain information from experts. AHP approach is a subjective methodology,
therefore, data can be obtained directly questioning the experts on the subject
matter. The experts in this study are the Deputy Registrar of UTM, Deputy Dean
of Faculty of Science, Assistant Registrar in Faculty of Science and Heads of
Departments in Faculty of Science.
5. Employ the pair-wise comparison. All the related elements are compared using
the priority scale pair by pair. AHP is used to derive relative priorities on
absolute scales (invariant under the identity transformation) from both discrete
and continuous paired comparisons in multilevel hierarchic structures. These
comparisons may be taken from actual measurements or from a fundamental
scale that reflects the relative strength of preferences and feelings (Saaty and
Vargas, 2006). A pair-wise comparison matrix or judgment matrix for all the
criteria is constructed. Saaty’s scale of measurement is used to rate the intensity
of importance between two elements (Saaty and Kearns, 1985) which is shown in
Table 3.1.
31
Table 3.1. Scale of Relative Importance
Intensity of
relative
Definition
Explanation
importance
1
3
5
7
Two activities contribute equally to
Equal importance.
the objective.
Moderate importance of Experience and judgment slightly
one over another.
Essential
favor one activity over another.
or
strong Experience and judgment strongly
importance.
favor one activity over another.
An activity is strongly favored and its
Demonstrated
dominance
importance.
is
demonstrated
in
practice.
The evidence favoring one activity
9
Extreme importance.
over another is of the highest possible
order of affirmation.
Intermediate
2,4,6,8
values
between the two adjacent When compromise is needed.
judgments.
If an activity has one of
the above numbers (e.g.
Reciprocals
3)
compared
with
a
of above
second activity, then the
non-zero
second activity has the
numbers
reciprocal
value
(1.e.,
1/3) when compared to
the first.
6. Estimate relative weights of elements. The basic idea of the AHP is to introduce
structure and objectivity into the largely subjective process of attaching weights
to a set of decision criteria in any multicriteria decision-making situation
(Tadisina and Bhasin, 1989). A vector of priorities (Eigenvector) in a matrix will
32
be calculated and then normalized to sum 1.0 or 100 per cent after the pair-wise
comparison matrix is formed. This can be done by dividing the elements of each
column of the matrix by the sum of that column. Then, obtaining the Eigen
vector by adding the elements in each resulting row (row sum) and dividing this
sum by the number of elements in the row (priority weight).
(a) Establish pair-wise comparison matrix A. A pair-wise comparison is carried
out based on the criteria of the upper hierarchy. Assuming n criteria exists,
there are n(n–1)/2 criteria of the pair-wise comparison to be derived. Let C1,
C2,…., Cn denote the set of criteria, while aij represents a judgment on a pair
of criteria Ci, Cj. An n by n matrix A can be represented as follows:
A = [aij]=
C1
C2
...
Cn
C1
1
a12
…
a1n
C2
1
a12
1
…
a2n
.
.
.
.
a.
.
.
.
.
Cn
1
a1n
1
a2 n
.
1
.
.
…
1
The result of pair-wise comparison of the n criteria are put at the upper
triangle of pair-wise comparison matrix A. The lower triangle shows the
value of the relative positions for the reciprocal values of the upper triangle.
Where aii = 1 and aji =
1
, I, j = 1, 2, …, n, two criteria (Ci, Cj) become one
aij
quantization value for an important relative judgment. In matrix A, aij can be
expressed as a set of numerical weights, W1, W2, …, Wn, in which the
recorded judgments must be assigned to the n criteria C1, C2,…., Cn. If A is a
33
consistency matrix, relations between weights Wi and judgments aij are
simply given by
A=
Wi
сaij ( for i, j = 1, 2,…, n) and matrix A as follows:
Wj
C1
C2
...
Cn
w1
w1
w1
w2
...
C1
w1
wn
C2
w2
w1
w2
w2
...
w2
wn
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
wn
w1
wn
w2
...
Cn
wn
wn
(b) Compute the eigenvalue and eigenvector. Matrix A multiplies the criteria
weight vector (x) equal to nx, i.e., (A – nI)x = 0, where x is the eigenvalue (n)
of eigenvector. Given that aij denotes the subjective judgment of decision
makers, the actual value (
Wi
) has a certain degree of difference. Therefore,
Wj
Ax = nx cannot be set up. To find out the weights, the largest eigenvalue,
λ max of A must be first found. It has been demonstrated by Saaty (1980) that
the eigenvector corresponding to the largest eigenvalue of the matrix provides
the relative priorities of the criteria. The largest eigenvalue, λ max can be
shown by
λ max =
Wj
n
¦a
j =1
ij
Wi
, i = 1,2,…, n
n
For uniqueness, the set of weights is normalized such that
¦ w = 1.
i
If A is a
j =1
consistency matrix, the eigenvector X can be calculated by the formula:
34
; A Ͳ λ max IͿX = 0
7. Calculate the degree of consistency. Consistency test is employed to validate
the responses. Consistency ratio (CR) is used to measure the consistency in
the pair-wise comparison. Since the comparisons are carried out through
personal or subjective judgments, some degree of inconsistency may be
occurred. To guarantee the judgments are consistent, the final operation
called consistency verification, which is regarded as one of the most
advantages of the AHP, is incorporated in order to measure the degree of
consistency among the pair-wise comparison by computing the consistency
ratio (Ho, 2008). Consistency is needed to ensure that any change in the
priority scale in pair-wise comparisons matrix will not affect the results made
by the decision maker. The inconsistency ratio of below 10 percent or 0.10
from the value of CR is still acceptable (Saaty and Kearns, 1985). If the
consistency ratio exceeds the limit, the decision maker reviews and revises
the pair-wise comparisons.
According to Saaty and Ramanujam (1983), the consistency index, CI can be
calculated by using the formula:
CI =
(λ max − n )
( n − 1)
and the ratio of CI to the random index (average consistency indices), RI as
suggested by Saaty and Kearns (1985) as shown in Table 3.2 for the same
order matrix is called the consistency ratio, CR. i.e. :
CR =
CI
.
RI
35
Table 3.2. Random Consistency Table.
Size of matrix,
n
Random
Consistency
1
2
3
4
5
6
7
8
9
10
0
0
0.58
0.90
1.12
1.24
1.32
1.41
1.45
1.49
8. Compute the entire hierarchical priority weight. The entire hierarchy priority
weight is computed after various hierarchies and criteria priority weights are
estimated to aid decision makers to select the most appropriate candidate at the
end of the selection process.
Figure 3.1 shows the summary of AHP steps used in this study:
36
ĞĨŝŶĞƚŚĞĚĞĐŝƐŝŽŶƉƌŽďůĞŵ
ĞǀĞůŽƉĂĐŽŶĐĞƉƚƵĂůĨƌĂŵĞǁŽƌŬ
^ĞƚƵƉƚŚĞĚĞĐŝƐŝŽŶŚŝĞƌĂƌĐŚLJ
ŽůůĞĐƚŝŶĨŽƌŵĂƚŝŽŶĨƌŽŵĞdžƉĞƌƚƐ
ŵƉůŽLJƚŚĞƉĂŝƌͲǁŝƐĞĐŽŵƉĂƌŝƐŽŶ
ƐƚŝŵĂƚĞƌĞůĂƚŝǀĞǁĞŝŐŚƚƐŽĨĞůĞŵĞŶƚƐ
ĂůĐƵůĂƚĞƚŚĞĚĞŐƌĞĞŽĨĐŽŶƐŝƐƚĞŶĐLJ
ŽŵƉƵƚĞƚŚĞĞŶƚŝƌĞŚŝĞƌĂƌĐŚLJƉƌŝŽƌŝƚLJǁĞŝŐŚƚ
Figure 3.1 AHP steps used in this study.
3.3.2
Implementation of AHP – Development of Model for Faculty Academic
Staff Selection
AHP is a powerful and flexible decision making process to help decision
makers in setting priority weights and make the best decision when both qualitative
and quantitative aspect of a decision need to be considered.
It not only helps
decision makers arrive at the best decision, but also provides a clear rationale that the
decision made is the best.
37
AHP helps to capture both subjective and objective evaluation measures by
providing a useful mechanism for checking the consistency of the evaluation
measures and alternatives suggested by the selection committee to reduce bias in
decision making. Combined with meeting automation, the selection committee can
minimize common pitfalls of committee decision making process, such as lack of
focus, planning, participation or ownership, which ultimately are costly distractions
that can prevent committee from making the right choice.
The use of AHP in the candidate evaluation process is straightforward. Pairwise comparisons are used to produce a matrix of employment characteristics
weighted with respect to their importance to overall desirability. Next, another set of
comparisons produces matrices comparing candidates for each of the characteristics
(Taylor III et al., 1998).
3.3.3
Decision making
In an AHP model that to be applied in a real situation, it is required that the
priority of criteria must be specified and all the decision makers must agree on the
definitions of the criteria and sub-criteria and sign off on the completeness of the
model.
In this study, a simulation model will be developed by researcher by
considering all the possible related factors which will affect the selection model
based on the selection process in Faculty of Science, UTM. In the selection process,
researcher must use the same definition of the selection criteria for it to work
effectively.
38
The candidates with highest relative priority weights of elements (criteria)
will be chosen as the faculty academic staff in this AHP aided selection model.
3.4
Research Framework
The summary of the research framework in this study is as shown in Figure
3.2.
39
^ƚĂƌƚ
>ŝƚĞƌĂƚƵƌĞZĞǀŝĞǁ
DĂƚŚĞŵĂƚŝĐĂůDŽĚĞůĨŽƌƚŚĞ
,WƉƌŽďůĞŵ
DĂƚŚĞŵĂƚŝĐĂůĂŶĂůLJƐŝƐĂŶĚ
ŶƵŵĞƌŝĐĂůĐŽŵƉƵƚĂƚŝŽŶ
ŶĂůLJƐŝƐŽĨƌĞƐƵůƚƐ
ZĞĨĞƌĞŶĐĞƐĨƌŽŵǀĂƌŝŽƵƐ
ƐŽƵƌĐĞƐ
ŽĐƵŵĞŶƚĂƚŝŽŶ
ŽŶƐƚƌƵĐƚŝŽŶ
ŽĨŚŝĞƌĂƌĐŚŝĐĂů
ƚ
WĂŝƌͲǁŝƐĞĐŽŵƉĂƌŝƐŽŶ
ƵƐŝŶŐDŝĐƌŽƐŽĨƚdžĐĞů
džƉĞƌƚ
ŚŽŝĐĞ
ϭϭ͘ϱ
ƐƚŝŵĂƚŝŽŶŽĨƌĞůĂƚŝǀĞ
ƉƌŝŽƌŝƚLJǁĞŝŐŚƚƐ
ŶĚ
Figure 3.2
<ŶŽǁůĞĚŐĞĂĐƋƵŝƐŝƚŝŽŶ
ĨƌŽŵĞdžƉĞƌƚƐ
Research framework.
40
3.5
Summary
Knowledge acquisition is done by interviewing experts in this study. A
simulation model of AHP aided decision making tool was developed in this study.
As described by Nussbauma et al. (1999), simulation is an appropriate technique to
aid personal selection.
Microsoft Excel 2007 software is used in this study in synthesizing the pairwise comparison matrices and calculating the consistency ratio (CR) and consistency
index (CI) in order to make it quick and simple by eliminating tedious calculation.
Expert Choice 11.5 was software used in this study.
The software provides a
structured approach and proven process for prioritization and decision making. It
helps decision makers arrive at the best decision and provides a clear rationale for the
decision made. It is mainly used in the synthesis of the priority weights, discussion
and analyzing the results (ranking of candidates) using sensitivity analyses in this
study.
41
CHAPTER 4
KNOWLEDGE ACQUISITION FOR THE DEVELOPMENT OF MODEL
A CASE STUDY OF UTM
4.1
Introduction
A case study of UTM regarding the selection process of academic staff in
Faculty of Science is done and detailed in this chapter.
The background of
Registrar’s office and Faculty of Science in UTM and their process of selection of
academic staff process are studied. The selection process, selection committee and
the selection criteria are studied in a great depth in order to obtain the relevant
information to carry on with this study.
42
4.2
Background
UTM is the largest engineering-based university located at Skudai, Johor,
Malaysia. It is renowned for being at the forefront of engineering and technological
knowledge and expertise. UTM has also established a reputation for innovative
education and leading-edge research, with a vision to educating technologists and
professionals towards the development of creative human capital and advanced
technological innovations.
It has more than 20 specialist institutes and research centres, in addition to
academic faculties to service technological education and research needs of the
university. There are more than 25,000 full-time undergraduate students at its main campus
in Johor. In addition, there are more than 3,000 postgraduate students in various fields of
specialization based on the last update done on Jun 2009.
Having produced more than 200,000 technical graduates and qualified
professionals over the years, UTM has earned its place as Malaysia's premier
university in Engineering and Technology which inspires creativity and innovation.
In line with this, UTM has its own selection procedure and method in the process of
selecting their academic staff in order to ensure the quality of education and the
success of the university to fulfill vision and mission of the university.
4.2.1 Registrar’s Office
In line with the vision of the Registrar’s Office of UTM’s which is to ensure
efficient information support for the UTM administration and mission of the
43
university which is to provide an excellent service that satisfy its customer, the
Registrar’s Office of UTM aims to provide efficient, effective and quality service
based on excellent knowledge within a positive and productive working environment.
The following are the functions of the Registrar’s Office, UTM:
(a) To implement the university human capital policy
(b) To implement the university human capital development policy
(c) To implement the university academic policy
(d) To implement the management development policy
(e) To ensure safety university environment
The Registrar’s Office of UTM is divided into six sections to ensure the
smoothness in providing its services to the university:
(a) Human Capital Management Section
(b) Human Capital Development Section
(c) Academic Management Section
(d) Security Management Section
(e) System Development Section
(f) Registrar’s Office of Kuala Lumpur Branch
Among these sections, Human Capital Management Section is playing a
crucial role in areas related to the management the staffs of UTM. The services
provided by this section include the planning of human capital, intake and
appointment of staff, service examination of staff and personnel services.
The intake and appointment of academic staff is one of the jurisdictions of
Human Capital Management Section in the Registrar’s Office.
44
4.2.2
Faculty of Science
Faculty of Science consists of three main departments: the Department of
Chemistry, Physics, and Mathematics, plus two additional institutes: the Ibnu Sina
Institute for Fundamental Science Studies (IIS) and the Advance Photonic Science
Institute (APSI).
According to the last update done on April 2009, the faculty
currently has 190 academic staffs in the faculty including the permanent staff (160),
contract staff (13) and temporary staffs (17). Figure 4.1 shows the number of Staff
and Technical Personnel and Figure 4.2 shows the detail of Academic Staff in the
faculty.
Figure 4.1
Number of Staff and Technical Personnel in the Faculty of
Science
45
/ŶƚĞƌŶĂƚŝŽŶĂů
^ƚĂĨĨ͗
/ŶĚŽŶĞƐŝĂŶсϮ
WĂŬŝƐƚĂŶŝсϮ
^ƚĂĨĨŽŶƐƚƵĚLJ
ůĞĂǀĞ͗
WŚсϮϮ
Figure 4.2
Detail of Academic Staff in the faculty.
46
4.3
Selection Process
UTM is currently using the traditional selection method in selecting their
academic staff. The traditional evaluation process in a selection process, methods
and criteria of academic staff selection are ineffective and it is necessary to find some
modern techniques since there will be arguments arose by its subjective and leaving
many areas open to bias and error natures.
The initial process of selecting and certification of an academic staff in UTM
is done respectively by each of the faculties and approved by the UTM’s Executive
Committee, JE after the second interview done by the Selection Committee in the
Registrar’s Office. There is a Selection Committee responsible for the selection at
the faculty level which is chaired by Dean of the faculty.
4.3.1 Registrar’s Office
There is an Academic Staff Intake and Appointment Work Order (Arahan Kerja
Pengambilan dan Perlantikan Staf Akademik) prepared to explain in detail the whole
process of academic staff selection. The academic staffs mentioned are:
(a) Associate Professor (Profesor Madya) - Grade DS54
(b) Senior Lecturer (Pensyarah Kanan) - Grade DS52
(c) Lecturer (Pensyarah) - Grade DS45
(d) Teacher (Guru) - Grade DG41
(e) Tutor - Grade DA41
47
According to the Work Order by Registrar’s Office, these are some of the
definitions used in the process of appointment of an academic staff in UTM:
(a) Academic Staff
Academic staff refers to the University’s teaching force, including the post of
Associate Professor (Grade DS53/54), Senior Lecturer (Grade DS51/52),
Lecturer (Grade DS45), Teacher (Grade DG41) and Tutor (Grade DA41).
(b) Designation Warrant (Waran Perjawatan)
This is a letter of allocation approval of permanent post from Treasury.
(c) University Staff Designation Statistic (Perangkaan Perjawatan Staf Universiti)
This is a distribution and filling of posts at faculties or departments which is
done monthly by the Human Resource Management of Registrar’s Office.
(d) Short-list (Senarai Pendek)
This is the list of candidates identified by the relevant faculties from a pool of
applicants. The candidates identified are applicants those have fulfilled the
university’s requirement.
(e) Selection Committee for the post of Teacher except for the post of
Professorship (Jawatankuasa Pemilih Bagi Jawatan Guru Selain Jawatan
Kursi)
This is a committee which will give advice and certification to Appointing
authorities (Pihak Berkuasa Melantik) in a staff selection to be service in the
university.
(f) Executive Committee (Jawatankuasa Eksekutif, JE)
This is the Appointing Authorities which will approve the certification in
academic staff selection. This Committee is chaired by the Vice Chancellor
and the members are: Deputy Vice Chancellor (Academic), Deputy Vice
Chancellor (Development), Deputy Vice Chancellor (Student Affairs),
Registrar, Bursary and the Head of Librarians.
48
Both of the faculties or departments will have to follow the same selection
guideline given by the Registrar’s Office in selecting the most appropriate candidate
to be the member of faculty in the selection process. The selection process is done
under the responsibility of Deputy registrar (services) (Timbalan Pendaftar
(Perkhidmatan), TP(P)), assistant of registrar (Human Resource Management)
(Penolong
Pendaftar(Human
Resource
Management)
,
PP(HRM)1)
and
administrative assistant of deputy registrar (Timbalan Pendaftar (Perkeranian/
Operasi), TP(P/O)). The details of the whole selection process are explained as the
following:
Step 1:
The selection process starts with the acceptation of candidates certified or approved
by the relevant faculty for the consideration of Executive Committee (JE) after a
series of procedures in the selection process at faculty level have been done. It is
critically important that at the outset there is clarity about precisely what is sought in
each appointment and that this is reflected in the detailed Job Description and Person
Specification prepared by the Faculty. It is crucial that everyone involved in a
Selection Committee at the faculty level understands the list of selection criteria and
use them as the focal point throughout candidate assessment. The selection must be
based on merit in relation to the selection criteria.
