Upload
hoangthuy
View
219
Download
1
Embed Size (px)
Citation preview
AA SSYYSSTTEEMM DDYYNNAAMMIICCSS AAPPPPRROOAACCHH TTOO
CCOONNSSTTRRUUCCTTIIOONN SSAAFFEETTYY CCUULLTTUURREE
By
TThhaannwwaaddeeee CChhiinnddaa
B.Eng, M.Eng
A thesis submitted in fulfillment of the requirements for the award of the degree of
DDooccttoorr ooff PPhhiilloossoopphhyy
From
Centre for Infrastructure Engineering and Management
Griffith School of Engineering
Faculty of Science, Environment, Engineering and Technology
GGrriiffffiitthh UUnniivveerrssiittyy
Gold Coast, Australia
October 2007
A System Dynamics Approach to Construction Safety Culture
i
DDEECCLLAARRAATTIIOONN
This work has not previously been submitted for a degree or diploma in any university.
To the best of my knowledge and belief, the thesis contains no material previously
published or written by another person except where due reference is made in the thesis
itself.
________________
Thanwadee Chinda
October 2007
�
A System Dynamics Approach to Construction Safety Culture
ii
AACCKKNNOOWWLLEEDDGGEEMMEENNTT
This research study would not be possible without the encouragement and assistance of
so many individuals and organizations.
First, and foremost, I thank my supervisor, Professor Sherif Mohamed, for the
opportunity, encouragement, guidance, and support he has given me over the years of
this research.
Thanks also to my associate supervisor, Dr. Rodney Stewart, for his assistance. Special
thanks go to Ms. Sandra Paine and Ms. Mary Ping for their help and support throughout
the study. I also thank Professor Yew-Chaye Loo for his vision in establishing the
School of Engineering, Griffith University, on the Gold Coast Campus. I am here
because of his energy and enthusiasm for engineering research.
I thank Mrs. Pullapha Noinonthong for her help in translating the survey questionnaire
into the Thai language. I am grateful to Mrs. Suangsurat Munka and the Department of
Labour Protection and Welfare, Ministry of Labour, Thailand, for helping me in the
process of data collection. Thanks also to the Thai construction organizations that
contributed their knowledge and experience to the questionnaire survey. Invaluable
assistance in data collection was given by Mr. Kumpanath Kaewtungmuang, Dr.
Wanchai Asvapoositkul, and Lt. Chanaton Surarak; thank you all very much.
I also thank my colleagues at the Griffith School of Engineering, especially Mr. Jaeho
Lee, Ms. Le Chen, and Mr. Kriengsak Panuwatwanich for their help in my data
analysis. I am also appreciative of the editorial assistance given by Ms. Carmel Wild
and Dr. Robyn Heales.
A System Dynamics Approach to Construction Safety Culture
iii
I am indebted to P’Kuum for his support, understanding, and thoughtfulness throughout
my Masters and PhD studies. May we remain good friends forever.
For supporting me personally, I thank P’Wut, P’Duang, P’Krieng, P’Ton, P’Sunan and
Bill, Heather, NaToo, NaNueng, NaNuan, and my friends at the Chiangmai Thai
restaurant. Thank you for making my life in Australia wonderful.
Last, but not least, I would like to dedicate this thesis to my beloved family: my Dad,
my Mom, P’Art, P’Aoy, my aunt, and my grandparents. Thank you for loving me,
believing in me, and giving me the strength to complete this large undertaking – a
successful PhD thesis. I love you all.
A System Dynamics Approach to Construction Safety Culture
iv
LLIISSTT OOFF PPUUBBLLIICCAATTIIOONNSS
The following papers were produced to disseminate the concepts and results of the work
undertaken by the author during the course of this PhD study.
Mohamed, S. and Chinda, T., 2005. Organizational safety culture: a system dynamics
approach. Proceedings of the 4th triennial international conference rethinking and
revitalizing construction safety, health, environment and quality, 17-20 May 2005, Port
Elizabeth, South Africa, 282-292.
Chinda, T. and Mohamed, S., 2006. Modelling construction safety culture using system
dynamics. In: D. Fang, R.M. Choudhry, and J.W. Hinze, eds. Proceedings of the CIB
W99 2006 international conference on global unity for safety and health in
construction, 28-30 June 2006, Beijing, China. Beijing: Tsinghua University Press, 165-
172.
Chinda, T. and Mohamed, S., 2007. Causal relationships between enablers of
construction safety culture. In: S.M. Ahmed, S. Azhar, and S. Mohamed, eds.
Proceedings of the fourth international conference on construction in the 21st century,
11-13 July 2007, Gold Coast, Australia. USA: CITC-IV, 438-445.
Chinda, T. and Mohamed, S., 2008. Structural equation model of construction safety
culture. Engineering, construction and architectural management, 15(2), 114-131.
Chinda, T. and Mohamed, S., 2008. System dynamic modeling of construction safety
culture. The 5th international conference on innovation in architecture, engineering, and
construction, 23-25 June 2008, Antalya, Turkey. Paper summitted for publication and
presentation.
�
A System Dynamics Approach to Construction Safety Culture
v
AABBSSTTRRAACCTT
Throughout the world, the construction industry has had a poor safety record, and is
disproportionately more dangerous when compared to other industries. The major cause
of construction accidents is attributed to unsafe behaviours and work practices, which
are viewed as the direct result of having a poor safety culture. The development of a
mature safety culture has been recognized as a vital element in the achievement of high
standards of safety, alongside an effective safety management system. A better
understanding of how to improve safety culture greatly assists an organization to
allocate appropriate safety resources, and thus improve its overall occupational health
and safety performance.
Recently, researches have been undertaken to measure the ‘health’ of construction
safety culture in an attempt to plan for safety culture improvement. Those studies,
however, have focused neither on the interactions among key safety culture elements,
nor on the consequences of safety initiatives being undertaken over time. Importantly,
construction organizations need to be able to measure their current safety culture
maturity level, and identify areas for safety improvement, to enable them to progress
through to higher maturity levels. Such actions are essential, as the implementation of
safety initiatives that do not address prioritized areas for improvement, may add little
value to the organization in its quest to improve its safety culture, and reduce costs in
the long term.
To address these issues, this study developed a model of construction safety culture, and
investigated the interactions and causal relationships between the five enablers (what
the organization should be doing) and Goals (what the organization aims to achieve),
and their consequences over time. The ‘construction safety culture index’, developed
through modelling construction safety culture, was used to measure the level of
construction safety culture maturity in the organization, and identify areas for safety
improvement.
A System Dynamics Approach to Construction Safety Culture
vi
In developing a construction safety culture model, this study adopted an internationally
recognized performance measurement system, the European Foundation for Quality
Management (EFQM) Excellence model, which consists of five key enablers
(Leadership, Policy and Strategy, People, Partnerships and Resources, and Processes),
to achieve a set of predetermined Goals. It was hypothesized that Leadership drives
(influences) the three enablers (Policy and Strategy, People, and Partnerships and
Resources), which, in turn, collectively influence the achievability of predetermined
Goals through the implementation and improvement of suitable Processes.
To confirm the above hypothesis, and to investigate the proposed relationships among
the Enablers and Goals, the statistical techniques of exploratory factor analysis and
structural equation modelling were performed. A questionnaire survey was used for data
collection purposes. The survey was sent to medium-to-large construction-contracting
organizations operating in Thailand. The targeted respondents were selected on the
assumption that they held senior appointments within their respective organizations.
The results revealed that Leadership was the main driver to effective safety culture, and
that this enabler strongly influenced Policy and Strategy, and People, but had a weak
relationship with Partnerships and Resources. Most of Leadership’s influence on
Partnerships and Resources appeared to be mediated through the People enabler. Thus,
it could be postulated that Thai construction managers focus more on human resources
and teamwork than on the provision of safety resources.
The results also showed that People and Policy and Strategy play a key role in
successful safety implementation. The cooperation of people within the organization,
and an effective safety policy and strategy, therefore, influence effective safety
implementation, which, in turn, enhances Goals achieved by the organization.
To capture the interactions among the five enablers and Goals, over a period of time
(e.g. to examine how feedback of Goals affects the implementation of the Leadership
enabler over time), system dynamics modelling was performed. The causal relationships
A System Dynamics Approach to Construction Safety Culture
vii
obtained from structural equation modelling were used to develop a construction safety
culture dynamic model; it was predicted that a higher enablers’ value would result in a
higher Goals value, and ultimately a higher construction safety culture index.
To plan for safety improvement, an organization can facilitate the testing of alternative
strategies, by simulating the developed dynamic model, to improve its construction
safety culture index, and to progress through to higher maturity levels without actually
having to implement them. This approach reduces any costs that may occur from not
implementing the best strategy.
The cyclical style of safety management was also modelled, to reflect real-life
situations, where management tends to withdraw its attention to safety following the
realisation of excellent performance record.
In conclusion, the developed construction safety culture dynamic model provides
insight into the interactions and influences that each enabler has on improving
construction safety culture over time. The construction safety culture index helps an
organization to assess how well its safety implementation is performed, and provides
guidance on how to plan for safety improvements.