Step 2:
Next, the information certified by the faculty will be checked by TP(P), PP(HRM)1
and PT(P/O).
The selection procedure will proceed to the next step if the
information provided is complete. Otherwise, this process will have to go back to
Step 1. The information from faculty which needs to be checked including the
application forms, Personnel Information of Candidate (Jadual Maklumat Calon) and
the copies of academic certificates.
49
Step 3:
A call letter of interview will be issued at this step. This must be done within 5
working days before the date of interview held. An offer letter will be issued directly
to the candidate without going through the process of interview for the candidate
who has achieved the CGPA of first class.
Step 4:
Preparation of information and documents is done at this stage for the meeting of the
Selection Committee (Jawatankuasa Pemilihaan, JP).
Step 5:
A JP Meeting is held for the preparation of the interview session.
Step 6:
An Executive Committee (Jawatankuasa Eksekutif, JE) working paper approval
(kertas kerja kelulusan JE) needs to be prepared at this stage and this need to be done
one week from the date of the meeting of JP.
Step 7:
An offer letter will be issued to the successful candidate and the unsuccessful
candidates will be notified one week from the date of approval.
50
Step 8:
The selection of an academic staff will be ended with the acceptation of staff
reporting under the responsibility of Registrar’s Office. An offer will be made to the
next appointable candidate if the preferred candidate does not report within the
period which is agreed by both of the University and the preferred candidate.
The summary of the intake and appointment of academic staff in UTM is
shown in the flow chart in Figure 4.3.
4.3.2
Selection Process at Faculty Level (Faculty of Science)
Academic staff in a faculty serves a vital role in the university and plays an
important part in the quality of its teaching. Therefore, the faculty is committed to
ensuring that it employs the most suitable qualified and experienced people.
The selection process of an academic staff in the faculty is guided by a clear
idea of job to be done and the relevant skills, knowledge, experience and
qualifications required.
The academic staff is normally drawn from a pool of
applicants, which obtained through a variety of recruitment options, including
advertisement and unsolicited applications.
The selection process at faculty level starts with the identification of a job
vacancy in the faculty by Department Heads from each of the Department assisted by
each of the Heads of each relevant field in the Departments. Department Head will
inform the Registrar’s Office through the Assistant Registrar in the faculty regarding
51
^ƚĂƌƚ
ĐĐĞƉƚĐĞƌƚŝĨŝĐĂƚŝŽŶŽĨ
ĂƉƉŽŝŶƚŵĞŶƚĨƌŽŵĨĂĐƵůƚLJ
ZĞƐƉŽŶƐŝďŝůŝƚLJŽĨ͗
dW;WͿ͕WW;,ZDͿϭ͕
Wd;WͬKͿ
Not
complete
ŚĞĐŬƚŚĞŝŶĨŽƌŵĂƚŝŽŶ
ŽĨĐĞƌƚŝĨŝĐĂƚŝŽŶ
ŽŵƉůĞƚĞ
/ƐƐƵĞĐĂůůůĞƚƚĞƌĨŽƌŝŶƚĞƌǀŝĞǁ
ƐĞƐƐŝŽŶ
^ƚĂŶĚĂƌĚ͗ƚůĞĂƐƚϱ
ǁŽƌŬŝŶŐĚĂLJƐ
,ŽůĚĂ:WŵĞĞƚŝŶŐĨŽƌŝŶƚĞƌǀŝĞǁ
ƉƵƌƉŽƐĞ
^ƚĂŶĚĂƌĚ͗dǁŽ
ǁĞĞŬƐďĞĨŽƌĞƚŚĞĚĂƚĞ
ŽĨŵĞĞƚŝŶŐ
WƌĞƉĂƌĞƚŚĞ:ĂƉƉƌŽǀĂů͚<ĞƌƚĂƐ
ŬĞƌũĂ͚
^ƚĂŶĚĂƌĚ͗KŶĞǁĞĞŬ
ĨƌŽŵƚŚĞĚĂƚĞŽĨ:W
ŵĞĞƚŝŶŐ
/ƐƐƵĞŽĨĨĞƌͬŶŽƚŝĨŝĐĂƚŝŽŶůĞƚƚĞƌƚŽ
ƐƵĐĐĞƐƐĨƵůͬƵŶƐƵĐĐĞƐƐĨƵůĐĂŶĚŝĚĂƚĞƐ
^ƚĂŶĚĂƌĚ͗KŶĞǁĞĞŬ
ĨƌŽŵƚŚĞĂƉƉƌŽǀĂůĚĂƚĞ
EŽƚĞƐ͗
/Ĩ ƚŚĞ ĨŝƌƐƚ
ƌĂŶŬĞĚ
ĐĂŶĚŝĚĂƚĞ
ĚŽĞƐ
ŶŽƚ
ƌĞƉŽƌƚ ǁŝƚŚŝŶ
ƚŚĞ
ƉĞƌŝŽĚ
ĂŐƌĞĞĚ
ďĞƚǁĞĞŶ ƚŚĞ
ĐĂŶĚŝĚĂƚĞ ĂŶĚ
ƚŚĞ ƵŶŝǀĞƌƐŝƚLJ͕
ĂŶ ŽĨĨĞƌ ůĞƚƚĞƌ
ǁŝůů ďĞ ŝƐƐƵĞĚ
ƚŽ ƚŚĞ ŶĞdžƚ
ƌĂŶŬĞĚ
ĂƉƉůŝĐĂŶƚ͘
ĐĐĞƉƚƚŚĞƌĞƉŽƌƚŽĨŶĞǁůLJ
ƌĞĐƌƵŝƚĞĚƐƚĂĨĨ
ZĞƐƉŽŶƐŝďŝůŝƚLJŽĨ
ZĞŐŝƐƚƌĂƌ͛ƐKĨĨŝĐĞ
ŶĚ
Figure 4.3.
4.3.2
Intake and Appointment of Academic Staff flow chart in UTM
Selection Process at Faculty Level (Faculty of Science)
52
the job vacancy in the faculty. After the notification of the job vacancy in the faculty
is done, Registrar’s Office will need to get the designation warrant issued by UTM
Bursary regarding the job vacancy. The Registrar’s Office can only proceed to the
recruitment process if there is allocation for the permanent post of academic staff for
the appropriate year from Jabatan Perkhidmatan Awam (JPA).
After the issue of designation warrant, Registrar’s Office will advertise the
job vacancy in the media chosen which is done once a year. Application forms sent
by applicants to Registrar’s Office will be sent to faculty for further action. The
Assistant Registrar in the faculty is responsible for the process of filtering or shortlisting the applicants before sending the complete application forms to the
Department Heads of the related fields.
Short-listing is done based only on the information contained in the
application form related to the job description and the person specification, together
with any other information supplied by the candidate with their application form.
The criteria in the person specification are consistently applied to all candidates.
Short-listing is done as an approach to reduce the number of candidates progressing
to the interview session. An interview record form is completed independently by
each interviewer to record their assessment of each individual interview immediately
after each interview.
Call letter for interview will be issued by the Assistant Registrar in the faculty
at the faculty level.
applicant.
An interview is held to obtain the ‘further details’ of an
Some information can only be determined at the interview stage.
Candidates who do not meet an essential attribute which is being assessed at the
short-listing stage will not be invited to attend for interview.
53
The Selection Committee in the faculty is responsible for the process of
conducting interview session and selecting an academic staff at faculty level. The
first interview of an applicant will be held by the Selection Committee at the faculty
level which consists of Dean, Deputy Dean (Academic), Deputy Dean
(Development), Assistant Registrar in the faculty, Head of related Department and
two professors or associate professors in the related field. Each of the Selection
Committee members will have to follow the guideline given by Registrar’s Office in
evaluating candidates in the selection process.
The selected academic staff which is done in the faculty will be certified by
the Dean of the Faculty.
Personal information, academic qualification, working
experience and suggested salary with the faculty certification by Dean of Faculty of a
candidate which is stated clearly in the Certificate Schedule of the Appointment of an
academic staff (Jadual Perakuan Perlantikan Staf Akademik) will be forwarded to the
Selection Committee in the Registrar’s Office. All of the documentation forwarded
will be reviewed by the Selection Committee in order to check the short-listing
decision is justifiable in respect of the essential and desirable criteria stipulated in the
person specification.
The rest of the procedure in the selection process is conducted by the
Registrar’s Office in order to select the appropriate candidates to be the faculty
member in the university as explained in Section 4.3.1. Executive Committee (JE) is
responsible in making the final decision on recruiting an academic staff for the
university.
The overall academic staff selection process in UTM is summarized as
shown in Figure 4.4.
54
ZĞŐŝƐƚƌĂƌ͛ƐKĨĨŝĐĞ
^ƚĂƌƚ
&ĂĐƵůƚLJŽĨ^ĐŝĞŶĐĞ
ĐĐĞƉƚĐĞƌƚŝĨŝĐĂƚŝŽŶŽĨ
ĂƉƉŽŝŶƚŵĞŶƚĨƌŽŵĨĂĐƵůƚLJ
^ƵďŵŝƚƌĞůĂƚĞĚ
ĚŽĐƵŵĞŶƚƐƚŽ
ZĞŐŝƐƚƌĂƌ͛ƐKĨĨŝĐĞ
/ŶƚĞƌǀŝĞǁ
EŽ
ŚĞĐŬƚŚĞŝŶĨŽƌŵĂƚŝŽŶ
ŽĨĐĞƌƚŝĨŝĐĂƚŝŽŶ
/ƐƐƵĞĐĂůůůĞƚƚĞƌ
ĨŽƌŝŶƚĞƌǀŝĞǁ
zĞƐ
/ƐƐƵĞĐĂůůůĞƚƚĞƌĨŽƌŝŶƚĞƌǀŝĞǁ
ƐĞƐƐŝŽŶ
^ĞŶĚ
ƉƉůŝĐĂƚŝŽŶ
ĨŽƌŵƐƐĞŶƚďLJ
ĂƉƉůŝĐĂŶƚƐƚŽ
ĨĂĐƵůƚLJ
ĚǀĞƌƚŝƐĞŵĞŶƚ
^ŚŽƌƚͲůŝƐƚŝŶŐ
,ŽůĚĂ:WŵĞĞƚŝŶŐĨŽƌŝŶƚĞƌǀŝĞǁ
ƉƵƌƉŽƐĞ
/ŶĨŽƌŵƐƐŝƐƚĂŶƚ
ZĞŐŝƐƚƌĂƌŝŶƚŚĞ
ĨĂĐƵůƚLJ
WƌĞƉĂƌĞǁŽƌŬŝŶŐƉĂƉĞƌ
ĂƉƉƌŽǀĂůŽĨ:
/ĚĞŶƚŝĨŝĐĂƚŝŽŶŽĨ
ũŽďǀĂĐĂŶĐLJŝŶƚŚĞ
ĨĂĐƵůƚLJ
/ƐƐƵĞŽĨĨĞƌͬŶŽƚŝĨŝĐĂƚŝŽŶůĞƚƚĞƌƚŽ
ƐƵĐĐĞƐƐĨƵůͬƵŶƐƵĐĐĞƐƐĨƵůĐĂŶĚŝĚĂƚĞƐ
ĐĐĞƉƚƚŚĞƌĞƉŽƌƚŽĨŶĞǁůLJ
ƌĞĐƌƵŝƚĞĚƐƚĂĨĨ
ŶĚ
Figure 4.4.
The overall Process of Academic Staff Selection in Faculty.
55
4.3.3
Selection Criteria
The university wishes to ensure that it attracts the most suitably qualified and
experienced person for its posts.
Selection criteria will help to select the best
candidate when more than one candidate meets all the essential criteria. Selection
criteria are those which are not key factors to carrying out the duties of the post but
which will give added value if candidates possess some or all of them. It should be
closely reflect the content of the position description and departmental needs.
To determine the selection criteria which are needed as a guideline to recruit
an academic staff in a faculty, there is a need to examine experience, education, and
intelligence and personality requirements.
Faculty of Science is following the guidelines and forms given by the Registrar’s
Office in the process of academic staff selection. The Registrar’s Office has the
guidelines for the criteria used in order to select an appropriate candidate. The
criteria used including:
(a) Knowledge (Professional knowledge, education background)
The quality of academic staff is the most important component in assuring the
quality, successful and excellent of the university.
Therefore, the
professional knowledge in the related field and the education background are
parts of the important criteria to look at in the selection process.
This
criterion relates to the level and breadth of knowledge required to do the job.
It is essential for the academic staff to have the required academic
qualification for the discipline they are teaching in, and to also have expertise
in one or more subdivisions or specialties within that discipline, as well as
research capabilities. It is vital that the teachers contribute to the
advancement of knowledge and to the intellectual growth of the students
56
through the scholarly activity of research and continuing education. Persons
appointed to academic positions must have demonstrated achievement within
their disciplines commensurate with their faculty rank.
(b) Working experience
Experienced academic staff can play an important role in the achievement of
a university. They must have the capability and continued commitment to be
effective teachers. Effective teaching requires knowledge of the discipline, an
understanding of pedagogy, methods of measuring student performance
consistent with the learning objective, and readiness to be subjected to
internal and external evaluation.
Experience is the accumulation of understanding gained from formal
education or through past job experience that the applicant would require to
meet the prerequisites of the position. Some knowledge can be gained on the
job, therefore, it is important to assess on how much and what types of
knowledge are required prior to a person starting in a role. This will assist in
avoiding overstating selection capabilities.
Thus, to know the working
experience of an applicant will be advantages.
(c) General knowledge
Another quality needed is general knowledge. The nature of this quality is
considered and put as one of the academic staff selection criteria.
(d) Personality
Personality is a set of qualities that make a person distinct from another. It is
made up the characteristic patterns of thoughts, feelings, and behaviors that
make a person unique. The sub-criteria which are taken into consideration in
the academic staff selection in the faculty are the manner, conversation style
politeness and appearance (tidiness) of the applicant.
57
(e) Intelligence / eloquence (respond to interviewer’s question)
This criteria is to see how intelligence or eloquence the applicant in order to
response to the interviewers’ questions during the interview session.
(f) Extracurricular activities in school / university (activities other than academic
in school / university)
(g) Presentation in English
For the effective teaching and learning process in the university, it is
important that the ability and skill of an applicant to communicate and give
lecture in English is being evaluated in the academic staff selection process
since the successful candidate will have to relate with students from a variety
of backgrounds and conduct tutorials, lectures in English. The ability and
effective of giving lecture in English and oral communication skills will be
assessed during the presentation.
(h) Others (not included in the marking rubric, e.g. Nationality, Gender, Age, etc.,
in application forms)
Although there are other key factors which is not listed in the marking rubric,
but sometime these factors are the key factors affecting the selection of an
applicant.
The evaluation of an applicant during the interview at faculty level is based
on criteria stated in the marking rubric. Applicants will be graded by using a rating
system. The rating system is based on the marks scored for each of the criteria
evaluated during the interview session. There are seven main criteria stated in the
marking rubric with the score boxes beside it for each of the criterion (Refer to
Appendix A).
Candidates will be given score which is in the rank from the minimum score
of 1 to the maximum of score 10 for each of the criterion during the interview
58
session. There is no specific rating for each of the sub-criteria stated. Candidate
who mostly fully meets all the criteria required and passes the interview and
presentation will be selected as the faculty academic staff by the Selection
Committee of Faculty of Science. Distribution of marks for the criteria evaluated is
as shown in Table 4.1.
Table 4.1.
No
Marks distribution in the marking rubric.
Criterion
Calculation of marks
Distribution of
marks
1.
Knowledge
a/10 x 25
25
2.
Working experience
b/10 x 25
25
3.
General knowledge
c/10 x 20
20
4.
Personality
d/10 x 10
10
5.
Intelligence / eloquence
e/10 x 10
10
6.
External activities
f/10 x 10
10
7.
Result
(a+b+c+d+e+f)/60 x 100
100
of
lecturing
in
English
Notes: a, b, c, d, e, f are the scores obtained by the applicant during the interview.
4.4
Conclusion
Although Faculty of Science in UTM has a series of systematic procedure in
the process of selecting its academic staff, but other than the requirement for a
marking rubric that included the criteria to be applied in assessing candidates in the
initial review, there is no direction concerning on how to elicit these criteria, how to
differentiate their importance and how to perform the actual assessments. This
59
procedure listed a sequence of activities but it does not convey a process to identify
the best candidate from among many applicants.
The existing guidelines and
procedures for the selection committees prescribed only ‘what to do’ basis but not
providing any guidance or procedures on the ‘how’ basis.
Therefore, as mentioned by Grandzol (2005), the selection committee will
leave to their own experiences, past practices, personal preferences, or any other set
of events that may emerge from the committee. Consistency may be non-existent as
search committees are frequently ad hoc in their nature, established on a vacancy-byvacancy basis.
Hence, there is a need to apply the sequential steps based on AHP principles
supported decision making tool in the academic staff selection in the faculty as well
as the university.
60
CHAPTER 5
THE AHP ACADEMIC STAFF SELECTION MODEL
5.1
Introduction
This chapter describes the development of the theoretical model for the
selection of appropriate academic staff in Faculty of Science using AHP aided
decision making tool. The process involved in development this model as well as the
mathematical aspect of the process are being discussed in detail in this chapter. The
key factors (selection criteria) of the model are highlighted.
Microsoft Excel spreadsheet software is used in the calculation of geometric
means from experts’ judgments and Expert Choice 11.5 software is being used to
simplify the synthesizing of the priorities for the academic staff selection criteria and
sub-criteria used in this study.
61
5.2
Modeling
Mathematical modeling is a process of developing a mathematical model. It
is the representation of the essential aspects of a system which presents knowledge of
that system in usable form. It is contended that the mathematical modeling deserves
attention as it serves science from the point of view that it is the primary testing and
development ground for the power of mathematical language as applied to real life
problems.
To develop a mathematical model in the selection of the most appropriate
academic staff in the Faculty of Science, there is a need to weigh the alternative
selection method with a list of appropriate selection criteria. To solve this multicriteria decision making problem, an Analytical Hierarchy Process (AHP) aided
decision making tool is adopted.
5.3
The AHP Model for Academic Staff Selection
Essentially, in the selection of an academic staff, the faculty is clear about the
qualities, skills, competencies and the knowledge needed for the vacant position. But
it is not articulate explicitly on how the selection committee can make the decision to
select an appropriate candidate to be the faculty member. Gibney and Shang (2007)
in their paper on Decision Making in Academia believed that the personnel selection
process can be aided by some formal decision making techniques, particularly the
AHP with its emphasis on decision making with intangible criteria.