A System Dynamics Approach to Construction Safety Culture
viii
TTAABBLLEE OOFF CCOONNTTEENNTTSS
DDEECCLLAARRAATTIIOONN ................................................................................................................................................................................................................................ ii
AACCKKNNOOWWLLEEDDGGEEMMEENNTT ....................................................................................................................................................................................................iiii
LLIISSTT OOFF PPUUBBLLIICCAATTIIOONNSS ............................................................................................................................................................................................ iivv
AABBSSTTRRAACCTT ...................................................................................................................................................................................................................................... vv
TTAABBLLEE OOFF CCOONNTTEENNTTSS ................................................................................................................................................................................................vviiiiii
LLIISSTT OOFF FFIIGGUURREESS ................................................................................................................................................................................................................ xxiivv
LLIISSTT OOFF TTAABBLLEESS ................................................................................................................................................................................................................ xxvviiiiii
AACCRROONNYYMMSS ................................................................................................................................................................................................................................ xxxxii
CCHHAAPPTTEERR 11
IINNTTRROODDUUCCTTIIOONN .......................................................................................................................................................................................................................... 11
1.1 GENERAL OVERVIEW .......................................................................................1
1.2 THE CONSTRUCTION INDUSTRY ...................................................................1
1.3 SAFETY CULTURE ..............................................................................................5
1.4 SAFETY CULTURE IN CONSTRUCTION ORGANIZATIONS .......................8
1.5 MEASURING SAFETY CULTURE ...................................................................10
1.5.1 Wright et al.’s Safety Culture Improvement Matrix (1999) ...................11
1.5.2 Molenaar et al.’s Characteristics of Safety Culture (2002) ....................12
1.5.3 Mohamed’s Balanced Scorecard for Benchmarking Safety Culture
(2003) .....................................................................................................14
1.6 RESEARCH NEED AND RESEARCH AIMS ...................................................15
1.7 THESIS ORGANIZATION .................................................................................16
A System Dynamics Approach to Construction Safety Culture
ix
CCHHAAPPTTEERR 22
RREESSEEAARRCCHH MMEETTHHOODDOOLLOOGGYY ............................................................................................................................................................................ 1199
2.1 GENERAL OVERVIEW .....................................................................................19
2.2 RESEARCH DESIGN AND RESEARCH FRAMEWORK ...............................19
2.2.1 Performance Measurement Systems: The Review .................................22
2.2.2 The Constructs of the EFQM Excellence Model: The Review ..............23
2.2.3 Data Collection: Questionnaire Survey ..................................................24
2.2.4 Data Screening and Preliminary Analyses .............................................28
2.2.5 Exploratory Factor Analysis: The Introduction ......................................28
2.2.6 Structural Equation Modelling: The Introduction ..................................29
2.2.7 System Dynamics Modelling: The Introduction ....................................30
2.3 SUMMARY ..........................................................................................................41
CCHHAAPPTTEERR 33
LLIITTEERRAATTUURREE RREEVVIIEEWW...................................................................................................................................................................................................... 4433
3.1 GENERAL OVERVIEW .....................................................................................43
3.2 PERFORMANCE MEASUREMENT SYSTEMS ..............................................43
3.2.1 The Malcolm Baldrige National Quality Award Framework .................44
3.2.2 The Balanced Scorecard Framework ......................................................48
3.2.3 The European Foundation for Quality Management Excellence Model 50
3.3 SELECTION OF A BASIC FRAMEWORK FOR THE CONSTRUCTION
SAFETY CULTURE ............................................................................................52
3.3.1 A Comparison between the MBNQA Framework and the EFQM
Excellence Model ....................................................................................52
3.3.2 A Comparison between the BSC Framework and the EFQM
Excellence Model ...................................................................................54
3.4 THE PROPOSED CONSTRUCTION SAFETY CULTURE MODEL ..............59
A System Dynamics Approach to Construction Safety Culture
x
3.4.1 Leadership ..............................................................................................60
3.4.2 Policy and Strategy .................................................................................62
3.4.3 People .....................................................................................................63
3.4.4 Partnerships and Resources .....................................................................65
3.4.5 Processes .................................................................................................66
3.4.6 Goals .......................................................................................................67
3.5 SAFETY CULTURE MATURITY MODEL ......................................................71
3.5.1 Safety Culture Maturity Levels ..............................................................71
3.5.2 Scoring Each Maturity Level ..................................................................74
3.6 SUMMARY ..........................................................................................................75
CCHHAAPPTTEERR 44
DDAATTAA CCOOLLLLEECCTTIIOONN AANNDD PPRREELLIIMMIINNAARRYY AANNAALLYYSSEESS ............................................................................................ 7777
4.1 GENERAL OVERVIEW .....................................................................................77
4.2 QUESTIONNAIRE SURVEY .............................................................................77
4.3 SAMPLE CHARACTERISTICS .........................................................................79
4.4 DATA SCREENING AND PRELIMINARY ANALYSES ................................84
4.4.1 Handling Missing Data ...........................................................................84
4.4.2 Test of Normality ...................................................................................86
4.4.3 Outliers Test ...........................................................................................88
4.4.4 Scale Reliability (Cronbach’s Alpha)......................................................90
CCHHAAPPTTEERR 55
EEXXPPLLOORRAATTOORRYY FFAACCTTOORR AANNAALLYYSSIISS AANNDD SSTTRRUUCCTTUURRAALL EEQQUUAATTIIOONN
MMOODDEELLLLIINNGG .................................................................................................................................................................................................................................. 9933
5.1 GENERAL OVERVIEW .....................................................................................93
A System Dynamics Approach to Construction Safety Culture
xi
5.2 THE EXPLORATORY FACTOR ANALYSIS ..................................................93
5.2.1 Assessment of the Suitability of the Data for the Analysis ....................94
5.2.2 Factor Extraction ....................................................................................95
5.2.3 Factor Rotation and Interpretation ..........................................................95
5.2.4 The EFA Results......................................................................................96
5.2.5 Conclusion of the EFA .........................................................................101
5.3 THE STRUCTURAL EQUATION MODELLING ...........................................103
5.3.1 Measurement Model .............................................................................104
5.3.2 Structural Model ...................................................................................111
5.3.3 Conclusion of the SEM ........................................................................115
CCHHAAPPTTEERR 66
SSYYSSTTEEMM DDYYNNAAMMIICCSS MMOODDEELLLLIINNGG OOFF CCOONNSSTTRRUUCCTTIIOONN SSAAFFEETTYY CCUULLTTUURREE --
MMOODDEELL BBUUIILLDDIINNGG .......................................................................................................................................................................................................... 111177
6.1 GENERAL OVERVIEW ...................................................................................117
6.2 SYSTEM DYNAMICS MODELLING .............................................................117
6.3 CAUSAL LOOP DIAGRAMS OF CONSTRUCTION SAFETY CULTURE .119
6.3.1 Causal Loop Diagram ...........................................................................119
6.3.2 A Causal Loop Diagram of the CSC Index ..........................................122
6.4 CONSTRUCTION SAFETY CULTURE DYNAMIC MODEL ......................127
6.4.1 Leadership Dynamic Model .................................................................127
6.4.2 People Dynamic Model ........................................................................130
6.4.3 Partnerships and Resources Dynamic Model .......................................132
6.4.4 Policy and Strategy Dynamic Model.....................................................133
6.4.5 Processes Dynamic Model ....................................................................134
6.4.6 Goals Dynamic Model ..........................................................................135
A System Dynamics Approach to Construction Safety Culture
xii
6.5 DYNAMIC SIMULATION RESULTS .............................................................137
6.5.1 Base Run Results ..................................................................................137
6.5.2 Base Run Results Examination ............................................................143
6.5.3 Model Verification and Validation .......................................................145
CCHHAAPPTTEERR 77
SSYYSSTTEEMM DDYYNNAAMMIICCSS MMOODDEELLLLIINNGG OOFF CCOONNSSTTRRUUCCTTIIOONN SSAAFFEETTYY CCUULLTTUURREE --
MMOODDEELL AAPPPPLLIICCAATTIIOONNSS ........................................................................................................................................................................................ 115511
7.1 GENERAL OVERVIEW ...................................................................................151
7.2 POLICY ANALYSIS .........................................................................................151
7.2.1 Base Run ...............................................................................................152
7.2.2 Policy Experiments between Organizations ‘A’ and ‘B’ .....................160
7.3 THE CYCLICAL STYLE OF SAFETY MANAGEMENT ..............................168
7.3.1 The Dynamic Model of the Cyclical Style of Safety Management ......171
7.3.2 The Simulation Results .........................................................................182
7.3.3 Conclusion of the Cyclical Style of Safety Management .....................186
CCHHAAPPTTEERR 88
SSTTUUDDYY FFIINNDDIINNGGSS AANNDD RREECCOOMMMMEENNDDAATTIIOONNSS FFOORR FFUUTTUURREE RREESSEEAARRCCHH ................ 118899
8.1 GENERAL OVERVIEW ...................................................................................189
8.2 MAJOR FINDINGS ...........................................................................................189
8.3 CONTRIBUTIONS TO THE EXISTING BODY OF KNOWLEDGE ............194
8.4 IMPLICATIONS FOR THE THAI CONSTRUCTION INDUSTRY ...............195
8.5 LIMITATIONS AND RECOMMENDATIONS FOR FUTURE RESEARCH 197
8.6 CLOSURE ..........................................................................................................199
A System Dynamics Approach to Construction Safety Culture
xiii
AAppppeennddiixx 11 QQuueessttiioonnnnaaiirree SSuurrvveeyy ...................................................................................................................................................... 220000
AAppppeennddiixx 22 RRaaww DDaattaa ............................................................................................................................................................................................ 221111
AAppppeennddiixx 33 SSttaannddaarrddiizzeedd SSccoorreess ((ZZ--SSccoorreess)) ........................................................................................................................ 222200
AAppppeennddiixx 44 MMeeaassuurreemmeenntt MMooddeell RReessuullttss.................................................................................................................................... 222299
AAppppeennddiixx 55 SSttrruuccttuurraall MMooddeell RReessuullttss .............................................................................................................................................. 224411
AAppppeennddiixx 66 SSDD EEqquuaattiioonnss ooff BBaassee RRuunn SSiimmuullaattiioonn .................................................................................................... 225522
AAppppeennddiixx 77 LLiinneeaarr RReeggrreessssiioonn RReessuullttss .......................................................................................................................................... 225577
AAppppeennddiixx 88 SSeennssiittiivviittyy AAnnaallyysseess RReessuullttss wwhheenn tthhee IInniittiiaall VVaalluuee ooff eeaacchh EEnnaabblleerr
iiss CChhaannggeedd .......................................................................................................................................................................................... 226600
AAppppeennddiixx 99 SSeennssiittiivviittyy AAnnaallyysseess RReessuullttss wwhheenn tthhee EExxttrraa EEffffoorrtt GGiivveenn ttoo IImmpprroovvee
eeaacchh EEnnaabblleerr iiss CChhaannggeedd.............................................................................................................................................. 226655
AAppppeennddiixx 1100 SSDD EEqquuaattiioonnss ooff OOrrggaanniizzaattiioonn ‘‘AA’’.................................................................................................................. 227700
AAppppeennddiixx 1111 SSDD EEqquuaattiioonnss ooff OOrrggaanniizzaattiioonn ‘‘BB’’.................................................................................................................. 227755
AAppppeennddiixx 1122 SSDD EEqquuaattiioonnss ooff tthhee CCyycclliiccaall SSttyyllee ooff SSaaffeettyy MMaannaaggeemmeenntt .................................... 228800
RREEFFEERREENNCCEESS ............................................................................................................................................................................................................................ 228855
A System Dynamics Approach to Construction Safety Culture
xiv
LLIISSTT OOFF FFIIGGUURREESS
Figure 1.1 Traditional concept of culture (Clarke, 1999) ...........................................7
Figure 1.2 Safety culture table model (Ho and Zeta, 2004)........................................9
Figure 1.3 Highest levels of safety culture hierarchy (Molenaar et al., 2002)..........12
Figure 1.4 The people branch (Molenaar et al., 2002)..............................................12
Figure 1.5 The process branch (Molenaar et al., 2002) ............................................13
Figure 1.6 The values branch (Molenaar et al., 2002) ..............................................13
Figure 1.7 Safety culture balanced scorecard (Mohamed, 2003) .............................14
Figure 2.1 Research design .......................................................................................20
Figure 2.2 Research activities and expected outputs ................................................21
Figure 2.3 Construction safety literature
(Adapted from Melville and Goddard, 1996)..........................................24
Figure 2.4 Methods of data collection (Adapted from Kumar, 2005) ......................25
Figure 2.5 AMOS window and its drawing tool icons .............................................30
Figure 2.6 STELLA menus of symbols for creating a model (Ithink, 2003)............36
Figure 3.1 Four basic elements of the MBNQA framework (NIST, 1993) ..............47
Figure 3.2 Four key perspectives of the BSC framework
(Lamotte and Carter, 2000) .....................................................................48
Figure 3.3 The EFQM Excellence model (EFQM, 2000).........................................51
Figure 3.4 Mapping the BSC framework onto the EFQM Excellence model
(Sheffield Hallam University, 2003) .......................................................57
Figure 3.5 Links between the safety management system and the EFQM
Excellence model (Adapted from Mbuya and Lema, 2004) ...................58
Figure 3.6 The proposed CSC model........................................................................59
A System Dynamics Approach to Construction Safety Culture
xv
Figure 3.7 Safety culture maturity model (Lardner et al., 2001) ..............................72
Figure 4.1 Years of experience in the Thai construction industry............................79
Figure 4.2 Years of experience in the present organization......................................80
Figure 4.3 Job titles of the respondents ....................................................................80
Figure 4.4 Safety responsibilities..............................................................................81
Figure 4.5 Safety activities engagement ...................................................................81
Figure 4.6 Formal safety policy in the organization .................................................82
Figure 4.7 Safety performance compared to the national average record ................82
Figure 4.8 The most influential enablers in improving safety culture ......................83
Figure 5.1 Baseline model of the CSC....................................................................102
Figure 5.2 The best-fit measurement model ...........................................................109
Figure 5.3 The best-fit structural model..................................................................112
Figure 5.4 The final CSC model .............................................................................113
Figure 6.1 Basic components of a SD model..........................................................118
Figure 6.2 Basic CSC dynamic model ....................................................................119
Figure 6.3 An example of a causal loop diagram ...................................................120
Figure 6.4 A causal loop diagram of the CSC index ..............................................123
Figure 6.5 A causal loop diagram of the five enablers and Goals ..........................126
Figure 6.6 The CSC dynamic model.......................................................................128
Figure 6.7 Leadership dynamic model....................................................................129
Figure 6.8 People dynamic model ..........................................................................131
Figure 6.9 Partnerships and Resources dynamic model .........................................132
Figure 6.10 Policy and Strategy dynamic model ......................................................133
Figure 6.11 Processes dynamic model......................................................................135
Figure 6.12 Goals dynamic model ............................................................................136
Figure 6.13 The CSC index dynamic model.............................................................137
A System Dynamics Approach to Construction Safety Culture
xvi
Figure 6.14 Graphical results of the Enablers score over time.................................138
Figure 6.15 Graphical results of the Goals score over time......................................139
Figure 6.16 Graphical results of the CSC index over time .......................................139
Figure 6.17 Graphical results of the five enablers over time ....................................143
Figure 6.18 Sensitivity results of the ‘used_lds’ value when its initial value is
changed..................................................................................................146
Figure 6.19 Sensitivity results of the CSC index when the initial value of Lds is
changed..................................................................................................147
Figure 6.20 Sensitivity results of the ‘used_lds’ value when the ‘plds’ value is
changed..................................................................................................148
Figure 6.21 Sensitivity results of the CSC index when the ‘plds’ value is changed 148
Figure 7.1 Graphical results of the five enablers of organization ‘A’ over time ....154
Figure 7.2 Graphical results of the Enablers score of organization ‘A’ over time.155
Figure 7.3 Graphical results of the Goals score of organization ‘A’ over time......155
Figure 7.4 Graphical results of the CSC index of organization ‘A’ over time .......156
Figure 7.5 Graphical results of the five enablers of organization ‘B’ over time ....158
Figure 7.6 Graphical results of the Enablers score of organization ‘B’ over time.159
Figure 7.7 Graphical results of the Goals score of organization ‘B’ over time......159
Figure 7.8 Graphical results of the CSC index of organization ‘B’ over time .......160
Figure 7.9 The accident cycle
(Adapted from NPS Risk Management Division, 2006).......................169
Figure 7.10 The normal accident cycle (Adapted from Jones, 2007) .......................170
Figure 7.11 The CSC index cycle as management withdraws attention to safety....172
Figure 7.12 The dynamic model of the cyclical style of safety management...........173
Figure 7.13 Leadership dynamic model....................................................................174
Figure 7.14 People dynamic model ..........................................................................176
A System Dynamics Approach to Construction Safety Culture
xvii
Figure 7.15 Partnerships and Resources dynamic model .........................................177
Figure 7.16 Policy and Strategy dynamic model ......................................................178
Figure 7.17 Processes dynamic model......................................................................179
Figure 7.18 Goals dynamic model ............................................................................180
Figure 7.19 Graphical results of the Enablers score as the effect of the attention
withdrawal .............................................................................................184
Figure 7.20 Graphical results of the Goals score as the effect of the attention
withdrawal .............................................................................................184
Figure 7.21 Graphical results of the CSC index as the effect of the attention
withdrawal .............................................................................................185
A System Dynamics Approach to Construction Safety Culture
xviii
LLIISSTT OOFF TTAABBLLEESS
Table 1.1 Definitions of safety culture (Adapted from Potter, 2003)........................6
Table 1.2 Safety culture components (Adapted from Flannery, 2001) ....................8
Table 1.3 Explanation of the scoring scale (Wright et al., 1999) ............................11
Table 2.1 Flow diagram model conventions (Morecroft, 1988)..............................36