62
5.3.1 Development of AHP hierarchy
AHP serves as a mathematical solution method for individual or group
decision making with multiple criteria. It is a decision making tool which provides a
proven, effective means to deal with complex decision making which is suitable for
complex decisions making involving the comparison of decision elements (criteria)
which are difficult to quantify in a multiple criteria problem. AHP assists with
identifying and weighting selection criteria, analyzing the data collected for the
criteria and expediting the decision-making process. It is based on the assumption
that when faced with a complex decision the natural human reaction is to cluster the
decision elements according to their common characteristics.
AHP structures the problem as a hierarchy by building a hierarchy of criteria
elements and making comparisons between each possible pair in each level as a
matrix. By making pair-wise comparison at each level of the hierarchy, the relative
weights or priorities can be developed to differentiate the importance of the criteria.
This gives a weighting for each element within a level and also a consistency ratio
which is useful for checking the consistency of the data. This is a method to derive
ratio scales from paired comparisons.
AHP hierarchy is a representation of a complex problem on a number of
levels. Minimal number of hierarchy levels for multiple attribute decision makings is
three.
The first level represents the goal of problem solving, the second level
represents the criteria and the third level represents alternatives (Brozova, 2004). It
is important to include the expert’s opinion and obtain consensus until the problem is
clearly defined in constructing the hierarchy tree. For this reason the criteria and the
alternatives obtained from the interview with the experts is used to construct the
hierarchy.
63
AHP consists of a simple hierarchy process that helps the academic staff
selection committee to evaluate candidates by using a set of chosen criteria. The
steps implemented in the development of AHP model as mentioned in Chapter 3
includes:
1. Development of a conceptual framework.
2. Setting of the decision hierarchy
3. Obtaining information from experts
4. Employing the pair-wise comparison
5. Estimating relative weights of elements
6. Calculating the degree of consistency
7. Computing the entire hierarchy weight
5.3.2 Development of a conceptual framework
Under the development of a conceptual framework, the problem of study is
defined and determined. Selection of the most appropriate academic staff has been
identified as the problem and goal in this study. The complexity of the problem in
this study which is the selection of appropriate academic staff decomposed into
different levels or components and the relations of the components are synthesized in
the next section.
5.3.3 Decision Hierarchy
It is crucial to formulate the preference structure which is referring to
deciding on the selection criteria most important to the faculty for the vacant position.
64
The selection committee members should take into account all constituents, rather
than focus exclusively on their own perceptions related to ‘best fit’ definition of a
colleague (Grandzol, 2005). This is the key factor to the whole selection process.
The misspecification of the key criteria that are critical to identifying not only
suitable, but best candidates will invalidate all the succeeding activities.
The selection of the most appropriate academic staff in the Faculty of Science,
which is the goal of the decision makers, is placed as level 0 of the model to serve as
goal node. Factors affecting the selection of candidates, which is classified into 5
main categories, is placed in level 1 of the model to serve as the main criteria. Level
2 consists of the sub-criteria of the model for the main criteria in level 1. Finally the
alternative solution (most appropriate candidate) is located at level 3 to serve as the
choice available for the decision makers.
5.3.3.1 Selection Criteria
There are 5 main criteria identified in this study to be included in the AHP
model developed, compared to the seven criteria used by UTM academic staff
selection committee since some of the criteria can be grouped under one criterion.
Some of the relevant criteria which are not mentioned in the UTM selection criteria
are added in the AHP aided decision making model. The criteria chosen in the AHP
model are the knowledge, working experience or interpersonal skill, general traits,
scholar or extracurricular activities and references. However, the criteria used can be
easily changed or modified in the model depending on the requirement of the faculty
in the process of selecting an appropriate academic staff.
Professional knowledge, education background and general knowledge are
categorized as the sub-criteria for knowledge. This is related to the level and breadth
of knowledge required for a successful candidate to carry out the duty. Meanwhile
the general knowledge is the wide body of information that a person acquires from
65
education and from life. This is important to show a candidate’s sensitivity to the
happening of the surrounding.
The criterion of working experience or interpersonal skill is further expanded
to include the sub-criteria of working duration, working field, teaching experience
and skill, independence, intelligence or eloquence and the ability to communicate in
English.
Working experience is the proven record of experience and achievement in a
field, profession or specialism. Employment history with dates is important in order
to determine an applicant's attitude towards their work.
Distinguishing between types of experiences is necessary. A person with
good qualifications and little experience may not be as efficient as a person with a
large amount of relevant experience but slightly less advantage in qualifications.
Although specific years of experience, in many cases, are no longer a requirement
included in a job specifications, but there is still the possibility that an applicant's age
and years of experience from the dates of employment histories and of qualifications
are looked into, which may influence the selection decisions.
Therefore, it is
important for an applicant to state the precise nature of the experiences, outline the
various elements of experience, and the specific skills when applying for the vacant
post.
Eloquence shows a candidate’s quality of persuasive and the power of
expression.
Whereas, intelligence can indicate a person’s adaptability to a new
environment, ability to comprehend relationships, ability to evaluate and judge and
the capacity for original
DQG productive
thought. Since the teaching and learning
process in university is conducted in English, the ability to communicate in English
both orally and in written context is a must to be possessed by an applicant.
66
The criterion of general traits is broken down into the sub-criteria of manner
or politeness, appearance, age and professional interest. A candidate’s traits will
ferret out the ability to perform required job responsibilities. Personality will
distinguish the qualities or characteristics of a candidate.
The forth criteria used for the evaluation of a candidate is the scholarly or
extracurricular activities that a candidate was involving. Although knowledge and
academic achievements are the prime criteria in the selection of an academic staff in a
faculty but the combination of the academic pressures experienced with the
involvement in extracurricular activities will illustrate a candidate’s interests and how
successfully a candidate can handle the pressures resolved in the past. It is essential
for an academic staff to have the required academic qualification for the discipline
they are teaching in, and to also have expertise in one or more subdivisions or
specialties within that discipline, as well as research capabilities.
Meanwhile, it is vital that the teachers contribute to the advancement of
knowledge and to the intellectual growth of the students through the scholarly activity
of research and continuing education. Therefore, number of publications, researches
done and rewards recognized and experience of a candidate with diverse population
are put under the criteria of scholar or extracurricular activities that a candidate was
involving.
Reference is another criterion used as the selection criteria of an academic
staff in this study. In the west, reference checking is an essential and important step
in the overall recruitment process. References will be contacted to substantiate and
validate the information received during the interview process and to provide further
insight into the candidate’s abilities, skills and knowledge as it relates to the position
applied for and to the area of expertise required.
67
Knowledge acquisition from experts in the faculty will help in synthesizing
the weight for each pair of the criteria or sub-criteria compared. Figure 5.1 shows
the hierarchical structure for the academic staff selection constructed using the
criteria and sub-criteria as discussed.
68
5.3.3.2 Hierarchical Structure
>ĞǀĞůϬ͗
'ŽĂů
>ĞǀĞůϮ͗
^ƵďͲĐƌŝƚĞƌŝĂ
>ĞǀĞůϭ͗
ƌŝƚĞƌŝĂ
>ĞǀĞůϯ͗
ůƚĞƌŶĂƚŝǀĞƐ
WƌŽĨĞƐƐŝŽŶĂů<ŶŽǁůĞĚŐĞ͕W&^
<ŶŽǁůĞĚŐĞ͕
<Et
ĚƵĐĂƚŝŽŶĂĐŬŐƌŽƵŶĚ͕h
'ĞŶĞƌĂů<ŶŽǁůĞĚŐĞ͕'E>
tŽƌŬŝŶŐƵƌĂƚŝŽŶ͕WZ
^ĞůĞĐƚŝŽŶŽĨƚŚĞŵŽƐƚĂƉƉƌŽƉƌŝĂƚĞĂĐĂĚĞŵŝĐƐƚĂĨĨ
tŽƌŬŝŶŐ&ŝĞůĚ͕&>
tŽƌŬŝŶŐ
ĞdžƉĞƌŝĞŶĐĞͬ
ŝŶƚĞƌƉĞƌƐŽŶĂů
ƐŬŝůů͕/d
dĞĂĐŚŝŶŐdžƉĞƌŝĞŶĐĞͬ^Ŭŝůů͕d,
/ŶĚĞƉĞŶĚĞŶĐĞ͕/W
/ŶƚĞůůŝŐĞŶĐĞͬĞůŽƋƵĞŶĐĞ͕/d>
ďŝůŝƚLJƚŽĐŽŵŵƵŶŝĐĂƚĞŝŶ
ŶŐůŝƐŚ͕>
DĂŶŶĞƌͬƉŽůŝƚĞŶĞƐƐ͕DEZ
'ĞŶĞƌĂů
ƚƌĂŝƚƐ͕
'Ed
ĂŶĚŝĚĂƚĞ
Ϯ
ĂŶĚŝĚĂƚĞ
ϯ
ƉƉĞĂƌĂŶĐĞ͕WZ
ŐĞ͕'
WƌŽĨĞƐƐŝŽŶĂů/ŶƚĞƌĞƐƚ͕W/d
WƵďůŝĐĂƚŝŽŶ͕W>
^ĐŚŽůĂƌͬ
džƚƌĂĐƵƌƌŝͲ
ĐƵůĂƌ
ĂĐƚŝǀŝƚŝĞƐ͕
d
ĂŶĚŝĚĂƚĞ
ϭ
ZĞƐĞĂƌĐŚĞƐ͕Z^,
ĂŶĚŝĚĂƚĞ
ϰ
ĂŶĚŝĚĂƚĞ
ϱ
ZĞǁĂƌĚƐ͕Zt
džƉĞƌŝĞŶĐĞǁŝƚŚĚŝǀĞƌƐĞ
ƉŽƉƵůĂƚŝŽŶ͕WW>
ZĞĨĞƌĞŶĐĞƐ͕
Z&
Figure 5.1. Hierarchical Structure for the Selection of Academic Staff
69
5.3.3.3 Importance of the criteria
After formulating the preference structure, the next step in the development
of AHP model is to differentiate the relative importance for each of the criteria. This
is done during the process of interviewing experts by using a pair-wise comparison
table as shown in Table 5.1 for each of the criteria and sub-criteria used in this study.
The subjective judgment of experts is illustrated using 1 to 9 scale as suggested by
Saaty and Kearns (1985).
There are different tables at each level of the hierarchy which grouped subcriteria for comparison within criteria one level above. For example, Table 5.1 lists
all the possible pair-wise comparison for the five level 1 criteria. There are another 4
separate tables listing the 3 sub-criteria of the knowledge criterion, 6 criteria of the
working experience / interpersonal skill criterion, 4 sub-criteria of general traits
criterion and 4 sub-criteria of scholar / extracurricular activities. In other hand,
during an interview, the expert will be required to complete 5 tables similar to the
one shown in Table 5.1.
Table 5.1. Sample Pair-wise Comparison Table.
ϭсYh>
ϭ <Et ϵ
Ϯ <Et ϵ
ϯ <Et ϵ
ϰ <Et ϵ
ϱ /d
ϵ
ϲ /d
ϵ
ϳ /d
ϵ
ϴ 'Ed ϵ
ϵ 'Ed ϵ
ϭϬ d
ϵ
ďďƌĞǀŝĂƚŝŽŶ
<Et
/d
'Ed
d
Z&
ŽŵƉĂƌĞƚŚĞƌĞůĂƚŝǀĞŝŵƉŽƌƚĂŶĐĞǁŝƚŚƌĞƐƉĞĐƚƚŽ'K>͗
^ĞůĞĐƚƚŚĞDŽƐƚƉƉƌŽƉƌŝĂƚĞĐĂĚĞŵŝĐ^ƚĂĨĨ
ϯсDKZd
ϱс^dZKE'
ϳсsZz^dZKE'
ϴ ϳ ϲ ϱ ϰ ϯ Ϯ ϭ Ϯ ϯ ϰ ϱ ϲ
ϴ ϳ ϲ ϱ ϰ ϯ Ϯ ϭ Ϯ ϯ ϰ ϱ ϲ
ϴ ϳ ϲ ϱ ϰ ϯ Ϯ ϭ Ϯ ϯ ϰ ϱ ϲ
ϴ ϳ ϲ ϱ ϰ ϯ Ϯ ϭ Ϯ ϯ ϰ ϱ ϲ
ϴ ϳ ϲ ϱ ϰ ϯ Ϯ ϭ Ϯ ϯ ϰ ϱ ϲ
ϴ ϳ ϲ ϱ ϰ ϯ Ϯ ϭ Ϯ ϯ ϰ ϱ ϲ
ϴ ϳ ϲ ϱ ϰ ϯ Ϯ ϭ Ϯ ϯ ϰ ϱ ϲ
ϴ ϳ ϲ ϱ ϰ ϯ Ϯ ϭ Ϯ ϯ ϰ ϱ ϲ
ϴ ϳ ϲ ϱ ϰ ϯ Ϯ ϭ Ϯ ϯ ϰ ϱ ϲ
ϴ ϳ ϲ ϱ ϰ ϯ Ϯ ϭ Ϯ ϯ ϰ ϱ ϲ
ĞĨŝŶŝƚŝŽŶ
<ŶŽǁůĞĚŐĞ
tŽƌŬŝŶŐĞdžƉĞƌŝĞŶĐĞͬŝŶƚĞƌƉĞƌƐŽŶĂůƐŬŝůů
'ĞŶĞƌĂůƚƌĂŝƚƐ
^ĐŚŽůĂƌͬĞdžƚƌĂĐƵƌƌŝĐƵůĂƌĂĐƚŝǀŝƚŝĞƐ
ZĞĨĞƌĞŶĐĞƐ
ϵсydZD
ϳ ϴ ϵ /d
ϳ ϴ ϵ 'Ed
ϳ ϴ ϵ d
ϳ ϴ ϵ Z&
ϳ ϴ ϵ 'Ed
ϳ ϴ ϵ d
ϳ ϴ ϵ Z&
ϳ ϴ ϵ d
ϳ ϴ ϵ Z&
ϳ ϴ ϵ Z&
70
The tables are drawn by referring to the questionnaire generated by using the
Expert Choice 11.5 which facilitates the application of AHP by allowing direct
comparison using the 1 to 9 scale as recommended by Saaty and Kearns (1985). The
number to the left of the center indicating a preference for the criterion listed on the
left and the opposite being true for the scale on the right.
5.3.3.4 Discordance levels
The setting of discordance levels needs to be done by the selection committee
in order to establish a minimum acceptable score for each of the criteria. Any
applicant not achieving this minimum level as set out in the specification will not be
considered to proceed to the rating process. This short listing process is done using
the level of discordance set by the selection committee. By setting this discordance
level, the number of potential candidates can be reduced in terms of efficient use of
time and effort.
5.3.3.5 Rating Scales
The process of creating a rating scale for the measurement or assessment of
each candidate needs to be done rigorously by the selection committee. Without the
careful definition of ratings scale, this may lead to unintended implicit weighting.
The AHP requires that criteria be specified a priori, and that all decision makers
agree on the definitions of the criteria and sub-criteria and sign off on the
completeness of the model. To foster agreement and a willingness to be fully behind
any group decision, any process using AHP should begin with a discussion among all
the decision makers to develop a working definition of the criteria (Gibney and
Shang, 2007).
71
The committee may select a single verbal scale covering all criteria, as
suggested by Grandol (2005), which consists of the options: excellent, very good,
good, fair and poor, depending on the criterion evaluated. Most of the rating scale
derived based on frequencies or time intervals. For example, the selection committee
can use number of months or years to determine the rating scale for working
experience and number of publications published for the sub-criterion of Publications.
5.3.4
Information from experts
One of the advantages of AHP is the qualitative information gathered from
experts from the process of knowledge acquisition can be transformed into
quantitative information or data to be used as input in AHP model. The data or
information is obtained directly from the experts through interviews.
The
quantitative data transformed is then used to find the priorities of each of the criteria
and its sub-criteria which are needed for the ranking of the most appropriate
academic staff in this study.
There are four methods to set the priorities for each of the criteria and subcriteria:
consensus, vote or compromise, geometric mean of the individual’s
judgments and separate models or players (Dyer and Forman, 1992). The basis for
using the geometric rather than the arithmetic mean to combine judgments of
different individuals has been justified mathematically by Saaty (1980). Since the
data is obtained from three experts in this study, it is calculated by using the
geometric mean from the three individual expert’s responses to obtain the consensus
judgments in this study.
Microsoft Excel 2007 is automated to calculate the geometric mean before
transferring the processed data into Expert Choice 11.5 to obtain the criteria and sub-
72
criteria weights (priorities). For example, the judgment for relative importance of
criteria 1 to criteria 2 is given x1, x2, x3, ..., xn by n experts respectively, the geometric
mean can be calculated by using the formula:
భ
౤
Geometric mean = ඥ‫ݔ‬ଵ ൈ ‫ݔ‬ଶ ൈ ‫ݔ‬ଷ ൈǤ Ǥ Ǥൈ ‫ݔ‬௡ As an example, if 3 committee members regard the relative importance of
Knowledge over References as 9, 8 and 7, respectively, then the aggregate
భ
య
importance of Knowledge will be ξͻ ൈ ͺ ൈ ͹ = 7.958, and this will be used as the
judgment. Using the geometric mean can be justified by noting that the reciprocal of
the group value 7.958 is identical with the value that will be obtained by taking the
geometric mean of the three individual reciprocal values.
5.3.5
Employment of pair-wise comparison
In the AHP model, criteria and sub-criteria of the problem are compared in
pairs with respect to the relative weight on a property they share in common. Pairwise comparison of each of the pair of the criterion and sub-criterion in this study is
reduced into a square matrix in which the array of numbers is arranged as shown in
Table 5.2, Table 5.3, Table 5.4, Table 5.5 and Table 5.6 in the next subsection.
5.3.5.1 Pair-wise comparison matrix
The pair-wise comparison matrices for each of the criteria and sub-criteria are
formed by using the processed data obtained from the calculation with geometric
mean as discussed earlier. Table 5.2, Table 5.3, Table 5.4, Table 5.5 and Table 5.6
73
are the matrices formed in this process. In the matrices formed, the elements in the
lower triangle in the matrices are the relative value for the reciprocal values of the
upper triangular, that is, aji =
ଵ
ୟ౟ౠ
.
Table 5.2. Pair-wise comparison matrix for the five selection criteria.
Criteria
KNW
EIT
GNT
ACT
RFC
KNW
1.000
7.319
2.884
2.000
7.958
EIT
0.137
1.000
1.260
1.260
6.649
GNT
0.347
0.794
1.000
1.442
7.652
ɉ୫ୟ୶
= 5.434
ACT
0.500
0.794
0.693
1.000
6.649
CI
= 0.109
RFC
0.126
0.150
0.131
0.150
1.000
CR
= 0.097
Table 5.3. Pair-wise comparison matrix for the criterion of Knowledge, KNW.
Sub-criterion of KNW
PFS
EDU
GNL
PFS
1.000
2.08
7.958
EDU
0.481
1.000
7.958
GNL
0.126
0.126
1.000
ɉ୫ୟ୶ = 3.062
CI = 0.031
CR = 0.053
74
Table 5.4.
Pair-wise comparison matrix for the criterion of Working
experience/ interpersonal skill, EIT.
Sub-criteria of EIT
PRD
FLD
TCH
IDP
ITL
ABL
PRD
1.000
0.158
1.000
4.160
0.146
0.275
FLD
6.316
1.000
1.000
7.000
1.260
0.550
TCH
1.000
1.000
1.000
3.979
1.000
1.000
IDP
0.240
0.143
0.251
1.000
0.131
0.137
ITL
6.840
0.794
1.000
7.652
1.000
1.000
ABL
3.634
1.817
1.000
7.319
1.000
1.000
ɉ୫ୟ୶ = 6.505
CI = 0.101
CR = 0.081
Table 5.5. Pair-wise comparison matrix for the sub-criterion of General Traits,
GNT.
Sub-criteria of GNT
MNR
APR
AGE
PIT
MNR
1.000
1.000
8.000
2.080
APR
1.000
1.000
7.958
3.979
AGE
0.125
0.126
1.000
0.131
PIT
0.481
0.251
7.652
1.000
ɉ୫ୟ୶ = 4.174
CI = 0.058
CR = 0.064
75
Table 5.6. Pair-wise comparison matrix for the sub-criterion of Scholar /
Extracurricular activities, ACT.
Sub-criteria of ACT
PBL
RSH
RWD
PPL
PBL
1.000
1.000
8.320
8.320
RSH
1.000
1.000
3.979
7.958
RWD
0.120
0.251
1.000
1.817
PPL
0.120
0.126
0.550
1.000
ɉ୫ୟ୶ = 4.057
CI = 0.019
CR = 0.021
The figures obtained through calculation with the geometric mean by using
Microsoft Excel 2007 are then transferred to the data grid in Expert Choice 11.5 as
shown in Figure 5.2.
Figure 5.2. Data grid for the input of criteria weights.
After entering the figures from the Microsoft Excel 2007 into the data grid,
the priorities of each of the criteria, as well as its sub-criteria, can be calculated
automatically using the Expert Choice 11.5.
Figure 5.3 to Figure 5.7 show the screen shots captured from the Expert
Choice 11.5 detailing the priorities of the criteria and sub-criteria in this study.
76
Figure 5.3. Priorities of criteria level 1.
5.3.5.2 Computation of Eigenvalue (ૃ max), Consistency Index (CI), Consistency
Ratio (CR) and weights or priorities of the criteria or sub-criteria.
By using the matrix of the sub-criterion of Knowledge in Table 5.3 as an
example, the computation of the ɉ୫ୟ୶ , CI and CR and weights or priorities of the
criteria or sub-criteria are shown as below.
Let the matrix formulated in Table 5.3 be Matrix A.
Therefore,
Matrix A =
(i)
ͳǤͲͲͲ
ͲǤͶͺͳ
ͲǤͳʹ͸
ʹǤͲͺͲ
ͳǤͲͲͲ
ͲǤͳʹ͸
͹Ǥͻͷͺ
͹Ǥͻͷͺ
ͳǤͲͲͲ
Eigenvalue, ૃ max
ɉI–A
ͳ Ͳ
= ɉͲ ͳ
Ͳ Ͳ
Ͳ
ͳǤͲͲͲ
Ͳെ ͲǤͶͺͳ
ͳ
ͲǤͳʹ͸
ʹǤͲͺͲ
ͳǤͲͲͲ
ͲǤͳʹ͸
͹Ǥͻͷͺ
͹Ǥͻͷͺ
ͳǤͲͲͲ
77
ɉ
= ͳ
ͳ
=
ͳ ͳ
ɉ ͳ െ
ͳ ɉ
ͳǤͲͲͲ ʹǤͲͺͲ ͹Ǥͻͷͺ
ͲǤͶͺͳ ͳǤͲͲͲ ͹Ǥͻͷͺ
ͲǤͳʹ͸ ͲǤͳʹ͸ ͳǤͲͲͲ
ɉ െ ͳǤͲͲͲ
ʹǤͲͺͲ
͹Ǥͻͷͺ
ͲǤͶͺͳ
ɉ െ ͳǤͲͲͲ
͹Ǥͻͷͺ
ͲǤͳʹ͸
ͲǤͳʹ͸
ɉ െ ͳǤͲͲͲ
By using the Characteristic Equation for the determinant,
|ɉ I – A | = 0
ɉ െ ͳ െ͹Ǥͻͷͺ
െͲǤͶͺͳ
െ(െ2.080)
െͲǤͳʹ͸ ɉ െ ͳ
െͲǤͳʹ͸
െͲǤͶͺͳ ɉ െ ͳ
(െ7.958)
=0
െͲǤͳʹ͸ െͲǤͳʹ͸
(ɉ - 1)
െ͹Ǥͻͷͺ
+
ɉെͳ
(ɉ െ ͳ)(ɉଶ െ 2ɉ ൅ ͳ െ ͳǤͲͲ͵) + (2.080)(-0.481ɉ ൅ ͲǤͶͺͳ െ ͳǤͲͲ͵) –
7.958(0.061 + 0.126ɉ െ ͲǤͳʹ͸) = 0
ɉଷ - 3ɉଶ - 0.006ɉ - 0.566 = 0
‫ ׵‬ɉ௠௔௫ = 3.062
(for a reciprocal matrix, ɉ୫ୟ୶ ൒ always, where n is the number of the
elements being compared )
(ii)
Consistency Index (CI)
Consistency Index
=
=
஛ౣ౗౮ ି௡
௡ିଵ
ଷǤ଴଺ଶିଷ
ଷିଵ
= 0.031
78
(iii)
Consistency Ratio (CR)
Since the comparisons are carried out through personal or subjective
judgments, some degree of inconsistency may occur. To guarantee the judgments
are consistent, the final operation, consistency verification, which is regarded as one
of the best advantages of AHP is incorporated in order to measure the degree of
consistency among the pair-wise comparisons by computing the consistency ratio
(Ho, 2008).
By referring to the random consistency index in Table 3.2, the random
consistency (RI) for the matrix size of 3 x 3 is 0.58.
Consistency Ratio (CR)
=
=
େ୍
ୖ୍
ଷǤ଴଺ଶ
଴Ǥହ଼
= 0.05 (< 0.1, acceptable)
(iv)
Weight or priority of the criteria or sub-criteria
The weights, w for the criteria or sub-criteria can be calculated by using the formulae:
AW = ɉW
and w1 + w2 + w3= 1
where ɉ‹•–Š‡ɉ୫ୟ୶ , the eigenvalue and W is the eigenvector.
By using the ɉ୫ୟ୶ obtained, the weights or priorities of the criteria or sub-criteria can
be calculated using the method involving eigenvector and eigenvalue.
79
‫ݓ‬ଵ
‫ݓ‬
Let W = ଶ
‫ݓ‬ଷ
, where w1, w2, w3 are the weights that need to be calculated.
AW = ɉW
ͳǤͲͲͲ
ͲǤͶͺͳ
ͲǤͳʹ͸
ʹǤͲͺͲ
ͳǤͲͲͲ
ͲǤͳʹ͸
‫ݓ‬ଵ
‫ݓ‬ଵ
͹Ǥͻͷͺ
‫ݓ‬
‫ݓ‬
͹Ǥͻͷͺ ଶ = 3.062 ଶ
‫ݓ‬ଷ
‫ݓ‬ଷ
ͳǤͲͲͲ
‫ݓ‬ଵ ൅ ʹǤͲͺ™ଶ ൅ ͹Ǥͻͷͺ™ଷ
ͲǤͶͺͳ‫ݓ‬ଵ ൅ ‫ݓ‬ଶ ൅ ͹Ǥͻͷͺ™ଷ
ͲǤͳʹ͸™ଵ ൅ ͲǤͳʹ͸™ଶ ൅ ‫ݓ‬ଷ
͵ǤͲ͸ʹ‫ݓ‬ଵ
= ͵ǤͲ͸ʹ‫ݓ‬ଶ
͵ǤͲ͸ʹ‫ݓ‬ଷ
(1െ3.062w1) + 2.08w2 + 7.958w3 = 0
0.481w1 + (1 െ 3.062w2) + 7.958w3 = 0
0.126w1 + 0.126w2 + (1 െ 3.062w3) = 0
െ2.062w1 + 2.08w2 + 7.958w3 = 0
……………………………………(5.1)
0.481w1 െ 2.062w2 + 7.958w3 = 0
……………………………………(5.2)
0.126w1 + 0.126w2 – 2.062w3 = 0
………………………………..….(5.3)
w1 + w2 + w3 = 1
…………………………………….(5.4)
From Equation (5.4),
w1 = 1 െw2 െ w3
……………………………………(5.5)
By substituting Equation (5.5) into Equation (5.1):
െ2.062(1 െw2 െ w3) + 2.08w2 + 7.958w3 = 0
2.062w2 +2.062 w3 + 2.08w2 + 7.958w3 = 0
w2 െ w3)+ 2.08w2 + 7.958w3 = 0
4.142w2 + 10.02 w3 െ2.062 = 0
……………………………………(5.6)
By substituting Equation (5.5) into Equation (5.2):
0.481(1 െw2 െ w3) െ 2.062w2 + 7.958w3 = 0
0.481െ 0.481w2 Ȃ 0.481w3 െ 2.062w2 + 7.958w3 = 0
80
0.481 െ 2.543w2 + 7.477w3 = 0
െ 2.543w2 + 7.477w3 + 0.481 = 0
……………………………………(5.7)
By substituting Equation (5.5) into Equation (5.3):
0.126(1 െw2 െ w3) + 0.126w2 – 2.062w3 = 0
0.126 Ȃ 0.126w2 Ȃ 0.126w3 + 0.126w2 – 2.062w3 = 0
0.126 – 2.188 w3 = 0
w3 =
଴Ǥଵଶ଺
ଶǤଵ଼଼
= 0.0576
By substituting w3 = 0.0576 into Equation (5.7):
െ 2.543w2 + 7.477(0.0576) + 0.481 = 0
2.543 w2 = 0.9099
w2 =
଴Ǥଽ଴ଽଽ
ଶǤହସଷ
= 0.3578
By substituting w2 = 0.3578, w3 = 0.0576 into Equation (5.4):
w1 = 1 – 0.3578 – 0.0576
= 0.5846
Therefore, w1 = 0.5846, w2 = 0.3578 and w3 = 0.0576.
The above discussion demonstrates in detail the steps required in calculating
the priorities for the sub-criteria under the criterion of knowledge. Table 5.7 shows
the summary of the sub-criteria and the priorities obtained from the above calculation.
81
Table 5.7. Priority of the sub-criteria of Knowledge.
Sub-criteria of KNW
Priority
PFS
0.5846
EDU
0.3578
GNL
0.0576
Similar calculation can be done to other matrices formed to obtain the
priorities or weights for each of the criteria and sub-criteria. Besides using manual
calculation, the value of eigenvalue and consistency ratio can be obtained by using
some online third party programmes and the Expert Choice 11.5.
(v)
Estimation of Relative Weights for Elements Using Expert Choice 11.5
Software
By using Expert Choice 11.5, the relative weights (priorities) of each criterion
and sub-criterion are synthesized.
From the qualitative information transformed
into quantitative data which was obtained from experts, the priority indices
synthesized for the criterion of Knowledge (KNW), General Traits (GNT), Scholar/
Extracurricular Activities (ACT), Working Experience/ Interpersonal Skill (EIT) and
References (RFC) are 0.492, 0.169, 0.155, 0.154 and 0.30 respectively as shown in
Figure 5.4 with the consistency ratio, CR 0.09. Among the criteria, Knowledge
(KNW) has shown the highest priority and References (RFC) has shown the lowest
priority in the selection process.
The result shows that the knowledge is the most
important requirement to be fulfilled by a candidate in order to be selected as the
academic staff in the Faculty of Science since academic staff selection is an
important process for the faculty as the decision affects the quality of education and
the success of the university.
82
Figure 5.4. The priorities for the criteria in the selection of the best candidate as
the most appropriate academic staff in Faculty of Science, UTM.
Under the sub-criterion of Knowledge (KNW), Professional Knowledge (PFS)
has the highest priority, 0.584, followed by Education Background (EDU), 0.358 and
General Knowledge (GNL), 0.057 with the consistency ratio, CR 0.06. The relative
priorities of these sub-criteria are shown in Figure 5.5.
Figure 5.5. The priorities for the sub-criterion of Knowledge (KNW).
For the criterion of Working Experience / Interpersonal Skill (EIT), the subcriterion Intelligence / Eloquence (ITL) has shown the highest priority index of 0.246
compared to Ability to Communicate in English (ABL), 0.243, Working Field (FLD),
0.237, Teaching Experience / Skill (TCH), 0.167, Working Duration, PRD, 0.078
and Independence (IDP), 0.029 with the consistency ratio of 0.08.
From the
priorities obtained, it can be seen that ITL has the priority index which is slightly
higher than the priority index of ABL. The difference of 0.03 shows that the subcriterion of ITL and ABL are having almost the same level of importance as the key
factors in the selection of academic staff under the criterion of Working Experience /
Interpersonal Skill (EIT). The priority indices for the sub-criteria of EIT are as
shown in Figure 5.6.
83
Figure 5.6.
The priorities for the sub-criteria of Working experience /
Interpersonal skill (EIT).
Figure 5.7 shows the priorities for the sub-criterion under General Traits
(GNT) in the selection of academic staff in Faculty of Science. Appearance (APR)
has scored the highest priority index of 0.432 compared to Manner/ Politeness
(MNR), 0.350, Professional Interest (PIT), 0.180 and Age (AGE), 0.038 with the CR
of 0.07. It shows that the appearance of a candidate is highly important compared to
other sub-criteria under the criterion of GNT whereas the age of a candidate is not
that important comparatively in the selection of an academic staff.
Figure 5.7. The priorities for the sub-criteria of General Traits (GNT).
Publication is the most important key factor under the criterion of Scholar/
Extracurricular Actitivies (ACT) followed by Researches (RSH) done, Rewards
(RWD) awarded and Experience with Diverse Population (PPL) gained. The priority
index for ACT, RSH, RWD and PPL is 0.480, 0.390, 0.80 and 0.490 respectively
with the CR of 0.02 in this model.
84
Figure 5.8.
The priorities for the sub-criteria of Scholar/ Extracurricular
Activities (ACT).
Figure 5.9 shows the local and global priorities synthesized for every
selection criteria and sub-criteria used in this AHP aided academic staff selection
model. The local priority represents the percentage of the parent node's priority that
is inherited by the child. The local priorities of the children of a node also sum to one.
In this study, the parent nodes are the selection criteria and the child nodes are the
sub-criteria under each of the selection criterion. Whereas the priority of each node
relative to the ‘Goal’ is called Global Priority.
For example, the global priority of Professional Knowledge (PFS) under the
criteria of Knowledge (KNW) is calculated as the following:
Global priority of PFS
= local priority of PFS x priority of KNW
= 0.584 x 0.492
= 0.287
85
Figure 5.9. The academic staff selection criteria and sub-criteria used in the
AHP model and their local and global priorities.
5.4
Summary
This chapter detailed the development of the AHP model as a systematic
evaluation model for the purpose of selecting the most appropriate academic staff in
Faculty of Science. Besides depending on the Expert Choice 11.5 to accomplish the
tedious calculations to obtain the priority and the consistency ratio, CR, for each of
the criterion or sub-criterion, the actual mathematical steps required in the manual
calculation of one of the level 2 sub-criteria are shown in detail.
The selection criteria adopted in the model developed also being discussed in
this chapter. The selection criteria can be easily modified and adopted by the model,
86
depending on the needs of the faculty to recruit an academic staff as the member of
the faculty.
Hence, the developed model is a flexible one where addition or
modification to the criteria can be done quite easily. Before implementing the model,
the member of academic staff selection committee will have to decide and study
rigorously on the selection criteria, discordance levels and the rating scales to avoid
unintended implicit weighting and wasting of time and effort done to recruit an
unsuitable staff.
In the next chapter, the profiles of 5 potential candidates are created. The
‘qualities’ of all the candidates are rated comparatively based on the criteria and subcriteria of the model discussed in this chapter. By using this model, the candidates
are then ranked. Figure 5.10 summarizes the overall work flow in this chapter.
87
<ŶŽǁůĞĚŐĞĂĐĐƋƵŝƐŝƚŝŽŶ
ͻ ĂƚĂĐŽůůĞĐƚƚŝŽŶ
DŝĐƌŽƐŽĨƚdžĐĞů
ͻ /ŶŝƚŝĂůĚĂƚƚĂƉƌŽĐĞƐƐŝŶŐ
ͻ 'ĞŽŵĞƚƌƌŝĐŵĞĂŶ
WĂŝƌǁŝƐĞŽŵƉĂƌŝƐŽŶDĂƚƌŝĐĞƐ
džƉĞƌƚŚŽŝĐĞ
ͻ ĂƚĂƚƌĂŶƐĨĞƌƌĞĚĨƌŽŵWĂŝƌǁŝƐĞ
ŽŵƉĂƌŝƐŽŶDĂƚƌŝĐĞƐ
Ž
WƌŝŽƌŝƚŝĞƐͬtĞŝŐŚƚƐΘŽŶƐŝƐƚĞŶĐLJ
ZĂƚŝŽ
ͻ ŝŐĞŶǀĂůƵĞͬŝŐĞŶǀĞĐƚŽƌDĞƚŚŽĚ
Ě
Figuree 5.10. Summary of the work flow in Chapter 5.
88
CHAPTER 6
RESULTS AND ANALYSIS
6.1
Introduction
In this chapter, AHP model as discussed in great depth in the last chapter will be
implemented to select the most appropriate candidate to be the faculty member of the
Faculty of Science. This is achieved by synthesizing the candidates’ priority weights by
implementing the AHP model developed in Chapter 5 in this Chapter. The priority
weights obtained from knowledge acquisition in Chapter 5 are used to rank the
candidates in the later part of this Chapter.
There are two main parts in this Chapter: Sub-section 6.2 explains the rating of
candidates in which the profiles of 5 candidates are generated, whereas Sub-section 6.3
details the results and analysis for the ranking of the candidates based on the generated
profiles using AHP model developed in Chapter 5. Analysis and discussion are done
based on the generated candidate profiles by using sensitivity analysis from Expert
Choice 11.5. Graphs and charts are drawn to aid in the discussion and analysis of the
results. Examples of calculations to obtain the priority weights are shown as well.