Table 2.2 Tests of model structure (Forrester and Senge, 1980).............................38
Table 2.3 Tests of model behaviour (Forrester and Senge, 1980)...........................39
Table 2.4 Tests of policy implications (Forrester and Senge, 1980) .......................40
Table 3.1 The MBNQA categories and items (Pannirselvam and Ferguson, 2001)46
Table 3.2 An example of a BSC template (Lamotte and Carter, 2000)...................49
Table 3.3 Similarities and differences of the MBNQA framework and the
EFQM Excellence model.........................................................................53
Table 3.4 Comparison between core concepts, the MBNQA framework, and the
EFQM Excellence model (Tummala and Tang, 1995)............................53
Table 3.5 Comparison between the BSC framework and the EFQM Excellence
model (Otley, 1999).................................................................................55
Table 3.6 High-level comparisons of the BSC framework and the EFQM
Excellence model (Lamotte and Carter, 2000)........................................56
Table 3.7 Six model constructs and their 34 attributes............................................70
Table 4.1 Missing values .........................................................................................85
Table 4.2 Skewness and kurtosis of the 34 attributes..............................................87
Table 4.3 The mean, the 5% trimmed mean, and the standard deviation of
the 34attributes ........................................................................................89
Table 4.4 Internal consistency of the five enablers and Goals ................................91
Table 5.1 Bartlett’s test of sphericity and the KMO index......................................94
A System Dynamics Approach to Construction Safety Culture
xix
Table 5.2 Three factors extracted from the remaining 25 items..............................97
Table 5.3 Two factors extracted from nine items of Factor 1 of Table 5.2 .............98
Table 5.4 Two factors extracted from nine items of Factor 2 of Table 5.2 .............99
Table 5.5 Five factors extracted from the EFA .....................................................100
Table 5.6 Internal consistency of five factors extracted from the EFA.................101
Table 5.7 The GOF indices of the baseline and the best-fit measurement models107
Table 5.8 Square multiple correlations and standardized coefficients of
observed variables .................................................................................110
Table 5.9 The GOF indices of the best-fit structural model ..................................113
Table 5.10 The direct and indirect path coefficients between the five enablers
and Goals ...............................................................................................115
Table 6.1 Simulation results of the five enablers and Goals .................................140
Table 6.2 Simulation results of the enablers, Goals, and CSC index....................141
Table 6.3 Experimentation with extra efforts given to improve the five enablers 145
Table 7.1 Simulation results of the five enablers of organization ‘A’ ..................153
Table 7.2 Simulation results of the Enablers, Goals, and CSC index of
organization ‘A’.....................................................................................154
Table 7.3 Simulation results of the five enablers of organization ‘B’...................157
Table 7.4 Simulation results of the Enablers, Goals, and CSC index of
organization ‘B’.....................................................................................158
Table 7.5 Simulation results of the five enablers of organization ‘A’ with
‘plds’ = 0.1.............................................................................................162
Table 7.6 Simulation results of the Enablers, Goals, and CSC index of
organization ‘A’ with ‘plds’ = 0.1 .........................................................162
Table 7.7 Simulation results of the five enablers of organization ‘A’ with
‘plds’ = 0.2.............................................................................................163
A System Dynamics Approach to Construction Safety Culture
xx
Table 7.8 Simulation results of the Enablers, Goals, and CSC index of
organization ‘A’ with ‘plds’ = 0.2 .........................................................164
Table 7.9 Simulation results of organization ‘A’ with ‘plds’, ‘pppl’, and
‘ppro’ = 0.2 ............................................................................................165
Table 7.10 Simulation results of organization ‘A’ with ‘plds’ = 0.3, and ‘pppl’
and ‘ppro’ = 0.1 .....................................................................................166
Table 7.11 Simulation results of organization ‘A’ with ‘plds’ and ‘ppro’ = 0.2,
and ‘pppl’, ‘pprs’, and ‘ppol’ = 0.1 .......................................................166
Table 7.12 The CSC index of organization ‘B’ when more of effort is given to
enhance each enabler .............................................................................168
Table 7.13 Simulation results of the cyclical style of safety management..............183
A System Dynamics Approach to Construction Safety Culture
xxi
AACCRROONNYYMMSS
ACSNI Advisory Committee on the Safety of Nuclear Installation
AMOS Analysis of Moment Structure
ANZSIC Australian and New Zealand Standard Industrial Classification
BSC Balanced Scorecard
CFA Confirmatory Factor Analysis
CFI Comparative Fit Index
Co_lds_pol Correlation Value between Leadership, and Policy and Strategy
Co_lds_ppl Correlation Value between Leadership and People
Co_lds_prs Correlation Value between Leadership, and Partnerships and Resources
Co_pol_pro Correlation Value between Policy and Strategy, and Processes
Co_ppl_pro Correlation Value between People and Processes
Co_ppl_prs Correlation Value between People, and Partnerships and Resources
Co_pro_goals Correlation Value between Processes and Goals
Co_prs_pol Correlation Value between Partnerships and Resources, and Policy and
Strategy
C.R. Critical Ratio
CSC Construction Safety Culture
dCSC_Index Desired Construction Safety Culture Index
DF Degree of Freedom
DF_goals_pro Decision Fraction between Goals and Processes
DF_pol_lds Decision Fraction between Policy and Strategy, and Leadership
DF_pol_prs Decision Fraction between Policy and Strategy, and Partnerships and
Resources
DF_ppl_lds Decision Fraction between People and Leadership
A System Dynamics Approach to Construction Safety Culture
xxii
DF_pro_pol Decision Fraction between Processes, and Policy and Strategy
DF_pro_ppl Decision Fraction between Processes and People
DF_prs_lds Decision Fraction between Partnerships and Resources, and Leadership
DF_prs_ppl Decision Fraction between Partnerships and Resources, and People
dgoals Desired Goals Value
dlds Desired Leadership Value
dpol Desired Policy and Strategy Value
dppl Desired People Value
dpro Desired Processes Value
dprs Desired Partnerships and Resources Value
DYNAMO Dynamic Models
EFA Exploratory Factor Analysis
EFQM European Foundation for Quality Management
ggoals Gap of Goals
glds Gap of Leadership
GOF Goodness of Fit
gpol Gap of Policy and Strategy
gppl Gap of People
gpro Gap of Processes
gprs Gap of Partnerships and Resources
IFI Incremental Fit Index
INIT Initial Value
KMO Kaiser-Meyer-Olkin
Lds Leadership
MBNQA Malcolm Baldrige National Quality Award
NFI Bentler-Bonett Normed Fit Index
A System Dynamics Approach to Construction Safety Culture
xxiii
NIST National Institute of Standards and Technology
plds Percentage of More Effort Provided to Improve Leadership
Pol Policy and Strategy
Ppl People
ppol Percentage of More Effort Provided to Improve Policy and Strategy
pppl Percentage of More Effort Provided to Improve People
ppro Percentage of More Effort Provided to Improve Processes
pprs Percentage of More Effort Provided to Improve Partnerships and
Resources
Pro Processes
Prs Partnerships and Resources
RFI Relative Fit Index
rgoals Goals Rate
rlds Leadership Rate
rldsf Leadership Rate Fraction
RMSEA Root Mean Square Error of Approximation
rpol Policy and Strategy Rate
rppl People Rate
rpro Processes Rate
rprs Partnerships and Resources Rate
SD System Dynamics
S.E. Standard Error
SEM Structural Equation Modelling
SMC, R2 Square Multiple Correlation
SMS Safety Management System
SPICE Standardized Process Improvement for Construction Enterprises
A System Dynamics Approach to Construction Safety Culture
xxiv
SPSS Statistical Package for Social Science
Stat. Statistics Value
STELLA Structural Thinking Experimental Learning Laboratory with Animation
TLI Tucker-Lewis Index
used_goals Goals Value
used_lds Leadership Value
used_pol Policy and Strategy Value
used_ppl People Value
used_pro Processes Value
used_prs Partnerships and Resources Value
�2 Chi Square
zacci Z-Score of the ‘Number of Accidents’ Item
zaccn Z-Score of the ‘Accountability’ Item
zalgn Z-Score of the ‘Safety and Productivity Alignment’ Item
zawrn Z-Score of the ‘Safety Awareness’ Item
zbnmk Z-Score of the ‘Benchmarking System’ Item
zcmmt Z-Score of the ‘Commitment’ Item
zcomm Z-Score of the ‘Communication’ Item
zcoop Z-Score of the ‘Stakeholders’ Cooperation’ Item
zcost Z-Score of the ‘Cost of Accidents’ Item
zcstm Z-Score of the ‘Customers’ Expectations’ Item
zdocu Z-Score of the ‘Safety Documentation’ Item
zfdbk Z-Score of the ‘Feedback’ Item
zfinc Z-Score of the ‘Financial Resources’ Item
zhmnr Z-Score of the ‘Human Resources’ Item
zhskp Z-Score of the ‘Housekeeping’ Item
A System Dynamics Approach to Construction Safety Culture
xxv
zimge Z-Score of the ‘Industrial Image’ Item
zinit Z-Score of the ‘Safety Initiatives’ Item
zintg Z-Score of the ‘Safety Integration in Business Goals’ Item
zinvm Z-Score of the ‘Workers’ Involvement’ Item
zjstf Z-Score of the ‘Job Satisfaction’ Item
zldbx Z-Score of the ‘Leading by Example’ Item
zmrle Z-Score of the ‘Workforce Morale’ Item
znobm Z-Score of the ‘No-Blame Approach’ Item
zprcp Z-Score of the ‘Shared Perceptions’ Item
zprsp Z-Score of the ‘Work Pressure’ Item
zresc Z-Score of the ‘Safety Resources’ Item
zresp Z-Score of the ‘Safety Responsibilities’ Item
zrisk Z-Score of the ‘Risk Assessment’ Item
zrlsp Z-Score of the ‘Workers’ Relationships’ Item
zsppt Z-Score of the ‘Supportive Environment’ Item
zstnd Z-Score of the ‘Safety Standards’ Item
zswbh Z-Score of the ‘Safe Work Behaviour’ Item
ztrng Z-Score of the ‘Training’ Item
zwkld Z-Score of the ‘Workload’ Item
A System Dynamics Approach to Construction Safety Culture
xxvi
[THIS PAGE IS INTENTIONALLY LEFT BLANK]
A System Dynamics Approach to Construction Safety Culture
1
11 IINNTTRROODDUUCCTTIIOONN
1.1 GENERAL OVERVIEW
Chapter 1 outlines the background to this study. It provides descriptions and
characteristics of the construction industry, and definitions of safety culture, as well as
previous research undertaken to measure safety culture. The research need, research
aims, and thesis organization are also presented.
1.2 THE CONSTRUCTION INDUSTRY
The Australian and New Zealand Standard Industrial Classification (ANZSIC) system
defines the construction industry as “units mainly engaged in construction, repair,
alteration, and renovation of buildings and other structures, and those engaged in
providing building or construction trade services and specific installation activities”
(Australian Bureau of Statistics, 1993).
The construction industry comprises many organizations, including property developers,
architects, engineers, quantity surveyors, accountants, lawyers, contractors,
subcontractors, labourers, and specialist tradespersons. It operates on international,
national, regional, and local scales, with participants ranging from large multinational
organizations to single person operations.
According to Jaafari (1996), the construction industry is different from the
manufacturing industry due to: 1) its fragmented structure; 2) its diffused responsibility;
3) its prototype nature; 4) its influences of public, regulatory agencies, and interest
groups; 5) its transient and itinerant labour force, which is not trained to operate under
A System Dynamics Approach to Construction Safety Culture
2
the quality assurance mode of construction; and 6) its virtual lack of research and
development. These factors are explained below:
� Its fragmented structure: The bulk of construction businesses are normally
generated by a large number of small firms that are less inclined to formal methods
of work and management.