89
6.2
Rating Each Candidate
To rate each candidate, the selection committee members will individually
review each candidate and complete the rating scorecards (refer to Table 5.1) for each of
the candidates. Members can discuss their individual ratings until achieving consensus
for those criteria rated qualitatively.
Simulated profiles of candidates after interview sessions are generated. These
simulated profiles are fed into the AHP model developed in this study. This leads to the
ranking of the candidates. It is of the assumption that the profiles simulated have been
transformed from the qualitative information to the quantitative data after discussions
among the members of the selection committee. It is also assumed that 5 candidates are
chosen for further consideration in the selection process after their profiles and results of
interviews have been carefully evaluated.
Table 6.1 to Table 6.18 show the pair-wise comparison of the 5 candidate
profiles generated. By referring to the first row in Table 6.1, it can be seen that for the
sub-criterion of PFS, Candidate 1 is 0.2 times more preferable than Candidate 2, 2 times
more preferable than Candidate 3, 0.5 times more preferable than Candidate 4 and 0.333
times more preferable than Candidate 5. Candidate 1 is 0.2 times more preferable than
Candidate 2 is also giving the meaning of Candidate 2 is 5 times more preferable than
Candidate 1. This is shown by the reciprocal of 0.2 (1/0.2 = 5) in row 2 and column 1 in
Table 6.1. The similar pair-wise comparisons done for EDU, GNL, PRD, FLD, TCH,
IDP, ITL, ABL, MNR, APR, AGE, PIT, PBL, RSH, RWD, PPL and RFC are shown in
Table 6.12 to Table 6.18.
90
Table 6.1. Pair-wise comparison of candidate profile for the sub-criterion of
Knowledge: Professional Knowledge (PFS)
Candidate
Candidate
Candidate
Candidate
Candidate
1
2
3
4
5
Candidate 1
1.000
0.200
2.000
0.500
0.333
Candidate 2
5.000
1.000
3.000
2.000
1.000
Candidate 3
0.500
0.333
1.000
1.000
0.500
Candidate 4
2.000
0.500
1.000
1.000
1.000
Candidate 5
3.000
1.000
2.000
1.000
1.000
Table 6.2. Pair-wise comparison of candidate profile for the sub-criterion of
Knowledge: Education Background (EDU)
Candidate
Candidate
Candidate
Candidate
Candidate
1
2
3
4
5
Candidate 1
1.000
1.000
1.000
0.500
1.000
Candidate 2
1.000
1.000
2.000
1.000
1.000
Candidate 3
1.000
0.500
1.000
0.500
1.000
Candidate 4
2.000
1.000
2.000
1.000
1.000
Candidate 5
1.000
1.000
1.000
1.000
1.000
Table 6.3. Pair-wise comparison of candidate profile for the sub-criterion of
Knowledge: General Knowledge (GNL)
Candidate
Candidate
Candidate
Candidate
Candidate
1
2
3
4
5
Candidate 1
1.000
1.000
1.000
3.000
2.000
Candidate 2
1.000
1.000
0.500
1.000
1.000
Candidate 3
1.000
2.000
1.000
1.000
1.000
Candidate 4
0.333
1.000
1.000
1.000
0.500
Candidate 5
0.500
1.000
1.000
2.000
1.000
91
Table 6.4. Pair-wise comparison of candidate profile for the sub-criterion of
Working Experience / Interpersonal Skill: Working Duration (PRD)
Candidate
Candidate
Candidate
Candidate
Candidate
1
2
3
4
5
Candidate 1
1.000
0.500
2.000
1.000
0.500
Candidate 2
2.000
1.000
2.000
0.500
1.000
Candidate 3
0.500
0.500
1.000
1.000
2.000
Candidate 4
1.000
2.000
1.000
1.000
0.500
Candidate 5
2.000
1.000
0.500
2.000
1.000
Table 6.5. Pair-wise comparison of candidate profile for the sub-criterion of
Working Experience / Interpersonal Skill: Working Field (FLD)
Candidate
Candidate
Candidate
Candidate
Candidate
1
2
3
4
5
Candidate 1
1.000
1.000
2.000
3.000
2.000
Candidate 2
1.000
1.000
0.500
2.000
1.000
Candidate 3
0.500
2.000
1.000
1.000
0.500
Candidate 4
0.333
0.500
1.000
1.000
1.000
Candidate 5
0.500
1.000
2.000
1.000
1.000
Table 6.6. Pair-wise comparison of candidate profile for the sub-criterion of
Working Experience / Interpersonal Skill: Teaching Experience / Skill (TCH)
Candidate
Candidate
Candidate
Candidate
Candidate
1
2
3
4
5
Candidate 1
1.000
2.000
1.000
1.000
1.000
Candidate 2
0.500
1.000
2.000
1.000
3.000
Candidate 3
1.000
0.500
1.000
1.000
1.000
Candidate 4
1.000
1.000
1.000
1.000
1.000
Candidate 5
1.000
0.333
1.000
1.000
1.000
92
Table 6.7. Pair-wise comparison of candidate profile for the sub-criterion of
Working Experience / Interpersonal Skill: Independence (IDP)
Candidate
Candidate
Candidate
Candidate
Candidate
1
2
3
4
5
Candidate 1
1.000
1.000
1.000
2.000
2.000
Candidate 2
1.000
1.000
1.000
2.000
0.500
Candidate 3
1.000
1.000
1.000
1.000
0.500
Candidate 4
0.500
0.500
1.000
1.000
1.000
Candidate 5
0.500
2.000
2.000
1.000
1.000
Table 6.8. Pair-wise comparison of candidate profile for the sub-criterion of
Working Experience / Interpersonal Skill: Intelligence / Eloquence (ITL)
Candidate
Candidate
Candidate
Candidate
Candidate
1
2
3
4
5
Candidate 1
1.000
2.000
1.000
0.500
1.000
Candidate 2
0.500
1.000
2.000
0.500
1.000
Candidate 3
1.000
0.500
1.000
1.000
1.000
Candidate 4
2.000
2.000
1.000
1.000
0.500
Candidate 5
1.000
1.000
1.000
2.000
1.000
Table 6.9. Pair-wise comparison of candidate profile for the sub-criterion of
Working Experience / Interpersonal Skill: Ability to Communicate in English
(ABL)
Candidate
Candidate
Candidate
Candidate
Candidate
1
2
3
4
5
Candidate 1
1.000
0.500
1.000
2.000
2.000
Candidate 2
2.000
1.000
0.500
1.000
1.000
Candidate 3
1.000
2.000
1.000
1.000
2.000
Candidate 4
0.500
1.000
1.000
1.000
1.000
Candidate 5
0.500
1.000
0.500
1.000
1.000
93
Table 6.10. Pair-wise comparison of candidate profile for the sub-criterion of
General Traits: Manner / Politeness (MNR)
Candidate
Candidate
Candidate
Candidate
Candidate
1
2
3
4
5
Candidate 1
1.000
1.000
1.000
0.500
0.500
Candidate 2
1.000
1.000
2.000
1.000
0.500
Candidate 3
1.000
0.500
1.000
2.000
1.000
Candidate 4
2.000
1.000
0.500
1.000
1.000
Candidate 5
2.000
2.000
1.000
1.000
1.000
Table 6.11. Pair-wise comparison of candidate profile for the sub-criterion of
General Traits: Appearance (APR)
Candidate
Candidate
Candidate
Candidate
Candidate
1
2
3
4
5
Candidate 1
1.000
1.000
1.000
1.000
1.000
Candidate 2
1.000
1.000
0.500
1.000
2.000
Candidate 3
1.000
2.000
1.000
0.500
1.000
Candidate 4
1.000
1.000
2.000
1.000
1.000
Candidate 5
1.000
0.500
1.000
1.000
1.000
Table 6.12. Pair-wise comparison of candidate profile for the sub-criterion of
General Traits: Age (AGE)
Candidate
Candidate
Candidate
Candidate
Candidate
1
2
3
4
5
Candidate 1
1.000
1.000
2.000
1.000
0.500
Candidate 2
1.000
1.000
2.000
1.000
0.500
Candidate 3
0.500
0.500
1.000
0.500
0.333
Candidate 4
1.000
1.000
2.000
1.000
1.000
Candidate 5
2.000
2.000
3.000
1.000
1.000
94
Table 6.13. Pair-wise comparison of candidate profile for the sub-criterion of
General Traits: Professional Interest (PIT)
Candidate
Candidate
Candidate
Candidate
Candidate
1
2
3
4
5
Candidate 1
1.000
1.000
2.000
1.000
2.000
Candidate 2
1.000
1.000
2.000
3.000
2.000
Candidate 3
0.500
0.500
1.000
3.000
3.000
Candidate 4
1.000
0.333
0.333
1.000
2.000
Candidate 5
0.500
0.500
0.333
0.500
1.000
Table 6.14. Pair-wise comparison of candidate profile for the sub-criterion of
Scholar / Extra Curricular Activities: Publication (PBL)
Candidate
Candidate
Candidate
Candidate
Candidate
1
2
3
4
5
Candidate 1
1.000
1.000
3.000
2.000
0.500
Candidate 2
1.000
1.000
2.000
1.000
1.000
Candidate 3
0.333
0.500
1.000
0.500
0.500
Candidate 4
0.500
1.000
2.000
1.000
1.000
Candidate 5
2.000
1.000
2.000
1.000
1.000
Table 6.15. Pair-wise comparison of candidate profile for the sub-criterion of
Scholar / Extra Curricular Activities: Researches (RSH)
Candidate
Candidate
Candidate
Candidate
Candidate
1
2
3
4
5
Candidate 1
1.000
0.500
1.000
2.000
1.000
Candidate 2
2.000
1.000
0.500
1.000
2.000
Candidate 3
1.000
2.000
1.000
1.000
1.000
Candidate 4
0.500
1.000
1.000
1.000
2.000
Candidate 5
1.000
0.500
1.000
0.500
1.000
95
Table 6.16. Pair-wise comparison of candidate profile for the sub-criterion of
Scholar / Extra Curricular Activities: Rewards (RWD)
Candidate
Candidate
Candidate
Candidate
Candidate
1
2
3
4
5
Candidate 1
1.000
1.000
2.000
1.000
1.000
Candidate 2
1.000
1.000
2.000
1.000
2.000
Candidate 3
0.500
0.500
1.000
1.000
2.000
Candidate 4
1.000
1.000
1.000
1.000
2.000
Candidate 5
1.000
0.500
0.500
0.500
1.000
Table 6.17. Pair-wise comparison of candidate profile for the sub-criterion of
Scholar / Extra Curricular Activities: Experience with Diverse Population (PPL)
Candidate
Candidate
Candidate
Candidate
Candidate
1
2
3
4
5
Candidate 1
1.000
0.500
1.000
1.000
0.500
Candidate 2
2.000
1.000
2.000
2.000
2.000
Candidate 3
1.000
0.500
1.000
2.000
0.500
Candidate 4
1.000
0.500
0.500
1.000
2.000
Candidate 5
2.000
0.500
2.000
0.500
1.000
Table 6.18. Pair-wise comparison of candidate profile for the criterion of
References (RFC)
Candidate
Candidate
Candidate
Candidate
Candidate
1
2
3
4
5
Candidate 1
1.000
2.000
2.000
2.000
1.000
Candidate 2
0.500
1.000
3.000
1.000
1.000
Candidate 3
0.500
0.333
1.000
0.500
2.000
Candidate 4
0.500
1.000
2.000
1.000
1.000
Candidate 5
1.000
1.000
0.500
1.000
1.000
96
From the pair-wise comparison matrices (tables) formed, although the relative
strong points of each of the candidates for each of the criterion and sub-criterion can be
easily seen, but it is rather hard to identify the best candidate to be selected by looking at
the separate pair-wise comparison matrices. Therefore, the quantitative data in the
matrices needs to be transferred to the data grid so that the ranking and the analysis of
the ranking results can be done using Expert Choice 11.5.
6.3
Results and Analysis
The ranking of candidates is done using Expert Choice 11.5 in this study. This
ranking can be done using Microsoft Excel as well, as long as the consistency ratio is
acceptable. Analysis of the results (ranking of candidates) is explained in detail and
shown by using graphs, tables and charts drawn based on the priority weights
synthesized.
6.3.1
Candidates Priority Weights
Candidates priority weights with respect to each of the criterion or sub-criterion
can be synthesized after the entering of the relative importance of each of the candidates
into the data grid in Expert Choice 11.5, as explained in Chapter 5.
97
6.3.1.1 Priority Weights
Table 6.19 shows the summary of the candidates’ priority weights synthesized
with respect to the criterion, sub-criterion and the Goal using Expert Choice 11.5. The
priority weights of KNW for Candidate 1, 2, 3, 4 and 5 are: 0.146, 0.285, 0.134, 0.10
and 0.225 respectively; EIT: 0.233, 0.205, 0.188, 0.188 and 0.187 respectively; GNT:
0.183, 0.216, 0.205, 0.204 and 0.192 respectively; ACT: 0.220, 0.226, 0.162, 0.193 and
0.199 respectively; RFC: 0.285, 0.212, 0.147, 0.185 and 0.171 respectively. Whereas
the overall ranking of the candidates with respect to the Goal, to select the best candidate,
is Candidate 2, 0.247, Candidate 5, 0.206, Candidate 4, 0.201, Candidate 1, 0.184 and
Candidate 3, 0.162.
It is difficult to identify the best candidate to be selected through the Priority
Weights Table (refer to Table 6.19). However, the comparison between the candidates
based on the priority weights synthesized is easier to be seen in graphical form. By
using Expert Choice 11.5, the result of the candidates’ ranking can be analyzed and
shown by using graphical sensitivity analyses. The sensitivity analysis will be explained
in detail in the next sub-section.
Figure 6.1 shows the priority weights synthesized for the ranking of candidates
with respect to Professional Knowledge. From the bar chart shown, Candidate 2 has
obtained the highest priority weight, 0.349, followed by Candidate 5, 0.253, Candidate 4,
0.178, Candidate 3, 0.111 and Candidate 1, 0.109.
98
Table 6.19. Candidates priority weights synthesized with respect to the criteria,
sub-criteria and the Goal.
Criterion/
sub-criterion
Candidate 1 Candidate 2 Candidate 3 Candidate 4 Candidate 5
KNW
0.146
0.285
0.134
0.210
0.225
1. PFS
0.109
0.349
0.111
0.178
0.253
2. EDU
0.171
0.225
0.149
0.259
0.196
3. GNL
0.280
0.170
0.223
0.137
0.190
EIT
0.233
0.205
0.188
0.188
0.187
1. PRD
0.170
0.217
0.130
0.267
0.217
2. FLD
0.305
0.195
0.177
0.131
0.192
3. TCH
0.236
0.256
0.164
0.188
0.156
4. IDP
0.260
0.194
0.167
0.146
0.232
5. ITL
0.196
0.176
0.168
0.232
0.227
6. ABL
0.228
0.207
0.255
0.167
0.143
GNT
0.183
0.216
0.205
0.204
0.192
1. MNR
0.145
0.202
0.205
0.195
0.252
2. APR
0.190
0.203
0.207
0.229
0.171
3. AGE
0.187
0.187
0.098
0.217
0.311
4. PIT
0.245
0.297
0.223
0.139
0.095
ACT
0.220
0.226
0.162
0.193
0.199
1. PBL
0.242
0.212
0.098
0.189
0.259
2. RSH
0.201
0.229
0.229
0.195
0.147
3. RWD
0.226
0.252
0.172
0.218
0.131
4. PPL
0.139
0.314
0.172
0.178
0.198
RFC
0.285
0.212
0.147
0.185
0.171
GOAL
0.184
0.247
0.162
0.201
0.206
99
Figure 6.1. Priority weights of candidates synthesized with respect to Professional
Knowledge.
Figure 6.2. Normalized Priority weights of candidates synthesized with respect to
Professional Knowledge.
Figure 6.2 shows the normalized priority weights of each candidate synthesized
with respect to the same sub-criterion of Knowledge, Professional Knowledge. The
priority weights are normalized to obtain the relative importance compared to each other.
Candidate 2 which has the highest priority index is given the index as 1.000. The
priority weights of the rest of the candidates are comparatively to Candidate 2. For
example, Candidate 5 is only 0.724 point as strong as Candidate 2 in the Professional
Knowledge whereas Candidate 1 is only 0.313 point as strong as Candidate 2 under the
same sub-criterion.
Therefore, Candidate 2 is having the strongest strength in
Professional Knowledge compared to the rest of the candidates.
Figure 6.3 to Figure 6.19 show the rest of the ranking of candidates based on
each of the criterion or sub-criterion in the selection process with the priority weights
shown.
100
Figure 6.3. Priority weights of candidates synthesized with respect to Education
Background.
Figure 6.4. Priority weights of candidates synthesized with respect to General
Knowledge.
Figure 6.5. Priority weights of candidates synthesized with respect to Working
Duration.
Figure 6.6. Priority weights of candidates synthesized with respect to Working
Field.
101
Figure 6.7. Priority weights of candidates synthesized with respect to Teaching
Experience / Skill.
Figure 6.8.
Priority weights of candidates synthesized with respect to
Independence.
Figure 6.9. Priority weights of candidates synthesized with respect to Intelligence /
Eloquence.
Figure 6.10. Priority weights of candidates synthesized with respect to Ability to
Communicate in English.
102
Figure 6.11. Priority weights of candidates synthesized with respect to Manner /
Politeness.
Figure 6.12. Priority weights of candidates synthesized with respect to Appearance.
Figure 6.13. Priority weights of candidates synthesized with respect to Age.
Figure 6.14. Priority weights of candidates synthesized with respect to Professional
Interest.
103
Figure 6.15. Priority weights of candidates synthesized with respect to Publication.
Figure 6.16. Priority weights of candidates synthesized with respect to Researches.
Figure 6.17. Priority weights of candidates synthesized with respect to Rewards.
Figure 6.18. Priority weights of candidates synthesized with respect to Experience
with Diverse Population.
104
Figure 6.19. Priority weights of candidates synthesized with respect to References.
Figure 6.20. Priority weights of candidates synthesized with respect to Goal – Best
candidate.
Figure 6.20 shows the overall priority weights synthesized with respect to the
goal for this study, i.e. to select the best candidate as a faculty member in Faculty of
Science, for each of the candidates. From the bar chart shown, Candidate 2 is having the
highest priority weight (0.247), followed by Candidate 5 (0.206), Candidate 4 (0.201),
Candidate 1 (0.184) and Candidate 3 is having the lowest ranking (0.162).
Figure 6.21 presents the priority weights for the criterion of RFC, sub-criteria
and the Goal for the 5 candidates in one bar chart.