� Its diffused responsibility: In construction projects, many individual professionals
and firms share responsibility for the specification, design, and construction of the
projects.
� Its prototype nature: Construction projects typically resemble the ‘prototype’
products in the manufacturing industry, carrying unique design features, site
characteristics, and functions. Thus, the potential for errors to creep in is always
presented due to the once-off nature of the relevant activities and production
processes.
� Its influences of public, regulatory agencies and interest groups: These influences
will ultimately affect the functions and configurations of the projects, which include
construction methods and associated safeguards to the environment, third party
issues, and beneficiaries.
� Its transient and itinerant labour force, which is not trained to operate under the
quality assurance mode of construction: The training of skilled labour is generally
based on learning how to do the work, not being one’s own inspector to produce
zero defects.
� Its virtual lack of research and development (R&D): Typically, R&D work in
construction is confined to that undertaken by the manufacturers of materials and
components incorporated into the projects. There is little R&D work on lines of
projects, such as commercial buildings as a ‘product line’, or managerial processes
in infrastructure works, etc.
A construction project is a unique task, has a predetermined date of delivery, is subject
to one or several performance goals (such as resource usage and quality), and consists
of a number of complex and/or interdependent activities (Packendorff, 1995). The
projects may vary from simple dwellings to complex structures, and normally involve
A System Dynamics Approach to Construction Safety Culture
3
many changes, such as frequent teamwork rotations, exposure to weather conditions,
and high rates of unskilled workers (Rosenfeld et al., 2006). To better understand the
construction industry, it is critical to understand its unique characteristics. Maloney
(2003) listed construction organization characteristics, namely:
� Construction projects in general, with a relatively short and finite duration, cause
project team stress by focusing on what has to be accomplished.
� Time pressure dictates management decisions and project payment, so delays are not
tolerated.
� Cost, work schedule, and productivity determine project profit.
� Supervision provides explicit direction as to what is to be accomplished, while
individuals performing the work may determine how the work is to be done.
� There is extensive goal clarification performance, i.e. planning, and scheduling.
� Construction organizations have significant differences in complexity, diversity, and
size, and often being temporary, multi-organizations, with as many as 20 different
contractors working on a single project site at one time.
� The principal or prime contractor (the contractor with whom the client has a
contract) subcontracts approximately 80% of the work to specialty contractors.
� A single subcontractor lacks control of the working environment.
� The operative employment of contractor organizations on a specific construction site
may be short and transient because workers are added to, and released from, crews
in response to project schedules.
These characteristics make the construction industry one of the most hazardous
industries, resulting in high rates of severe and fatal work-related accidents (Maloney,
2003). In the United Kingdom (UK), for example, the industry accounts for one third of
all work-related fatalities and, on average, five construction workers are killed every
two weeks, while one member of the public is killed every month by construction
activities (HSC, 2003). In the United States of America (USA), although construction
jobs account for just 5% of the total workforce, they account for more than 17% of
annual workplace deaths (Goetsch, 2003). In Australia, the construction fatality rate in
A System Dynamics Approach to Construction Safety Culture
4
2001-02 was five per 100,000 employees, which made it double the all-industry average
(NOSHC, 2005).
Emerging economies and less developed countries are no exception to high fatality
rates. In Thailand, for example, the rate of accidents and fatalities in the construction
industry is reported as highest of all industries (International Labour Organization,
2005). Additionally, construction workers are five times more likely to suffer a
permanent disability than are those in other industries. In India, one of the world’s fast
growing economies, the construction industry accounts for a major share of work-
related accidents (Damodaran, 2006). A similar trend is experienced in the international
construction ‘hotspot’ – the United Arab Emirates – where construction accidents
dominate work-related accident records (The UAE Ministry of Labour and Social
Affairs, 2001).
Construction accidents cause many human tragedies, de-motivate workers, disrupt site
activities, delay project progress, and adversely affect the overall cost, productivity, and
reputation of the construction industry (Mohamed, 1999). According to Kartam (1997),
construction accidents may arise from a variety of causes, which can generally be
classified as: 1) physical incidents posing hazardous situations; and 2) behavioural
incidents caused by unsafe acts. The latter has been identified as the main cause of
construction accidents (Sawacha et al., 1999), and is viewed by many as the direct result
of having a poor safety culture (Smith and Roth, 1991).
Since poor safety culture can lead to risks to human lives, much attention has been paid,
over the past few years, to organizational safety culture, especially to its definitions,
dimensions, and enablers, as well as to the development of tools for assessing and
monitoring its ‘health’, in order to identify areas for safety performance improvement.
The establishment of a good culture of safety can undoubtedly help organizations to
control and reduce their construction costs, and increase the efficiency of their
operations in the long term (Fung et al., 2005). The next section (Section 1.3) presents
an overview of previous research into safety culture.
A System Dynamics Approach to Construction Safety Culture
5
1.3 SAFETY CULTURE
The term safety culture was first introduced by the International Atomic Energy Agency
(IAEA) following their analysis into the nuclear reactor accident at Chernobyl, Ukraine,
in 1986 (Gadd and Collins, 2002). The identification of a poor safety culture, as a
contributing factor to this accident, led to a large number of studies investigating and
attempting to measure safety culture in a variety of high-hazard industries (Little, 2002).
No single definition of what constitutes a safety culture exists. However, the majority of
research studies commonly describe it as including norms, rules, and behaviours that are
presented with respect to safety, as well as characteristics, beliefs, and values that are
exhibited (Potter, 2003). One of the most widely used safety culture definitions is that
developed by the Advisory Committee on the Safety of Nuclear Installation (ACSNI,
1993). This broad-based definition was based on the findings of a study group on
human factors, and had been adopted for use in this study (see below). Other safety
culture definitions are listed in Table 1.1.
Safety culture is the product of individual and group values, attitudes, perceptions, competencies, and patterns of behaviour that determine the commitment to, and the style and proficiency of, an organization’ s health and safety management. Organizations with a positive safety culture are characterized by communications found on mutual trust, shared perceptions of the importance of safety, and confidences in the efficacy of preventive measures (ACSNI, 1993).
A System Dynamics Approach to Construction Safety Culture
6
Table 1.1 Definitions of safety culture (Adapted from Potter, 2003)
Definition Source
“That observable degree of effort by which all organizational members
direct their attentions/actions towards improving safety on a daily basis.”
Cooper, 2000
“Those aspects of the organizational culture which will impact on attitudes
and behaviours related to increasing or decreasing risk.”
Guldenmund, 2000
“The attitudes, beliefs, and perceptions shared by natural groups as
defining norms and values, which determine how they act and react in
relation to risks and risk control systems.”
Hale, 2000
“The involving perceptions and attitudes, as well as the behaviour of
individuals within an organization.”
Harvey et al., 2002
“The ideas and beliefs that all members of the organization share about
risk, accidents, and ill health.”
Cooper, 2002
“An environmental setting where everyone feels responsible for safety,
and pursues it on a daily basis, going beyond ‘the call of duty’ to identify
unsafe conditions and behaviours, and intervene to correct them... people
‘actively care’ on a continuous basis for safety... (which) is not a priority
that can be shifted depending on situational demands, rather safety is a
value linked with all other situational priorities.”
Geller, 2001
Safety culture is made up of a collection of individual cultures and other subcultures
within the environmental constraints and promotions of an organization (see Figure
1.1). Developing a safety culture needs the cooperation of both individuals and groups
in an organization. It should not be viewed as a separate process, but one that forms an
integrative part of the wider organizational culture (Clarke, 1999).
A System Dynamics Approach to Construction Safety Culture
7
Individual
Culture
Vocational Culture
Organizational Culture
National Culture
Safety Culture
Defined or Manifested
Extrinsic Elements
Intrinsic Elements
Values
Sum
Norms
Beliefs
Assumptions
Rituals
Symbols
Behaviours
Group
Figure 1.1 Traditional concept of culture (Clarke, 1999)
Many researchers have investigated the components of safety culture, finding
similarities between them. However, there appears to be a moderate difference of
opinion about their compositions. Table 1.2 compares four studies of safety culture
components. Management commitment, communication, and training are critical
components in the development of a good safety culture. Organizations thus need to
concentrate on these factors to improve their safety culture.
A System Dynamics Approach to Construction Safety Culture
8
Table 1.2 Safety culture components (Adapted from Flannery, 2001)
Components Zohar (1980) ICAO (1992) Lardner et al. (2000) INEEL (2004)
Management
commitment to safety*
� � � �
Two-way
communication
� � �
Job satisfaction � � �
Safety training* � � � �
Housekeeping � � �
Learning organization � �
Workers’ involvement � �
Shared perceptions
about safety
� �
Safety resources � �
Personal accountability � �
Workable and realistic
safety rules
� � �
Note: * Common to all four studies
1.4 SAFETY CULTURE IN CONSTRUCTION ORGANIZATIONS
Blockley (1995) proposed that positive changes in safety in the construction industry
will not be fully effective until safety culture is improved. A better understanding of
safety culture will help construction organizations to strategically allocate safety
resources, and thus improve their overall occupational health and safety performance on
sites.
Recently, many research studies have been undertaken in the area of construction safety
culture (CSC). Kartam et al. (2000), for example, studied issues, procedures, and
A System Dynamics Approach to Construction Safety Culture
9
problems of construction safety in Kuwait, and concluded that safety culture
improvement, especially in areas such as management training and commitment in
safety, was needed to prevent construction injuries and accidents. Little (2002)
identified 14 key elements across four themes (ownership and commitment, systems and
procedures, training and competence, and communication) to improve safety culture in
the UK construction industry. He stated that leadership ownership and commitment is
the bedrock upon which improvements in construction safety can be built, and that
commitment must be visible to the workers through, for example, senior management
health and safety tours.