Ϭ͘ϬϬϬ
Ϭ͘ϬϱϬ
Ϭ͘ϭϬϬ
Ϭ͘ϭϱϬ
Ϭ͘ϮϬϬ
Ϭ͘ϮϱϬ
Ϭ͘ϯϬϬ
Ϭ͘ϯϱϬ
Ϭ͘ϰϬϬ
d,
/W
/d>
> DEZ WZ '
W/d
W>
Z^, Zt WW>
Z& 'K>
Figure 6.21. Priority weights of the sub-criterion, criterion RFC and Goal in the selection process.
W&^ h 'E> WZ &>
ĂŶĚŝĚĂƚĞϱ
ĂŶĚŝĚĂƚĞϰ
ĂŶĚŝĚĂƚĞϯ
ĂŶĚŝĚĂƚĞϮ
ĂŶĚŝĚĂƚĞϭ
105
105
106
6.3.1.2 Calculation of Priority Weights
By referring to the local sub-criteria weights synthesized as shown in Figure 5.9
and the priority weights of candidates in Table 6.19, the calculation of the priority of
each of the candidates can be done. The following examples show the details of the
calculation for the criteria weight and candidates’ priority weights.
(i)
Calculation of criterion weight with respect to KNW for the candidates
based on the sub-criteria weight
Table 6.20. Candidate priority weights synthesized with respect to the KNW.
Local Weight
Sub-criterion
(Priority) of Subcriterion, ࢖࢐
Priority weight of Priority weight of
Candidate 3, ࢗ૜࢏
Candidate 5, ࢗ૞࢏
PFS
0.287
0.111
0.253
EDU
0.176
0.149
0.196
GNL
0.028
0.223
0.190
KNW
0.492
0.134
0.225
Let ‫݌‬௝ , j = 1, 2, 3 be the weights of the 3 sub-criteria and ‫ݍ‬௜௝ , i = 1, 2, .., 5 are the
priority weights of the 5 candidates on the sub-criterion j, j = 1, 2, 3, then the weights of
the 5 candidates with respect to KNW can be obtained as mentioned by Rafikul (2003):
wi = σଷ௝ୀଵ ‫݌‬௝ ‫ݍ‬௜௝ ,
i = 1, 2 ,.., 5
(6.1)
107
By using Equation (6.1) and referring to Table 6.20, the calculation of priority weight
with respect to the criterion of KNW for Candidate 5 and 3 are shown as the following:
Priority weight for Candidate 5 with respect to KNW
= p1q51 + p2q52 + p3q53
= 0.584(0.253) + 0.358(0.196) + 0.057(0.190)
= 0.22875
Priority weight for Candidate 3 with respect to KNW
= p1q31 + p2q32 + p3q33
= 0.584(0.111) + 0.358(0.149) + 0.057(0.223)
= 0.130877
The priority weights for the criterion of EIT, GNT, ACT and RFC for each of the
candidates can be obtained by using the same method as shown above.
(ii)
Calculation of overall priority weights with respect to Goal of each
candidates based on the criteria weight synthesized
Table 6.21. Priority weights of the criteria weight and Candidate 3 and 5.
Criterion Weight,
Priority weight of
Priority weight of
࢖࢐
Candidate 3, ࢗ૜࢏
Candidate 5, ࢗ૞࢏
KNW
0.492
0.134
0.225
EIT
0.154
0.188
0.187
GNT
0.169
0.205
0.192
ACT
0.155
0.162
0.199
RFC
0.030
0.147
0.171
0.162
0.206
Criterion
GOAL
108
By modifying Equation (6.1), let ‫݌‬௝ , j = 1, 2,.., 5 be the weights of the 5 criteria
and ‫ݍ‬௜௝ , i = 1, 2, .., 5 are the priority weights of the 5 candidates on the 5 criteria j, j = 1,
2, .., 5 then the weights of the 5 candidates with respect to Goal in this study can be
calculated by using the formula:
wi = σହ௝ୀଵ ‫݌‬௝ ‫ݍ‬௜௝ ,
i = 1,2,.., 5
(6.2)
The calculation of priority weight with respect to the Goal for Candidate 5 and 3 are as
shown in the following calculation by using Equation (6.2) and referring to Table 6.21:
Priority weight of Candidate 5 with respect to Goal, w5
= p1q51 + p2q52 + p3q53 + p4q54 + p5q55
= 0.492(0.225) + 0.154(0.187) + 0.169(0.192) + 0.155(0.199) + 0.030(0.171)
= 0.207921
Priority weight of Candidate 3 with respect to Goal, w5
= p1q31 + p2q32 + p3q33 + p4q34 + p5q35
= 0.492(0.134) + 0.154(0.188) + 0.169(0.205) + 0.155(0.162) + 0.030(0.147)
= 0.159045
Notes: Due to the limited 3 decimal places of the priority weights synthesized using
Expert Choice 11.5, there is a slight difference between the priority weights calculated
manually and those obtained from Expert Choice 11.5.
109
6.3.2
Sensitivity Analysis
Sensitivity Analysis is used to investigate the sensitivity of the alternatives
(candidates) to changes in the priorities of the objectives (criteria and sub-criteria).
There are five types of sensitivity analyses which can be performed using Expert Choice
11.5:
1. Performance
2. Dynamic
3. Gradient
4. 2-D plot
5. Head-to-Head.
Analyses can be performed from the Goal node or from the current node in the
hierarchy such as an objective. The purpose of sensitivity analyses is to graphically
present the changes of alternatives (candidates) with respect to the importance of the
objectives (criteria or sub-criteria).
Each sensitivity analysis from the 5 types of
analyses can be performed from the Goal or from a selected criterion or sub-criterion. In
all cases, there must be at least two levels below the selected node.
6.3.2.1 Performance Sensitivity
The Performance Sensitivity Graph displays the performance of candidates with
respect to all of the criteria or sub-criteria as well as the overall performance. The
Performance Sensitivity Graph for the candidates with respect to Goal is shown in
Figure 6.22. The priority indices for each of the candidates under criteria indicated can
be seen from the lines drawn in the Performance Sensitivity Graph. From the graph
110
drawn, it is shown that Candidate 2 is rated the highest priority index in KNW, GNT and
ACT and second in EIT and RFC compared to Candidate 1, 3, 4 and 5. The overall
priority index of Candidate 2 is the highest since KNW has given the highest weight or
priority of 0.492 compared to the criterion of EIT, 0.154, GNT, 0.169, ACT, 0.155 and
RFC, 0.030. The weight given for each of the sub-criterion is shown by bars drawn in
the graph which is synthesized based on the knowledge acquisition from experts in this
study (refer to Figure 5.9).
KďũĞĐƚŝǀĞďĂƌƐ
Figure 6.22. Performance sensitivity graph with respect to Goal.
The Performance graph is a dynamic one in which the relationship between the
alternatives (candidates) and their objectives (sub-criteria) can be temporarily altered by
dragging the objective bars up or down. The lines connecting the alternatives from one
objective to another included is helping in finding the particular alternative lies as it
moves from one objective to another.
111
6.3.2.2 Dynamic Sensitivity
Dynamic Sensitivity analysis is used to dynamically change the priorities of the
objectives to determine how these changes affect the priorities of the alternative choices.
Figure 6.23 shows the screen shot of the dynamic sensitivity graph synthesized with
respect to the Goal. From the graph, it can be seen that the weight or priority of KNW is
given 49.2% from the overall priority in the selection model, whereas EIT, 15.4%, GNT,
16.9%, ACT, 15.5% and RFC, 3%. Based on the weights or priorities synthesized from
the knowledge acquisition from experts which is implemented in the model used in this
study, Candidate 2 is rated the highest in the overall ranking in this selection model
which is 24.7%.
This is followed by Candidate 5, 20.6%, Candidate 4, 20.1%,
Candidate 1, 18.4% and Candidate 3, 16.2%.
Figure 6.23. Dynamic sensitivity graph with respect to the Goal.
112
If an objective (criterion) is thought to be more or less important than originally
indicated, the objective's bar can be dragged to the right or left to increase or decrease
the objective's priority and to see the impact on alternatives (candidates). This can be
done by dragging the objective's priorities back and forth in the left column, the
priorities of the alternatives will change in the right column. For example, as the priority
of one objective increases (by dragging the bar to the right) the priorities of the
remaining objectives decrease in proportion to their original priorities, and the priorities
of the alternatives are recalculated. The detail of the changes will be discussed in the
sub-section of 6.3.2.6.
6.3.2.3 Gradient Sensitivity Graph
This graph shows the alternatives' (candidates’) priorities with respect to one
objective at a time. The vertical red line represents the priority of the selected objective
and is read from the X-Axis intersection. The priorities for the alternatives are read from
the Y-Axis and it is determined by the intersection of the alternative's line with the
objective's (vertical) priority line.
For example, Figure 6.24 shows a gradient sensitivity graph drawn with respect
to the Goal for the 5 candidates ranked. The vertical red line represents the priority of
the KNW, (0.492, referring to Figure 5.9) which is read from the X-Axis intersection.
The priority index for Candidate 2 is the highest, followed by candidate 5, Candidate 4,
Candidate 1 and Candidate 3 which is determined by the intersection of the candidates’
line with the vertical priority line.
113
Figure 6.24. Gradient sensitivity graph with respect to Goal.
The objective's priority can be easily changed by dragging the red bar to either
the left or right, a blue bar showing the new objective's priority will be displayed. The
Gradient Sensitivity shows "key tradeoffs" when two or more alternatives intersect each
other. This is even more important if the intersection is close to the objectives priority.
6.3.2.4 Head-to-Head Sensitivity Analysis
Head-to-Head Sensitivity Analysis shows the comparison between two
alternatives (candidates) against the objectives in a decision. One alternative is listed on
the left side of the graph and the other is listed on the right. The alternative on the left is
fixed, while selecting a different tab on the graph can vary the alternative on the right.
Down the middle of the graph are listed the objectives in the decision. If the left-hand
114
alternative is preferred to the right-hand alternative with respect to an objective, a
horizontal bar is displayed towards the left. If the right-hand alternative is better, the
horizontal bar will be on the right. If the two choices are equal, no bar is displayed. The
overall result is displayed at the bottom of the graph and shows the overall percentage
that one alternative is better than the other; this is the difference. The overall priority can
either be shown based on the objective weights (typical) or un-weighted.
Figure 6.25 and Figure 6.26 show the screen shots of the Head-to-Head
Sensitivity Analysis Graph for the comparison between Candidate 1 and 2 and Candidate
4 and 1 respectively. Candidate 2 is showing a greater priority than Candidate 1 in
Figure 6.24 whereas Candidate 4 is showing the greater priority than candidate 1 in
Figure 6.26.
Figure 6.25. Head-to-Head Sensitivity Analysis Graph for the comparison between
Candidate 1 and Candidate 2.
115
Figure 6.26. Head-to-Head Sensitivity Analysis Graph for the comparison between
Candidate 4 and Candidate 1.
6.3.2.5 Two-Dimensional Sensitivity
This Two-Dimensional Sensitivity Graph shows the performance of the
alternatives (candidates) with respect to any two objectives (criteria). This graph is
sometimes known as the Bubble Plot. One objective is represented on the X-Axis and
another on the Y-Axis. The circles represent the alternatives. The area of the 2D plot is
divided into quadrants. The most favorable alternatives as defined by the objectives and
judgments in the model will be shown in the upper right quadrant (the closer to the upper
right hand corner the better) while, conversely, the least favorable alternatives will be
shown in the lower left quadrant. Alternatives located in the upper left and lower right
quadrants indicate key tradeoffs where there is conflict between the two objectives.
When the projection line of alternatives is on, projection not only shows how
preferable the alternatives are with respect to the two selected objectives (criteria) but it
shows a composite projection line indicating the preference of each alternative taking
116
into account all the objectives' priorities. The farther to the right on the line, the better
the alternative in which each alternative has two circles. The large circles represent how
preferable the alternatives are with respect to the two objectives that have been. The
smaller circles along the composite line indicate the overall preference of the alternatives
with respect to all the objectives.
Figure 6.27 shows the performance of candidates using Two-Dimensional
Sensitivity Graph with respect to KNW and ACT. Candidate 2 has shown a greater
priority than the rest whereas Candidate 3 is showing the least in priority among the
candidates.
Figure 6.27. Two-Dimensional Sensitivity Graph synthesized with respect to KNW
and ACT.
117
6.3.2.6 Impact of Changes in Weighting of Criteria Using Sensitivity Analysis
Sensitivity Analysis is used in surveying the criterion weight to determine the
influences on a candidate hierarchy in this study. As an example, a survey is made as of
how the criteria and sub-criteria weights influence the 5 candidates’ hierarchy.
According to the earlier result, by referring to Figure 6.21 to Figure 6.24, the candidates’
hierarchy is of the following order: Candidate 2, Candidate 5, Candidate 4, Candidate 1
and Candidate 3. When the weight/ priority of KNW decreases from 49.2% to 36.4%,
EIT decreases from 15.4% to 2.3%, GNT decreases from 16.9% to 15.6%, ACT
increases from 15.5% to 20.4% and RFC increases from 3.0% to 25.3%, the best
candidate to be selected will change from Candidate 1 to Candidate 5.
As shown in Figure 6.28 and Figure 6.29, the hierarchy of candidates has
changed to the following order: Candidate 2, Candidate 1, Candidate 5, Candidate 4 and
Candidate 3 with the changes made to the criteria weights as mentioned above. This
show that the weighting of criteria can easily be changed depending on the faculty
selection requirement and by using sensitivity analysis graphs the influences can be
easily seen by the decision makers before making their decision.
118
Figure 6.28.
Performance Sensitivity for nodes below Goal (after changes of
criteria weight made).
Figure 6.29. Dynamic Sensitivity for nodes below Goal (after changes of criteria
weight made)
119
Figure 6.30. Gradient Sensitivity Graph drawn with respect to ACT (after changes
of criteria weight made).
Figure 6.30 represents the Gradient sensitivity drawn with respect to ACT. The
blue vertical line shows the new priority indices of ACT after the changes have been
made. The new priority indices of the candidates can be read from the Y-Axis and it is
determined by the intersection of the candidates’ line with the vertical (blue) priority line.
Table 6.22 shows the difference for the ranking of candidates after the changes of
criteria weight (priority) are made.
Table 6.22. Ranking of candidates after changes of criteria weight is made.
Criteria
Model Weight
Ranking of
New Weight
Ranking of
(%)
candidates
(%)
candidates
KNW
49.2
1. Candidate 2
36.4
1. Candidate 2
EIT
15.4
2. Candidate 5
2.3
2. Candidate 1
GNT
16.9
3. Candidate 4
15.6
3. Candidate 5
ACT
15.5
4. Candidate 1
20.4
4. Candidate 4
RFC
3.0
5. Candidate 3
25.2
5. Candidate 3
120
>ĞǀĞůϬ͗
'ŽĂů
>ĞǀĞůϭ͗
ƌŝƚĞƌŝĂ
>ĞǀĞůϮ͗
^ƵďͲĐƌŝƚĞƌŝĂ
>ĞǀĞůϯ͗
ůƚĞƌŶĂƚŝǀĞƐ
W&^;>͗Ϭ͘ϱϴϰ͕'͗Ϭ͘ϮϴϳͿ
<Et
;Ϭ͘ϰϵϮͿ
h;>͗Ϭ͘ϯϱϴ͕'͗Ϭ͘ϭϳϲͿ
'E>;>͗Ϭ͘Ϭϱϳ͕'͗Ϭ͘ϬϮϴͿ
WZ;>͗Ϭ͘Ϭϳϴ͕'͗Ϭ͘ϬϭϮͿ
^ĞůĞĐƚŝŽŶŽĨƚŚĞŵŽƐƚĂƉƉƌŽƉƌŝĂƚĞĂĐĂĚĞŵŝĐƐƚĂĨĨ
&>;>͗Ϭ͘Ϯϯϳ͕'͗Ϭ͘ϬϯϲͿ
/d
;Ϭ͘ϭϱϰͿ
d,;>͗Ϭ͘ϭϲϳ͕'͗Ϭ͘ϬϮϲͿ
/W;>͗Ϭ͘ϬϮϵ͕'͗Ϭ͘ϬϬϰͿ
/d>;>͗Ϭ͘Ϯϰϲ͕'͗Ϭ͘ϬϯϴͿ
>;>͗Ϭ͘Ϯϰϯ͕'͗Ϭ͘ϬϯϳͿ
DEZ;>͗Ϭ͘ϯϱϬ͕'͗Ϭ͘ϬϱϵͿ
'Ed
;Ϭ͘ϭϲϵͿ
ĂŶĚŝĚĂƚĞϮ
;Ϭ͘ϮϰϳͿ
ĂŶĚŝĚĂƚĞϯ
;Ϭ͘ϭϲϮͿ
WZ;>͗Ϭ͘ϰϯϮ͕'͗Ϭ͘ϬϳϯͿ
';>͗Ϭ͘Ϭϯϴ͕'͗Ϭ͘ϬϬϲͿ
W/d;>͗Ϭ͘ϭϴϬ͕'͗Ϭ͘ϬϯϬͿ
W>;>͗Ϭ͘ϰϴϬ͕'͗Ϭ͘ϬϳϱͿ
d
;Ϭ͘ϭϱϱͿ
ĂŶĚŝĚĂƚĞϭ
;Ϭ͘ϭϴϰͿ
Z^,;>͗Ϭ͘ϯϵϬ͕'͗Ϭ͘ϬϲϭͿ
ĂŶĚŝĚĂƚĞϰ
;Ϭ͘ϮϬϭͿ
ĂŶĚŝĚĂƚĞϱ
;Ϭ͘ϮϬϲͿ
Zt;>͗Ϭ͘ϬϴϬ͕'͗Ϭ͘ϬϭϮͿ
WW>;>͗Ϭ͘Ϭϰϵ͕'͗Ϭ͘ϬϬϴͿ
Z&
;Ϭ͘ϬϯͿ
Figure 6.31. Hierarchy Tree with the Priority Weights.
121
Figure 6.31 shows the hierarchy tree for the AHP model developed in this study
with the priority weights for each of the criterion, sub-criterion and the candidates. There
are two types of priority weights shown for each of the sub-criteria at level 2, local
priority and global priority.
As explained in Sub-section 5.3.5.2, local priority represents the percentage of
the parent node’s (level 1 nodes, criteria) priority which in inherited by the child (level 2
nodes, sub-criteria) and the local priorities of the children of a node are sum to one. For
example, PFS has a local priority weight of 0.584, EDU, 0.358 and GNL, 0.057. This
shows that, PFS has 58.4%, EDU, 35.8% and GNL, 5.7% from the parent node, which is
KNW. The sum of the local priorities of the KNW is (0.584 + 0.358 + 0.057) ൎ 1.
Whereas global priority of a node represents the portion of the parent’s (criterion)
priority inherited by the child (sub-criterion) and the global priorities of all the children
equal to the parent’s global priority. The global priority of a child equals the local
priority of the child time global priority of the parent. For example, the following shows
the calculation to obtain the global priority weight for KNW in which KNW has the
priority of 0.492 (local priority and global priority weights are the same for the level 1
node).
Table 6.23. Calculation of global priority weight.
Local priority
Global priority weight
Total of Global priority
weights for
weight
sub-criteria
PFS
0.584
0.584 ൈ 0.492 = 0.287
of KNW
EDU
0.358
0.358 ൈ 0.492 = 0.176
= 0.287 + 0.176 + 0.028
GNL
0.057
0.057 ൈ 0.492 = 0.028
ൎ 0.492
122
Local priority does not show the importance of a sub-criterion for the overall
selection process. For example, although the local priority weight for PFS is 0.584, but
the global priority weight is only 0.287 which means that the importance of PFS
contribute to the overall selection process is only 28.7% compared to the priority of PFS
contribute to KNW which is 58.4%.
6.4
Summary
There are two types of priority weights synthesized in this study:
1. Priority weight of criteria and sub-criteria (Chapter 5)
2. Priority weight of candidates (Chapter 6)
Five simulated candidate profiles are generated in order to represent the
candidates selected for the final selection for the Faculty of Science. The ranking of
candidates is done using the AHP model developed in this chapter. Sensitivity Analysis
graphs from Expert Choice 11.5 provided a very useful tool in analyzing and enabling
the discussion on the ranking of candidates made in great dept.
From the results, (refer to Figure 6.30), it can be seen that Candidate 2 is
showing the highest priority weight than Candidate 1, 3, 4 and 5. Therefore, Faculty of
Science should choose Candidate 2 as the academic staff as its priority weight has the
highest value according to the ranking using the AHP model developed in this study.
The result of priority weights of the 5 candidates (alternatives) further reveals that the
order of these alternatives in this study is Candidate 2 > Candidate 5 > Candidate 4 >
Candidate 1 > Candidate 3.
123
CHAPTER 7
SUMMARY AND CONCLUSION
7.1
Introduction
This Chapter presents a summary of the work done throughout this study.
The AHP model developed in this study is summarized and the ranking of candidates
are done as well as the advantages of AHP aided decision making model are
described in this chapter. The implementation of AHP aided decision making model
in the academic staff selection process in Faculty of Science, UTM is summarized by
using a flow chart.
Some suggestion for future works in the area of selecting a best candidate to
be an academic staff in the Faculty of Science, UTM is given.
124
7.2
Summary
This paper intends to implement the use of AHP in weighting the information
for an academic staff selection process in Faculty of Science in UTM. AHP is a
method which considers both qualitative and quantitative approaches to research and
intends to combine them into a single methodology. More specifically, it uses a
qualitative way to decompose an unstructured problem into a decision hierarchy and
induces an iterative process to solve any inconsistent responses.
Pair-wise
comparison with a prescribed absolute scale and performs the consistency test to
validate the consistency of respondents is employed in this study. Inconsistency
refers to a lack of transitivity of preferences (Saaty, 1980). Those respondents who
could not build up their judgments logically would not achieve the consistent
comparisons (Cheng and Heng, 2001).
A case study on the selection of academic staff in Faculty of Science, UTM is
done. Essentially, this study explains how the identification of key factors helps in
better selection of a best candidate to be the academic staff in Faculty of Science. It
reflects the increasing importance of managerial information that forms a key role in
the selection of academic staff in the faculty. A more standardized or established
selection model which is customized for the selection of academic staff in Faculty of
Science, UTM is developed and the demonstration on how the AHP provide a wellstructured, coherent, and justifiable selection practices is shown in this study.
The decision criteria used in the model developed are Knowledge, Working
Experience/ Interpersonal Skill, General Traits, Scholar/ Extra Curricular Activities
and References. These criteria were evaluated to determine the order of candidates
for the selection of the most appropriate academic staff for the faculty. By the help
of the AHP model, the ambiguities involved in the data can be effectively
represented and processed to make a more effective decision. Figure 7.1 shows
where to implement the AHP aided decision making model in the selection of
academic staff in Faculty of Science in this study.
125
/ĚĞŶƚŝĨŝĐĂƚŝŽŶŽĨ
ũŽďǀĂĐĂŶĐLJŝŶ
ƚŚĞĨĂĐƵůƚLJ
/ŶĨŽƌŵƐƐŝƐƚĂŶ
Ŷƚ
ZĞŐŝƐƚƌĂƌŝŶƚŚĞĞ
ĨĂĐƵůƚLJ
Selection Moddel using
Analytic Hierarcchy Process
(AHP
P)
&ŝƌƐƚƐŚŽƌƚͲůŝƐƚŝŶ
ŶŐ
/ƐƐƵĞĐĂůůůĞƚƚĞƌƌ
ĨŽƌŝŶƚĞƌǀŝĞǁ
/ŶƚĞƌǀŝĞǁ
^ĞĐŽŶĚƐŚŽƌƚͲ
ůŝƐƚŝŶŐ
^ƵďŵŝƚƌĞůĂƚĞĚ
Ě
ĚŽĐƵŵĞŶƚƐƚŽ
ZĞŐŝƐƚƌĂƌ͛ƐKĨĨŝĐĐĞ
ϭ
ͻ ĞǀĞůŽƉŵĞŶƚŽĨĂ
ĐŽŶĐĞƉƚƵĂůĨƌĂĂŵĞǁŽƌŬ
Ϯ
ͻ ^ĞƚƚŝŶŐŽĨƚŚĞĚĞĐŝƐŝŽŶ
ŚŝĞƌĂƌĐŚLJ
ϯ
ͻ KďƚĂŝŶŝŶŐŝŶĨŽ
ŽƌŵĂƚŝŽŶ
ĨƌŽŵĞdžƉĞƌƚƐ
ϰ
ͻ ŵƉůŽLJŝŶŐƚŚĞĞƉĂŝƌͲǁŝƐĞ
ĐŽŵƉĂƌŝƐŽŶ
ϱ
ͻ ƐƚŝŵĂƚŝŶŐƌĞůĂƚŝǀĞ
ǁĞŝŐŚƚƐŽĨĞůĞĞŵĞŶƚƐ
ϲ
ͻ ĂůĐƵůĂƚŝŶŐƚŚĞĚĞŐƌĞĞŽĨ
ĐŽŶƐŝƐƚĞŶĐLJ
ϳ
ͻ ŽŵƉƵƚŝŶŐƚŚĞĞŶƚŝƌĞ
ŚŝĞƌĂƌĐŚLJǁĞŝŐŐŚƚ
ϴ
ͻ ZĂƚŝŶŐŽĨĞĂĐŚ
ŚĐĂŶĚŝĚĂƚĞƐ
Figure 7.1. Implem
mentation of AHP aided academic staff seleection model in the
selection process in Faculty
F
of Science, UTM.
126
As a result of this study, Candidate 2 is determined as the best alternative
which has the highest priority weight of 0.247 in this study. Candidate 2 is selected
based on the relative judgments made by the experts in the knowledge acquisition
process as described in Chapter 4 and 5. In a study done by Cheng and Heng (2001),
it is mentioned in their paper that the more a person knows about a situation, the
more consistent the results that can be expected from this person which means that
the experiences and knowledge for a decision maker is vital in the determination of
the priority weight of criteria and sub-criteria in the selection process.
Researcher agreed with the advantages of AHP which had pointed out by
Cheng and Heng (2001) in which the AHP has the following advantages:
1
AHP adopts a pair-wise comparison process by comparing two objects at one
time to formulate a judgment as to their relative weight
2
With an adequate measurement, this method is more accurate (with less
experimental error) to achieve a higher level of consistency since it requires
the respondents to think precisely before giving their answers
By using the AHP model, committee members or decision makers who were
not extensively trained in AHP can quickly understood the process and could easily
generate the required matrices. AHP permits pair-wise comparison judgments which
are documented and can be re-examined to express the relative strength or intensity
of impact of the elements (criteria) in the hierarchy. Hence, the criteria and the subcriteria as well as their relative importance can be easily changed when necessarily
by the selection committee.
The AHP model is free from biases as many decision makers rely on ‘gut
feelings’ in making decisions since it can transform the qualitative information
gathered into quantitative data in the selection process. Therefore, AHP provides the
127
opportunity for the decision makers to include feelings and intuition by fully and
correctly articulating their thoughts.
In order to account for the boundary conditions under the model holds, AHP
users need to incorporate sensitivity analyses in the selection process. The AHP
outcome presented in this study depends on the relative judgments made by the
experts from the process of knowledge acquisition. Changes in the judgments of the
main criteria will lead to a change in the outcome and explain the final decision. The
sensitivity of the outcome can be easily investigated and understood by using the
Expert Choice 11.5.
AHP saves time in the long run by making sure meeting of the decision
maker advance along a track to a conclusion rather than circling in confusion.
Comparison between 2 candidates can also be easily seen by using the sensitivity
analysis (Head-to-Head Sensitivity) from Expert Choice 11.5.
In short, selection
committee could benefit tremendously by embracing the AHP-based selection
process proposed in this study. As mentioned by Gibney and Shang (2007), AHP is
a practical, versatile and powerful tool that explicitly identifies the factors that matter
and provides a consistent structure and process for evaluating candidates whereas as
concluded by Grandzol (2005) in his paper that the process articulated by AHP aided
faculty member selection model possesses the following advantages:
1.
Avoid faculty recruitment to become a tedious, time consuming and
frustrating for all parties involved (committee members, administrators and
applicants) when it is not well-defined, effective and efficient
2. Minimize consumption of resources
3. Capture all pertinent preference issues
4. Fair and equitable to all participants (faculty and applicants)
128
5. Reusable
7.3
Suggestion for Future Work
Due to the time constrain of this research work, a simpler version of
hierarchy structure was built and used.
The following recommendations are
suggested for the further research in this study:
1.
Research area can be expanded to develop an AHP model for the academic
staff selection of the university.
2. Rating of candidates can be done in more detail for every single element of
the criteria and its sub-criteria.
3. In order to improve the result obtained, further research can be carried out
with a more perfect match of criteria to be use in the selection process of
Faculty of Science or UTM.
129
REFERENCES
Arvey, R. D. and Campion, J. E. (1982). The employment interview: A summary and
review of recent research. Personnel Psychology. 35(2), 281-322.
Avery, R. D. and Renz, G. L. (1992). Fairness in the Selection of Employees. Journal
of Business Ethics. 11, 331 – 340.
Boose, J. H. (1989). A Survey of Knowledge Acquisition Techniques and Tools.
Knowledge Acquisition. 1, 3 – 37.
Brent, A. C., Rogers, D. E. C., Ramabitsa-Siimane, T. S.M., Rohwer, M. B. (2007). O.R
Applications: Application of the analytical hierarchy prosess to establish health
care waste management systems that minimize infection risk in developing
countries. European Journal of Operational Research. 181, 403 – 424.
Brozova, H. (2004). The Analytic Hierarchy Process for the Decision Tree with Multiple
Criteria. AGRIC. ECON. – CZECH. 50(2), 77 – 82.
Cheng, E. W. L. and Heng, L. (2001). Information Priority-Setting for Better Resource
Allocation Using Analytic Hierarchy Process (AHP). Information Management &
Computer Security. 9(2), 61 – 70.
Ciptomulyono, U. (2008). Fuzzy Goal Programming Approach for Deriving Priority
Weights in the Analytical Hierarchy Process (AHP) method. Journal of Applied
Science Research. 4(2), 171 – 177.
130
Da˘gdeviren, M. (2008). A hybrid multi-criteria decision-making model for personnel
selection in manufacturing systems. Journal of Intelligent Manufacturing.
Dagdeviren, M., Yuksel, I. (2007). Personnel Selection Using Analytic Network Process.
østanbul Ticaret Üniversitesi Fen Bilimleri Dergisi Yıl: 6 Sayı:11Bahar. 99-118.
Dyer, R. F., Forman, E. H. (1992). Group Decision Support with the Analytic Hierarchy
Process. Decision Support System. 8, 99 – 124.
Ertu÷rul, I. and Karakaúo÷lu, N. (2007). Fuzzy TOPSIS Method for Academic Member
Selection in Engineering Faculty. M. Iskander (ed.), Innovations in E-learning,
Instruction Technology, Assessment and Engineering Education. 151–156.
Faggini, M. and Lux, T. (2009). New Economic Windows: Coping with the Complexity
of Economics. Springer Milan.
Formann, A. K. (1992). Academic Personnel Selection: Description and Prognosis of
the Decision Made by the Committee for the Selection of Candidates for a Full
Professorship. Scientometrics. 3, 401 – 414.
Gibney, R. and Shang, J. (2007). Decision making in academia: A case of the dean
selection process. Mathematical and Computer Modelling. 1030-1040.
Golec, A. and Kahya, E. (2007). A Fuzzy Model for Competency-based Employee
Evaluation and Selection. Computers & Industrial Engineering. 52, 143–161.
Grandzol, J. R. (2005). Improving the Faculty Selection Process in Higher Education: A
Case for the Analytic Hierarchy Process. IR Applications Using Advanced Tools,
Techniques, and Methodologies. 6, 1 – 11.
Grensing, L. (1986). A Small Business guide to Employee Selection.
Canada.
International Self-Counsel Press Ltd.
Gunion, R. M. (1998). Some Virtues of Dissatisfaction in the Science and Practice of
Personnel Selection. Human Resource Management Review. 8(4), 351 – 365.
131
Hishamuddin, M. S. (2000). Handbook on Managerial Selection Techniques. UTM,
Skudai, Malaysia. Cetak Ratu Sdn. Bhd.
Ho, W. (2008). Integrated Analytic Hierarchy process and its Applications – A
Literature Review. European Journal of Operational Research. 186, 211 – 228.
Hwang, C.L. and Yoon, K. (1981). Multiple Attribute Decision Making. Methods and
Applications. Lecture Notes in Economics and Mathematical System. Vol. 186.
Springer-Verlag, Berlin.
Kahraman, C. (2008). Springer Optimization and Its Application Volume 16: Fuzzy
Multi-Criteria Decision-Making:
Theory and Applications with Recent
Developments. Springer US.
Lam, K and Zhao, X. (1998).
An Application of quality function deployment to
improve the quality of teaching. International Journal of Quality and Reliability
Management. 15(4), 389 – 413.
Leigh, I. W., Smith, I. L., Bebeau, M. J., Lichtenberg, J. W., Nelson, P. D., Portnoy, S.,
Rubin, N. J., & Kaslow, N. J. (2007). Competency assessment models.
Professional Psychology: Research and Practice. 38(5), 463–473.
Linkov, I., Satterstrom, F. K. , Tkachuk, A., Levchenko, A., Seager, T.P., Figueira, J. R.
and Tervonen, T. (2007). NATO Security through Science Series:
Managing
Critical Infrastructure Risks. Netherlands : Springer.
Mamat, N. J. Z. and Daniel, J.K. (2007). Statistical analyses on time complexity and
rank consistency between singular value decomposition and the duality approach
in AHP: A case study of faculty member selection. Mathematical and Computer
Modelling. 46, 1099 – 1106.
Mamat, N. J. Z., Daniel, J. K. (2004). Faculty Member Selection; A Comparative Study
of AHP and its Variants. 17th International Conference on Multiple Criteria
Decision Making (MCDM 2004). 6 – 11 Aug 2004. Canada : Whistler.
132
Nussbauma, M., Singerb, M., Rosasc, R., Castilloa, M., Fliesa, E., Larad, R., Sommersd,
R. (1999). Decision Support System for Conflict Diagnosis in Personnel Selection.
Information & Management. 36, 55 – 62 .
Olson, D. L. (2008). Handbook on Decision Support System I. Berlin Heidelberg :
Springer.
Rafikul, I. (2003). The Analytic Hierarchy Process: An Effective Multi-criteria Decision
Making Tool. Malaysia: International Islamic University Malaysia.
Saaty, T. L. (1980). The Analytic Hierarchy Process. New York. Mc Graw Hill, NY.
Saaty, T. L. and Vargas, L. G. (2006). Decision Making with the Analytic Network
Process:
Economic, Political, Social and Technological Applications with
Benefits, Opportunities, Costs and Risks. Springer US.
Saaty, T. L., Ramanujam, V. (1983). An Objective Approach to Faculty Promotion and
Tenure by the Analytic Hierarchy Process. Research in Higher Education. 18(3),
311-331.
Saaty, T.L., Kearns, K.P. (1985). Analytical Planning: The Organization of Systems.
London: Oxford Press Ltd.
Shih, H. S., Huang, L. C., Shyur, H. J. (2005). Recruitment and Selection Processes
Through an Effective GDSS. Computers and Mathematics with Applications. 50,
1543 – 1558.
Slaughter, J. E., Bagger, J., Li, A. (2006). Context Effects o Group-Based Employee
Selection Decisions. Organizational Behavior and Human Decision Processes.
100, 47 – 59.
Tadisina, S. K. and Bhasin, V. (1989). Doctoral Program Selection Using Pair-wise
Comparisons. Research in Higher Education. 30(4), 403 – 418.
133
Taylor III, F. A., Ketcham, A. F. and Hoffman, D. (1998). Personnel Evaluation with
AHP. Management Decision. 36(10), 679 – 685.
Tung, R. L. (1981). Selection and training of personnel for overseas assignments.
Columbia Journal of World business. 16(1), 68-78.
Tung, S. L. and Tang, S. L. (1998). A comparison of the Saaty's AHP and modified AHP
for right and left eigenvector inconsistency. Elsevier Science B.V.
Tzeng, G.H., Teng, M.H., Chen, J.J., Opricovic, S. (2002). Multicriteria selection for a
restaurant location in Taipei. International Journal of Hospitality Management.
21, 171 – 187.
Wasil, E and Golden, B. (2003). Editorial: Celebrating 25 years of AHP-based Decision
Making. Computers & Operations Research. 30, 1419 – 1420.
Wilson, J. R. and Corlett, E. N. (1995). Evaluation of Human Work: A Practical
Ergonomics Methodology. London. Taylor and Francis.
Zhang, G. Q. and Lu, J. 2003. An Integrated Group Decision-Making Method Dealing
with Fuzzy Preferences for Alternatives and Individual Judgments for Selection
Criteria. Group Decision and Negotiation. Kluwer Academic Publishers. 12, 501–
515.
133
APPENDIX A
SAMPLE OF DOCUMENTS OBTAINED FROM KNOWLEDGE
ACQUISITION OF CASE STUDY
134
Figure A1. Marking rubric.
135
Figure A2. Tutor selection approval table.
136
Figure A3. Lecturer/ Tutor selection approval (paage 1).
137
Figure A4. Lecturer/ Tutor selection approval (paage 2).
138
Figure A5. Lecturer/ Tutor selection approval (paage 3).
139
Figure A6. Lecturer/ Tutor selection approval (paage 4).
140
.