�
Ho and Zeta (2004) studied safety in the Hong Kong construction industry, and
established four key cultural factors (environment, behaviour, organization, and person)
that affect the CSC. They concluded that safety culture and its significance vary from
one country to another due to cultural differences. Indeed workers may behave
differently due to their background differences (race, nation, religion, and community).
According to Hofstede (1980), the cultural diffences are based on five value
dimensions; power distance, uncertainty avoidance, individualism versus collectivism,
masculinity versus femininity, and long-term versus short-term focus. For this reason, a
‘safety culture table model’ was developed (see Figure 1.2). It consists of four main
CSC factors (environment, behaviour, organization, and person). It was claimed that an
organization’s safety culture will fail if it lacks the support of any of these factors.
Figure 1.2 Safety culture table model (Ho and Zeta, 2004)
A System Dynamics Approach to Construction Safety Culture
10
Fung et al. (2005) compared safety culture divergences among three levels of
construction personnel (top management, supervisors, and frontline workers). They
revealed that top management and supervisors have significant safety culture
divergences from frontline workers, especially in areas of organizational commitment,
communication, and the reporting of accidents and near misses. The researchers
proposed that managers and supervisors launch safety promotional campaign to raise
safety awareness. Further, open communication was recommended as a way to decrease
safety culture divergences among these three groups.
1.5 MEASURING SAFETY CULTURE
Such research studies, discussed above, highlight the importance of safety culture in the
construction industry. It is apparent that the enhancement of safety culture helps
organizations to reduce the number of accidents, improve the industry’s image, and
enhance safety performance (Kartam et al., 2000; Tang et al., 2003; Teo et al., 2005).
Before attempting to improve safety culture, organizations need to measure their current
safety culture implementation, and then plan for safety improvements. Unfortunately, as
Speirs and Johnson (2002) attested, safety culture is difficult to measure because it is
not a product or an outcome, but rather a process. Consequently, it is necessary to
distinguish between the outcomes of safety management and the process by which it is
acquired. This, however, is not an easy task to perform. Indeed Flin and Mearns (1999)
questioned whether the ‘state of safety’ could truly be measured in organizations. They
argued that safety attitude surveys, and other such quantitative methods of gauging
safety, are descriptive rather than normative. Thus, they saw the need for qualitative
research that includes researchers spending time in organizations they are studying, to
get a feel for the culture, and to understand how people interact throughout
organizations.
Despite the above arguments, a brief summary of the main studies attempting to
measure, benchmark, and present the ‘aggregate’ scores as an indicator of the ‘health’
of organizational safety culture, is presented below.
A System Dynamics Approach to Construction Safety Culture
11
1.5.1 Wright et al.’s Safety Culture Improvement Matrix (1999)
Wright et al. (1999) developed a so-called ‘safety culture improvement matrix’ based on
an internationally recognized business model, the European Foundation for Quality
Management (EFQM) Excellence model. This matrix, which is a self-assessment tool,
may be ‘scored’ in two methods. The first method simply entails a judgment of whether
each and every criterion of the matrix’s nine criteria (leadership, policy and strategy,
people management, resources, processes, customer satisfaction, employee satisfaction,
impact on society, and behavioural results) has been satisfied. This result is obtained by
asking top management to what extent the particular element has been satisfied.
Elements that are wholly satisfied can be coloured green, partly satisfied yellow, and
unsatisfied red.
The second method entails scoring individual questions (from zero to 100 points, based
on the judgment of the assessors), and calculates both specific element and overall
scores. Once a weighted score for each element is obtained, it is added together to
compute the total safety culture score. An explanation of the scoring scale is presented
in Table 1.3.
Table 1.3 Explanation of the scoring scale (Wright et al., 1999)
Score Explanation
0 None or minimal anecdotal evidence of activity on this point
25 Some evidence of activity on this point, such as some aspects of a results element are
measured, some examples of good leadership communications, etc.
50 Evidence of soundly based approach that delivers about half of the examples cited, but
overlooks some important points
75 Evidence of refinement and improvement of activity; results show sustained high
standards of performance and good integration of activity into normal operations and
planning
100 Systematic and refined activity has been totally integrated into normal working
patterns (but is still visible), with results showing that the organization is the best in its
class, and evidence that the activity will be sustained over time. Excellent comparisons
with internal targets
A System Dynamics Approach to Construction Safety Culture
12
1.5.2 Molenaar et al.’s Characteristics of Safety Culture (2002)
Molenaar et al. (2002) identified a total of 31 characteristics of a positive safety culture,
based on a number of construction organizations with outstanding safety records. These
characteristics were grouped into three branches: people, process, and values (see
Figure 1.3). There are 13 characteristics in the ‘people’ branch (see Figure 1.4), 11
characteristics in the ‘process’ branch (see Figure 1.5), and seven characteristics in the
‘values’ branch (see Figure 1.6).
Figure 1.3 Highest levels of safety culture hierarchy (Molenaar et al., 2002)
Figure 1.4 The people branch (Molenaar et al., 2002)
Safety Culture
People Process Values
People
Top Management Field Personnel Subcontractor
Importance
Initiate
Communication
Training
Accountability
Importance Empowerment Safety Personnel Pre-construction
Past Performance
Incentive
Attendance
Importance
A System Dynamics Approach to Construction Safety Culture
13
�
Figure 1.5 The process branch (Molenaar et al., 2002)
Figure 1.6 The values branch (Molenaar et al., 2002)
Within each branch, the characteristics were organized into a hierarchical structure, with
quantifiable questions to operationally measure the characteristics. All the questions are
based on previously proven research, and the results serve as a ‘snap-shot’ assessment
of organizational safety culture.
Process
Safety Plan Training and Education Disincentives
Involvement
Change
Feedback Change
Enforcement
Duration
Consistency
Dedicated Time Effectiveness
Assessment and Change Incentives
Regularity Value
Values
Safety Values Behaviour-Based Safety
Importance
Actions
Responsibility
Length of Employment
Identification and Correction
Participation
Hazard Prevention
A System Dynamics Approach to Construction Safety Culture
14
1.5.3 Mohamed’s Balanced Scorecard for Benchmarking Safety Culture (2003)
Mohamed (2003) adopted a Balanced Scorecard tool (as shown in Figure 1.7) to
benchmark organizational safety culture. He argued that this tool has the potential to
provide a medium by which to translate safety plans and processes into a clear set of
goals, which are, in turn, translated into a system of performance measures. The tool
offers the advantage of providing a mix of objective and subjective performance
measures that can effectively communicate a powerful strategic focus on safety to the
entire organization. It is also conducive to organizational learning by providing
feedback on targets of performance measures that have not been achieved. Further, it
has a number of different, but complementary, perspectives that help enable
organizations to pursue incremental safety performance improvements.
Figure 1.7 Safety culture balanced scorecard (Mohamed, 2003)
Customer Perspective
Goals Measures
Operational Perspective
Goals Measures
Learning Perspective
Goals Measures
Management Perspective
Goals Measures
What must we do to ensure efficient implementation of rules and procedures?
What must management excel at to achieve zero-accident culture?
How do our employees/ project partners/clients see us?
How do we continue to learn and improve?
A System Dynamics Approach to Construction Safety Culture
15
1.6 RESEARCH NEED AND RESEARCH AIMS
The variety of tools, briefly presented above, are an indication of how researches are
rapidly progressing towards the development of a reliable and valid instrument to
measure organizational safety culture. A major shortcoming with these tools, though, is
the inability to appropriately capture and present causal links between what the
organization is doing and what it aims to achieve (in this study called the Enablers and
Goals, respectively). Another element of weakness lies in a lack of understanding about
the interactions among different organizational safety culture enablers, as well as the
extent of their individual, or combined, effects on the organization’s ability to achieve
safety performance improvements. For example, a safety management system, which is
a complex interactive set of enablers, may not function according to what had been
originally planned and predicted for a variety of reasons, e.g. the difference in
perception of safety culture that has the potential to determine how successful the
process of system implementation is (Dedobbeleer and Beland, 1991).
There has also been little examination of the extent to which there is a consensus among
workers and managers regarding the contributions of the identified enablers in
determining perceptions of safety culture. Based on the SPICE (Standardized Process
Improvement for Construction Enterprises) process improvement framework, it is easy
to argue that implementing safety initiatives that are not addressing prioritized areas for
improvement may add little value to the organization in its quest to improve its safety
culture (Sarshar et al., 2000). In other words, organizations should realistically assess
their organizational safety culture maturity level, and progress sequentially through
different levels of cultural maturity.
In summary, then, there is a need to examine the interactions and interrelationships
among the CSC enablers, so that construction organizations are able to better
understand the influences of enablers on safety culture. To meet the need, this study
aims to contribute a greater understanding of the CSC enablers to the construction
industry. The aims are:
A System Dynamics Approach to Construction Safety Culture
16
� Focusing on the interactions among the CSC enablers and Goals, as well as their
consequences over time;
� Investigating causal relationships between those enablers and Goals;
� Providing a model to measure the CSC maturity levels; and
� Identifying areas for improvement in order to progress through to higher maturity
levels.
With the above aims in mind, this study set out to develop a CSC dynamic model to
simulate the interactions and causal relationships between the CSC enablers, and to
predict the influence of each enabler on safety goals, over a period of time. In
developing the CSC dynamic model, this study utilized a system dynamics (SD)
modelling technique to analyse and solve the problems, with a focus on policy analysis
and design. The details of the SD modelling are described in Chapter 6.
1.7 THESIS ORGANIZATION
The thesis is organized into eight chapters. The structure of each chapter is as follows:
� Chapter 1 describes the characteristics of the construction industry, the descriptions
of safety culture, and the attempts to measure safety culture. Gaps and shortcomings
of previous research studies are identified, and the research needs, as well as the
research aims, are stated.
� Chapter 2 outlines the adopted research methodology, including the research design,
the research activities and expected outputs, the literature review relating to the
development of the CSC model, the data collection methods, and an introduction of
the exploratory factor analysis (EFA), the structural equation modelling (SEM), and
the system dynamic (SD) modelling.
A System Dynamics Approach to Construction Safety Culture
17
� Chapter 3 reviews the literatures on performance measurement systems that were
potential basic models for the CSC. The selection was based on a number of criteria.
The EFQM Excellence model, the selected measurement system, was examined in
detail, including its key constructs and their associated attributes. The CSC model
was then proposed, based on the EFQM Excellence model. Lastly, the five levels of
CSC maturity were introduced at the end of this chapter.
� Chapter 4 details the questionnaire survey’s development and data collection. The
data interpretations, data screening, and preliminary analyses were performed to
increase confidence in the data collected.
� Chapter 5 presents the exploratory factor analysis (EFA), using the SPSS program,
to confirm the construct validity of the five enablers of the proposed CSC model.
Structural equation modelling was then performed to investigate the causal
relationships between the six constructs (five enablers and Goals) of the CSC
model. The final CSC model is presented at the end of this chapter.
� Chapter 6 sees the final CSC model being used to develop the CSC dynamic model
utilizing the system dynamics (SD) modelling technique. The developed dynamic
model was verified and validated to increase confidence in the model. The CSC
index, developed through the SD simulation, was used to measure the current CSC
maturity level of the organization.
� Chapter 7 presents a number of simulations, with different safety policies, that were
undertaken to examine different scenarios to enhance the CSC index and progress
through to higher CSC maturity levels. The cyclical style of safety management was
also modelled to reflect real-life situations where management withdraws attention
from safety, which then leads to a reduced CSC index.
� Chapter 8, the final chapter, concludes with the major findings, contributions to the
existing body of knowledge, implications for the Thai construction industry, and
limitations and recommendations for future research.
The next chapter (Chapter 2) outlines the research methodology adopted in this study.
A System Dynamics Approach to Construction Safety Culture
18
[THIS PAGE IS INTENTIONALLY LEFT BLANK]
A System Dynamics Approach to Construction Safety Culture
19
2 RREESSEEAARRCCHH MMEETTHHOODDOOLLOOGGYY
2.1 GENERAL OVERVIEW
As discussed in Chapter 1, the shortcomings of previous research studies lie in their not
considering the interactions and causal relationships between what the organization is
doing, and what it aims to achieve. Moreover, no index is available to use in assessing
CSC maturity levels. Thus, this study aims to develop a CSC dynamic model to explain
the interactions among the key constructs (five enablers and Goals) of the CSC. The
CSC index, developed through the SD simulation, was a tool used to help measure the
CSC maturity level, and to address areas for safety improvement.
This chapter presents different research steps/activities used to achieve the research
aims stated in Chapter 1.
2.2 RESEARCH DESIGN AND RESEARCH FRAMEWORK
The research design (shown diagrammatically in Figure 2.1) involved a review of the
safety culture literature to identify the gaps of previous research studies. The research
needs, followed by the research aims, were identified to fill the research gaps. The
research aims required the development of the CSC model. The data was then collected
via a questionnaire survey. Following the data analyses, the CSC dynamic model and
the CSC index were constructed to fulfil the research aims of the study.
A System Dynamics Approach to Construction Safety Culture
20
Figure 2.1 Research design
The ultimate goal of this study was to develop a CSC index for measuring the CSC
maturity level of an organization. To achieve this goal, a number of research activities
and expected outputs were planned (see Figure 2.2). The details of each activity are
described below.
Literature Review
Research Aims
Model Development
Data Collection
Gaps of Previous Research Studies
Data Analyses
Analysis Results
Research Need
Fulfil Research Aims
A System Dynamics Approach to Construction Safety Culture
21
Figure 2.2 Research activities and expected outputs
The Proposed CSC Model Literature Review on the Performance Measurement Systems
(See Chapter 3)
Data Collection
The Baseline Model of the CSC
The Final CSC Model
The CSC Dynamic Model
The Verified and Validated Model
Literature Review on the Constructs of the EFQM Excellence Model
(See Chapter 3)
Exploratory Factor Analysis (EFA) (See Chapter 5)
Structural Equation Modelling (SEM) (See Chapter 5)
System Dynamics (SD) Modelling (See Chapter 6)
Verification and Validation (See Chapter 6)
Activities Outputs
Enablers, Goals, and their Associated Attributes of the Proposed CSC Model
Questionnaire Survey (See Chapter 4)
The CSC Index SD Simulation (See Chapter 7)
Data Screening and Preliminary Analyses (See Chapter 4)
Screened Data
A System Dynamics Approach to Construction Safety Culture
22
2.2.1 Performance Measurement Systems: The Review
To develop the CSC model, this study considered three major performance
measurement systems, including the Malcolm Baldrige National Quality Award
(MBNQA) framework, the Balanced Scorecard (BSC) framework, and the European
Foundation for Quality Management (EFQM) Excellence model (see details in Chapter
3). While these measurement systems are widely used in measuring an organization’s
performance (Caravatta, 1997; Wongrassamee et al., 2003), only one was selected,
however, to be used for the CSC model development.
The selection of a basic measurement system to develop the CSC model was based on a
comparison of the advantages and disadvantages of each system. The EFQM Excellence
model was the measurement system of choice, for the following reasons:
� The EFQM Excellence model, compared with the MBNQA framework, includes
areas of financial performance and impact on society, which, in the construction
industry, are considered important in order to improve the CSC (Wright et al.,
1999).
� Compared to the BSC framework, the EFQM Excellence model covers more aspects
of Partnerships and Resources, Customer Results, and Society Results. Indeed,
Wright et al. (1999) postulated that resources, such as human and financial
resources, are important in the planning of safety improvements. Further, Mohamed
(2003) suggested that the Customer is one of the four perspectives against which to
benchmark organizational safety culture in construction.
� The EFQM Excellence model identified strengths, and areas for improvement,
across the organization’s processes. This capability satisfied identified research
aims.
� The EFQM Excellence model focused on continuous improvement, including the
use of constructive feedback to improve performance. This capability suited the SD
modelling technique that was used in developing the CSC dynamic model.
A System Dynamics Approach to Construction Safety Culture
23
To better understand the EFQM Excellence model, its key constructs, as well as the
attributes associated with each construct, are examined in detail in Chapter 3.
2.2.2 The Constructs of the EFQM Excellence Model: The Review
The EFQM Excellence model comprises nine constructs, including five ‘enablers’ and
four ‘results’. The five ‘enablers’ are Leadership, Policy and Strategy, People,
Partnerships and Resources, and Processes; while the four ‘results’ are People Results,
Customer Results, Society Results, and Key Performance Results. However, the focus of
this study was mainly on the improvements of, and interactions among, enablers’
criteria, to achieve better results, so that the four ‘results’ criteria could be combined
together into a single construct called Goals (see details in Chapter 3). As a result, the
CSC model was proposed; it consists of six constructs (five enablers and a single set of
Goals). Each construct comprises a number of attributes to explain the construct. For
example, the Leadership construct consists of four attributes (top management
commitment, effective two-way communication, management accountability, and
management leading by example), which were selected by their frequent citations
within recent construction safety literature (such as textbooks, scientific journals,
conference proceedings, theses and dissertations, and company reports). This approach,
advocated by Melville and Goddard (1996), is described in Figure 2.3.
A total of 34 attributes covering the six constructs of the CSC were chosen following
their frequent citations (see the attributes in Table 3.7 of Chapter 3). Within the
constructs, Leadership consists of four attributes, Policy and Strategy five attributes,
People seven attributes, Partnerships and Resources four attributes, Processes seven
attributes, and Goals seven attributes. These attributes were used in developing the
questionnaire survey for data collection; the details of the questionnaire survey are
described next.
A System Dynamics Approach to Construction Safety Culture
24
Textbooks
Company Reports
Scientific Journals
Conference Proceedings
Theses and Dissertations
Construction Safety Literature
Good starting place for safety culture literature, and good sources
for the analyses’ methodologies
e.g. Byrne (2001), Coakes and Steed (2003), Ithink (2003), and
Tabachnick and Fidell (2007)
Up-to-date information
e.g. Engineering, construction and architectural management; Journal of
construction engineering and management; and Safety sciences
Gatherings of researchers in particular field, where scientific
results are presented
e.g. Proceedings of the CIB W99 2006 international conference on global unity for safety and health in construction, 28-30 June 2006,
Beijing, China
Up-to-date, and often unpublished, information
e.g. Hsu (2002), Huy (2002), and Ali (2006)
Practical value, and can be easily downloaded from the
internet
e.g. HSC (2003), INEEL (2004), International Labour
Organization (2005), and NOSHC (2005)
Figure 2.3 Construction safety literature (Adapted from Melville and Goddard, 1996)
2.2.3 Data Collection: Questionnaire Survey
2.2.3.1 Survey Content
In conducting a research study, three different types of information (facts, behaviours,
and opinions) may be obtained from survey respondents (Dane, 1990). Facts are
anything that can be verified independently; they are phenomenon or characteristics
available to anyone who knows how to observe them. The information collected from
facts is often called ‘demographic characteristics’ (such as age, gender, name of the
organization, years of experience, and job title). Behaviours are actions completed by a
respondent. Like facts, behaviours can be verified, but only if they are witnessed or
A System Dynamics Approach to Construction Safety Culture
25
indirect evidence is obtained. Opinions are expressions of a respondent’s preference,
feeling, or behavioural intention. They can be objectively measured. In this study,
respondents’ opinions or perceptions on the CSC are used in a number of data analyses.