Figure A7. Lecturer/ Tutor selection approval (paage 5).
141
Figurre A8. Profile amendment of work instructtion.
142
Figure A9. Academic staff recruitment and selection flow
f
chart.
143
Figure A10. Detail of
o work instruction for recruitment and seelection of
academic staff (pagee 1).
144
Figure A11. Detail of
o work instruction for recruitment and seelection of
academic staff (pagee 2).
145
Figure A12. Detail of
o work instruction for recruitment and seelection of
academic staff (pagee 3).
146
Figure A13. Detaail of work instruction for recruitment and selection of
academic staff (pagee 4).
147
Figure A14. Detail of
o work instruction for recruitment and seelection of
academic staff (pagee 5).
148
APPENDIX B
PAIR-WISE COMPARISON TABLE
(TO BE COMPLETED BY EXPERTS)
Table B1. Comparison of the relative importance with respect to: GOAL
(Select the Best Candidate).
149
ϭ
Ϯ
ϯ
ϰ
ϱ
ĐĂĚĞŵŝĐƐƚĂĨĨ^ĞůĞĐƚŝŽŶĐƌŝƚĞƌŝĂ;ůĞǀĞůϭͿ͗
<ŶŽǁůĞĚŐĞ͕<Et;WƌŽĨĞƐƐŝŽŶĂůŬŶŽǁůĞĚŐĞ͕ĞĚƵĐĂƚŝŽŶďĂĐŬŐƌŽƵŶĚͿ
tŽƌŬŝŶŐĞdžƉĞƌŝĞŶĐĞͬŝŶƚĞƌƉĞƌƐŽŶĂůƐŬŝůů͕/d;ǁŽƌŬŝŶŐƉĞƌŝŽĚ͕ǁŽƌŬŝŶŐĨŝĞůĚ͕
ƉƌŽĨĞƐƐŝŽŶĂůŝŶƚĞƌĞƐƚͿ
'ĞŶĞƌĂůƚƌĂŝƚƐ͕'Ed;ŵĂŶŶĞƌ͕ƉŽůŝƚĞŶĞƐƐͬĂƉƉĞƌĂŶĐĞͿ
^ĐŚŽůĂƌͬĞdžƚƌĂĐƵƌƌŝĐƵůĂƌĂĐƚŝǀŝƚŝĞƐŝŶǀŽůǀĞĚƉƌĞǀŝŽƵƐůLJ͕d;ƉƵďůŝĐĂƚŝŽŶ͕ƌĞƐĞĂƌĐŚĞƐ͕
ƌĞǁĂƌĚƐ͕ĞdžƉĞƌŝĞĐŶĞǁŝƚŚĚŝǀĞƌƐĞƉŽƉƵůĂƚŝŽŶͿ
ZĞĨĞƌĞŶĐĞƐ͕Z&
ϭсYh>ϯсDKZdϱс^dZKE'ϳсsZz^dZKE'ϵсydZD
ϭ
<Et
ϵ
ϴ
ϳ
ϲ
ϱ
ϰ
ϯ
Ϯ
ϭ
Ϯ
ϯ
ϰ
ϱ
ϲ
ϳ
ϴ
ϵ
/d
Ϯ
<Et
ϵ
ϴ
ϳ
ϲ
ϱ
ϰ
ϯ
Ϯ
ϭ
Ϯ
ϯ
ϰ
ϱ
ϲ
ϳ
ϴ
ϵ
'Ed
ϯ
<Et
ϵ
ϴ
ϳ
ϲ
ϱ
ϰ
ϯ
Ϯ
ϭ
Ϯ
ϯ
ϰ
ϱ
ϲ
ϳ
ϴ
ϵ
d
ϰ
<Et
ϵ
ϴ
ϳ
ϲ
ϱ
ϰ
ϯ
Ϯ
ϭ
Ϯ
ϯ
ϰ
ϱ
ϲ
ϳ
ϴ
ϵ
Z&
ϱ
/d
ϵ
ϴ
ϳ
ϲ
ϱ
ϰ
ϯ
Ϯ
ϭ
Ϯ
ϯ
ϰ
ϱ
ϲ
ϳ
ϴ
ϵ
'Ed
ϲ
/d
ϵ
ϴ
ϳ
ϲ
ϱ
ϰ
ϯ
Ϯ
ϭ
Ϯ
ϯ
ϰ
ϱ
ϲ
ϳ
ϴ
ϵ
d
ϳ
/d
ϵ
ϴ
ϳ
ϲ
ϱ
ϰ
ϯ
Ϯ
ϭ
Ϯ
ϯ
ϰ
ϱ
ϲ
ϳ
ϴ
ϵ
Z&
ϴ
'Ed
ϵ
ϴ
ϳ
ϲ
ϱ
ϰ
ϯ
Ϯ
ϭ
Ϯ
ϯ
ϰ
ϱ
ϲ
ϳ
ϴ
ϵ
d
ϵ
'Ed
ϵ
ϴ
ϳ
ϲ
ϱ
ϰ
ϯ
Ϯ
ϭ
Ϯ
ϯ
ϰ
ϱ
ϲ
ϳ
ϴ
ϵ
Z&
ϭϬ
d
ϵ
ϴ
ϳ
ϲ
ϱ
ϰ
ϯ
Ϯ
ϭ
Ϯ
ϯ
ϰ
ϱ
ϲ
ϳ
ϴ
ϵ
Z&
Table B2. Comparison of the relative importance with respect to Knowledge.
^ƵďĐƌŝƚĞƌŝĂŽĨ<ŶŽǁůĞĚŐĞ͗
ϭ WƌŽĨĞƐƐŝŽŶĂůŬŶŽǁůĞĚŐĞ͕W&^
Ϯ ĚƵĐĂƚŝŽŶďĂĐŬŐƌŽƵŶĚ͕h
ϯ 'ĞŶĞƌĂů<ŶŽǁůĞĚŐĞ͕'E>
ϭсYh>ϯсDKZdϱс^dZKE'ϳсsZz^dZKE'ϵсydZD
ϭ
W&^
ϵ
ϴ
ϳ
ϲ
ϱ
ϰ
ϯ
Ϯ ϭ Ϯ ϯ ϰ
ϱ
ϲ
ϳ ϴ ϵ h
Ϯ
W&^
ϵ
ϴ
ϳ
ϲ
ϱ
ϰ
ϯ
Ϯ ϭ Ϯ ϯ ϰ
ϱ
ϲ
ϳ ϴ ϵ 'E>
ϯ
h
ϵ
ϴ
ϳ
ϲ
ϱ
ϰ
ϯ
Ϯ ϭ Ϯ ϯ ϰ
ϱ
ϲ
ϳ ϴ ϵ 'E>
150
Table B3. Comparison of the relative importance with respect to working
experience/ interpersonal skill.
^ƵďĐƌŝƚĞƌŝĂŽĨǁŽƌŬŝŶŐĞdžƉĞƌŝĞŶĐĞͬŝŶƚĞƌƉĞƌƐŽŶĂůƐŬŝůů͗
ϭ
Ϯ
ϯ
ϰ
ϱ
ϲ
tŽƌŬŝŶŐƉĞƌŝŽĚ͕WZ
tŽƌŬŝŶŐĨŝĞůĚ͕&>
dĞĂĐŚŝŶŐĞdžƉĞƌŝĞŶĐĞͬƐŬŝůů͕d,
/ŶĚĞƉĞŶĚĞŶĐĞ͕/W
/ŶƚĞůůŝŐĞŶĐĞͬĞůŽƋƵĞŶĐĞ͕/d>
ďŝůŝƚLJƚŽĐŽŵŵƵŶŝĐĂƚĞŝŶŶŐůŝƐŚ͕>
ϭсYh>ϯсDKZdϱс^dZKE'ϳсsZz^dZKE'ϵсydZD
ϭ
WZ
ϵ
ϴ
ϳ
ϲ
ϱ
ϰ
ϯ
Ϯ ϭ Ϯ ϯ ϰ
ϱ
ϲ
ϳ ϴ ϵ
&>
Ϯ
WZ
ϵ
ϴ
ϳ
ϲ
ϱ
ϰ
ϯ
Ϯ ϭ Ϯ ϯ ϰ
ϱ
ϲ
ϳ ϴ ϵ
d,
ϯ
WZ
ϵ
ϴ
ϳ
ϲ
ϱ
ϰ
ϯ
Ϯ ϭ Ϯ ϯ ϰ
ϱ
ϲ
ϳ ϴ ϵ
/W
ϰ
WZ
ϵ
ϴ
ϳ
ϲ
ϱ
ϰ
ϯ
Ϯ ϭ Ϯ ϯ ϰ
ϱ
ϲ
ϳ ϴ ϵ
/d>
ϱ
WZ
ϵ
ϴ
ϳ
ϲ
ϱ
ϰ
ϯ
Ϯ ϭ Ϯ ϯ ϰ
ϱ
ϲ
ϳ ϴ ϵ
>
ϲ
&>
ϵ
ϴ
ϳ
ϲ
ϱ
ϰ
ϯ
Ϯ ϭ Ϯ ϯ ϰ
ϱ
ϲ
ϳ ϴ ϵ
d,
ϳ
&>
ϵ
ϴ
ϳ
ϲ
ϱ
ϰ
ϯ
Ϯ ϭ Ϯ ϯ ϰ
ϱ
ϲ
ϳ ϴ ϵ
/W
ϴ
&>
ϵ
ϴ
ϳ
ϲ
ϱ
ϰ
ϯ
Ϯ ϭ Ϯ ϯ ϰ
ϱ
ϲ
ϳ ϴ ϵ
/d>
ϵ
&>
ϵ
ϴ
ϳ
ϲ
ϱ
ϰ
ϯ
Ϯ ϭ Ϯ ϯ ϰ
ϱ
ϲ
ϳ ϴ ϵ
>
ϭϬ
d,
ϵ
ϴ
ϳ
ϲ
ϱ
ϰ
ϯ
Ϯ ϭ Ϯ ϯ ϰ
ϱ
ϲ
ϳ ϴ ϵ
/W
ϭϭ
d,
ϵ
ϴ
ϳ
ϲ
ϱ
ϰ
ϯ
Ϯ ϭ Ϯ ϯ ϰ
ϱ
ϲ
ϳ ϴ ϵ
/d>
ϭϮ
d,
ϵ
ϴ
ϳ
ϲ
ϱ
ϰ
ϯ
Ϯ ϭ Ϯ ϯ ϰ
ϱ
ϲ
ϳ ϴ ϵ
>
ϭϯ
/W
ϵ
ϴ
ϳ
ϲ
ϱ
ϰ
ϯ
Ϯ ϭ Ϯ ϯ ϰ
ϱ
ϲ
ϳ ϴ ϵ
/d>
ϭϰ
/W
ϵ
ϴ
ϳ
ϲ
ϱ
ϰ
ϯ
Ϯ ϭ Ϯ ϯ ϰ
ϱ
ϲ
ϳ ϴ ϵ
>
ϭϱ
/d>
ϵ
ϴ
ϳ
ϲ
ϱ
ϰ
ϯ
Ϯ ϭ Ϯ ϯ ϰ
ϱ
ϲ
ϳ ϴ ϵ
>
151
Table B4. Comparison of the relative importance with respect to General Traits.
^ƵďĐƌŝƚĞƌŝĂŽĨ'ĞŶĞƌĂůƚƌĂŝƚƐ͗
ϭ
Ϯ
ϯ
ϰ
DĂŶŶĞƌͬƉŽůŝƚĞŶĞƐƐ͕DEZ
ƉƉĞĂƌĂŶĐĞ͕WZ
ŐĞ͕'
WƌŽĨĞƐƐŝŽŶĂů/ŶƚĞƌĞƐƚ͕W/d
ϭсYh>ϯсDKZdϱс^dZKE'ϳсsZz^dZKE'ϵсydZD
ϭ
DEZ
ϵ
ϴ
ϳ
ϲ
ϱ
ϰ
ϯ
Ϯ ϭ Ϯ ϯ ϰ
ϱ
ϲ
ϳ ϴ ϵ
WZ
Ϯ
DEZ
ϵ
ϴ
ϳ
ϲ
ϱ
ϰ
ϯ
Ϯ ϭ Ϯ ϯ ϰ
ϱ
ϲ
ϳ ϴ ϵ
'
ϯ
DEZ
ϵ
ϴ
ϳ
ϲ
ϱ
ϰ
ϯ
Ϯ ϭ Ϯ ϯ ϰ
ϱ
ϲ
ϳ ϴ ϵ
W/d
ϰ
WZ
ϵ
ϴ
ϳ
ϲ
ϱ
ϰ
ϯ
Ϯ ϭ Ϯ ϯ ϰ
ϱ
ϲ
ϳ ϴ ϵ
'
ϱ
WZ
ϵ
ϴ
ϳ
ϲ
ϱ
ϰ
ϯ
Ϯ ϭ Ϯ ϯ ϰ
ϱ
ϲ
ϳ ϴ ϵ
W/d
ϲ
'
ϵ
ϴ
ϳ
ϲ
ϱ
ϰ
ϯ
Ϯ ϭ Ϯ ϯ ϰ
ϱ
ϲ
ϳ ϴ ϵ
W/d
Table B5. Comparison of the relative importance with respect to Scholar /
Extracurricular activities.
^ƵďĐƌŝƚĞƌŝĂŽĨ^ĐŚŽůĂƌͬdžƚƌĂĐƵƌƌŝĐƵůĂƌĂĐƚŝǀŝƚŝĞƐ͗
ϭ
Ϯ
ϯ
ϰ
WƵďůŝĐĂƚŝŽŶ͕W>
ZĞƐĞĂƌĐŚĞƐ͕Z^,
ZĞǁĂƌĚƐ͕Zt
džƉĞƌŝĞŶĐĞǁŝƚŚĚŝǀĞƌƐĞƉŽƉƵůĂƚŝŽŶ͕WW>
ϭсYh>ϯсDKZdϱс^dZKE'ϳсsZz^dZKE'ϵсydZD
ϭ
W>
ϵ
ϴ
ϳ
ϲ
ϱ
ϰ
ϯ
Ϯ ϭ Ϯ ϯ ϰ
ϱ
ϲ
ϳ ϴ ϵ
Z^,
Ϯ
W>
ϵ
ϴ
ϳ
ϲ
ϱ
ϰ
ϯ
Ϯ ϭ Ϯ ϯ ϰ
ϱ
ϲ
ϳ ϴ ϵ Zt
ϯ
W>
ϵ
ϴ
ϳ
ϲ
ϱ
ϰ
ϯ
Ϯ ϭ Ϯ ϯ ϰ
ϱ
ϲ
ϳ ϴ ϵ
ϰ
Z^,
ϵ
ϴ
ϳ
ϲ
ϱ
ϰ
ϯ
Ϯ ϭ Ϯ ϯ ϰ
ϱ
ϲ
ϳ ϴ ϵ Zt
ϱ
Z^,
ϵ
ϴ
ϳ
ϲ
ϱ
ϰ
ϯ
Ϯ ϭ Ϯ ϯ ϰ
ϱ
ϲ
ϳ ϴ ϵ
WW>
ϲ
Zt
ϵ
ϴ
ϳ
ϲ
ϱ
ϰ
ϯ
Ϯ ϭ Ϯ ϯ ϰ
ϱ
ϲ
ϳ ϴ ϵ
WW>
WW>
152
APPENDIX C
SENSITIVITY ANALYSIS
153
Figure C1. Performance sensitivity graph with respect to KNW.
Figure C2. Performance sensitivity graph with respect to EIT.
154
Figure C3. Performance sensitivity graph with respect to GNT.
Figure C4. Performance sensitivity graph with respect toACT.
155
Figure C5. Dynamic sensitivity graph with respect to KNW.
Figure C6. Dynamic sensitivity graph with respect to EIT.
156
Figure C7. Dynamic sensitivity graph with respect to GNT.
Figure C8. Dynamic sensitivity graph with respect to ACT.
157
Figure C9. Gradient sensitivity graph with respect to KNW.
Figure C10. Gradient sensitivity graph with respect to EIT.
158
Figure C11. Gradient sensitivity graph with respect to GNT.
Figure C12. Gradient sensitivity graph with respect to ACT.
159
W&^
W&^
h
h
'E>
'E>
KsZ>>
KsZ>>
W&^
W&^
h
h
'E>
'E>
KsZ>>
KsZ>>
W&^
W&^
h
h
'E>
'E>
KsZ>>
KsZ>>
W&^
W&^
h
h
'E>
'E>
KsZ>>
KsZ>>
W&^
W&^
h
h
'E>
'E>
KsZ>>
KsZ>>
Figure C13. Head-to-Head sensitivity graph with respect to KNW.
160
WZ
WZ
&>
&>
d,
d,
/W
/W
/d>
/d>
>
>
KsZ>>
KsZ>>
WZ
WZ
&>
&>
d,
d,
/W
/W
/d>
/d>
>
>
KsZ>>
KsZ>>
WZ
WZ
&>
&>
d,
d,
/W
/W
/d>
/d>
>
>
KsZ>>
KsZ>>
WZ
&>
d,
/W
/d>
>
KsZ>>
WZ
WZ
&>
d,
/W
/d>
>
KsZ>>
WZ
&>
&>
d,
d,
/W
/W
/d>
/d>
>
>
KsZ>>
KsZ>>
Figure C14. Head-to-Head sensitivity graph with respect to EIT.
161
DEZ
DEZ
WZ
WZ
'
'
W/d
W/d
KsZ>>
KsZ>>
DEZ
DEZ
WZ
WZ
'
'
W/d
W/d
KsZ>>
KsZ>>
DEZ
WZ
'
'
W/d
W/d
KsZ>>
KsZ>>
DEZ
DEZ
DEZ
WZ
DEZ
WZ
WZ
'
'
W/d
W/d
KsZ>>
KsZ>>
DEZ
WZ
WZ
'
'
W/d
W/d
KsZ>>
KsZ>>
Figure C15. Head-to-Head sensitivity graph with respect to GNT.
162
W>
W>
Z^,
Z^,
Zt
Zt
WW>
WW>
KsZ>>
KsZ>>
W>
W>
Z^,
Z^,
Zt
Zt
WW>
WW>
KsZ>>
KsZ>>
W>
W>
Z^,
Z^,
Zt
Zt
WW>
WW>
KsZ>>
KsZ>>
W>
W>
Z^,
Z^,
Zt
Zt
WW>
WW>
KsZ>>
KsZ>>
W>
W>
Z^,
Z^,
Zt
Zt
WW>
WW>
KsZ>>
KsZ>>
Figure C16. Head-to-Head sensitivity graph with respect to ACT.
163
Figure C17. Two-Dimensional sensitivity graph synthesized with respect to EIT.
Figure C18. Two-Dimensional sensitivity graph synthesized with respect to
GNT.