2.2.3.2 Survey Methods
According to Kumar (2005), there are two major approaches (secondary and primary
sources, see Figure 2.4) to gathering information about a situation, person, problem, or
phenomenon. Information from secondary sources is normally available and needs only
be extracted. Primary data, on the other hand, can be collected by observations,
interviews, or questionnaires.
Methods of Data Collection
Secondary Sources
Documents
�Government publications�Earlier research�Census�Personal records�Client histories�Service records
Primary Sources
Observation Interviewing Questionnaire
Participant
Non-participant
Mailed questionnaire
Collective questionnaire
Structured
Unstructured
Figure 2.4 Methods of data collection (Adapted from Kumar, 2005)
A System Dynamics Approach to Construction Safety Culture
26
In this study, three methods of data collection (observation, interview, and
questionnaire) were initially considered. Observation is a purposeful systematic and
selective way of watching and listening to an interaction as it takes place. Usually, there
are two types of observations: participant and non-participant. In participant
observations, a researcher participates in the activities of the group being observed in
the same manner as its members, with or without their knowing that they are being
observed. Non-participant observations occur, on the other hand, when the researcher
does not get involved in the activities of the group. According to Kumar (2005), the
problems with using observation as a method of data collection may include the
possibility of observer bias, and the misinterpretation of observations.
A personal interview, sometimes called a face-to-face interview, has a major advantage
in that it allows the researcher to record not only verbal responses, but also any facial or
bodily expressions. These nonverbal responses may give the researcher greater insight
into the respondents’ true opinions and beliefs. The other advantages include: 1) the
respondents can ask for the questions to be clarified; 2) the researchers can ask follow-
up questions, if they think they will provide more reliable data; 3) supplementary
material, such as audio/video materials, can be used to increase the respondents
understanding of the questions; and 4) the response rates are generally high. This type
of survey, however, is time-consuming, and it can be very costly. It may also generate
interviewer bias, if the interviewer is not well trained. In addition, participants may be
more likely to give socially desirable responses, because it is deemed appropriate by
society (Jackson, 2003).
A written questionnaire is self-administered, and can be sent through the traditional mail
system or by email. It is important that a mail survey be clearly written and self-
explanatory because no one will be available to answer questions regarding the survey,
once it has been mailed out. Questionnaire surveys have several advantages, for
example, they generally have less sampling bias (a tendency for one group to be
overrepresented in a sample) than personal interviews. They also allow the researcher to
collect data on more sensitive information. Participants, who may be unwilling to
discuss personal information with someone face-to-face, may be willing to answer such
questions in a written survey. This method is usually less expensive and can cover a
A System Dynamics Approach to Construction Safety Culture
27
large geographical area. Further, the participants can take as much time as they need to
answer the questions without feeling the pressure of someone waiting for the answers.
However, the major problems concern the low response rate, and the misinterpretation
of the questions (McBurney, 1994).
In summary, the three methods of data collection offer different advantages and
disadvantages. For example, the observation is the best approach to collect information
required for the researcher interested in the behaviours rather than the perceptions of
individuals. While the personal interview offers the opportunity for questionnaire
clarification, it is costly and time-consuming, and may suffer from interviewer bias.
Finally, the questionnaire survey is less expensive and has no answers bias, but it does
have a poor response rate.
Due to the limitation of costs and time, this study used the questionnaire survey as the
data collection method of choice. The drop-off and collect approach was used (together
with the mailing method) to increase the response rate. Further, with this approach, help
was provided during the drop-off and collect times to clarify any misunderstandings.
The intention of this research was to study the construction industry sector in Thailand.
Medium to large construction-contracting organizations, with staff of 100 or more, were
selected for the sampling. Targeted respondents were selected on the assumption that
they held senior appointments, such as executive directors, managing directors, and
senior project managers, within their respective organizations. To avoid any
misinterpretation or misunderstanding, each question was written in English, with the
Thai translation underneath. The Thai translation was reviewed and corrected by a local
Thai professional translator.
The questionnaire survey included open-ended, partially open-ended, and rating-scale
questions. The open-ended questions related to data, such as the name of the
organization, job title, and working experience. With the partially open-ended questions,
the respondents were asked to choose one of the provided answers that best represented
their beliefs, or to select the ‘other’ option, if the answers provided were not
A System Dynamics Approach to Construction Safety Culture
28
appropriate. The rating-scale questions asked the respondents to indicate their degree of
agreement or disagreement with each statement provided (that represented each attribute
of the CSC) using a five-point Likert-type scale (points 1 and 5 represented strongly
disagree and strongly agree, respectively). The Likert scale was chosen as it measures
the magnitude of the opinion, not simply the direction (McBurney, 1994). The
questionnaire survey and data collection are described in detail in Chapter 4.
2.2.4 Data Screening and Preliminary Analyses
To increase confidence in the data collected, via the questionnaire survey, it was
examined for characteristics such as response rate, working experience of the
respondents, and the position of the respondents (see Chapter 4). The Statistical Package
for Social Sciences (SPSS) program version 11.5 was used to ensure data consistency,
and to allow the results to be meaningfully interpreted. Thus a number of data screening
and preliminary analyses, including the handling of missing data, the normality test, the
outliers test, and the reliability test, were performed. The screened data were then
further analysed using more complex analyses, including the exploratory factor analysis
(EFA), the structural equation modelling (SEM), and the system dynamics (SD)
modelling.
2.2.5 Exploratory Factor Analysis: The Introduction
According to Seo et al. (2004), the exploratory factor analysis (EFA) is a precursor to
the structural equation modelling. In this research, the EFA was performed to gather
information about the interrelationships among a set of attributes, and to yield a factor-
based scale of the CSC. In other words, the EFA was used to confirm the validity of the
five enabler-constructs of the proposed CSC model.
A System Dynamics Approach to Construction Safety Culture
29
The method of principal axis factoring, with an eigenvalue over one, together with the
varimax rotation, was chosen for the analysis. It was expected that the baseline model of
the CSC would be achieved from this analysis (see Chapter 5).
2.2.6 Structural Equation Modelling: The Introduction
To provide further evidence of construct validity and to investigate the causal
relationships between the six constructs of the CSC (five enablers and the single set of
Goals), the baseline model of the CSC was performed using structural equation
modelling (SEM). SEM comprises two main tests: the measurement model, and the
structural model. Firstly, it is important to confirm the measurement model before the
structural model can be conducted. The results of the SEM helped clarify the causal
relationships, as well as the degrees of influence, of the six CSC constructs (see details
in Chapter 5). These clarified relationships were then used in the CSC dynamic model
development (see Chapter 6).
While a number of programs can be used to conduct the SEM (such as AMOS, CALIS,
EQS, LISREL, Mplus, Mx Graph, RAMONA, and SEPATH) (Kline, 2005), the AMOS
program version 6.0 was chosen for this analysis. AMOS (Analysis of Moment
Structure) is a Microsoft Windows program made up of two core modules: AMOS
Graphics and AMOS Basic (Kline, 2005). In AMOS Graphics, the program provides a
graphical user interface through which the user can specify the model by drawing it on
the screen. Further, all other aspects of the analysis are controlled through this interface.
Indeed a complete set of tools is available under AMOS Graphics for drawing,
modifying, or aligning graphical elements of model diagrams. Each tool is represented
by an icon, and performs one particular function. AMOS window, together with its
palette displaying icons, is shown in Figure 2.5 below.
A System Dynamics Approach to Construction Safety Culture
30
Figure 2.5 AMOS window and its drawing tool icons
2.2.7 System Dynamics Modelling: The Introduction
The system dynamics (SD) modelling was used to develop the CSC dynamic model. It
was first introduced by Forrester (1961) as a method for modelling and analysing the
behaviour of complex social systems, particularly in an industrial context. It has been
used to examine various social, economic, and environmental systems, where a holistic
view is important, and feedback loops are critical to the understanding of the
interrelationships (Rodrigues and Bowers, 1996). Simonovic (2005) stated that a SD
simulation approach relies on an understanding of complex interrelationships existing
among different elements within a system. This understanding is achieved by
developing a model that can simulate and quantify the behaviour of the system over
time. Such simulations are considered essential in understanding the dynamics of the
system.
A System Dynamics Approach to Construction Safety Culture
31
However, it is difficult to evaluate a SD, as there are no performance criteria for such an
evaluation (Barnes et al., 2005). Nevertheless, some of its strengths and limitations are
stated below. The strengths of SD applications are that:
� It looks at the policies as well as the processes: SD enables policies to be included
in the model as well as processes, so that any problems with the policies can be
addressed. The type of policy may be formal or informal. High overtime levels, for
example, may result from an informal policy, interpreted as the feeling one has to
work long hours to look good. Such a policy can be included into a SD model.
� It provokes serious systems thinking: The idea of the SD is to look at the problem as
a whole, including those influencing factors that affect the behaviour of the system
(such as cause and effect interrelationships between system variables). The outcome
is a more consistent solution.
� It includes high (qualitative, conceptual) level, as well as low (quantitative,
detailed) level, analysis: SD incorporates both qualitative analyses, such as causal
loop diagrams, and quantitative analyses, incorporating rates and levels. These are
useful as they provide a good basis for decision-making.
Some limitations of SD applications are listed below:
� It may be difficult to apply at detailed levels: This difficulty results from the amount
of mathematical analysis required, especially when it lacks the use of a computer
program.
� It has a problem with the delay time factor: At the conceptual level, there is no
reference to the length of the delay between two elements; putting a ‘delay’ into the
diagram only proves that there is a delay that will affect the outcome. At the
quantitative level, it may be difficult to accurately predict the length of the delay,
which may then affect the simulation result. A number of simulations may have to
be run, with varying delay lengths, to obtain a general idea of what the effects might
be.
� It is difficult to set the boundary of the system: To set the boundary of the system, all
factors that significantly affect the problem must be represented. In practice, it is
hard to judge which factors should be included or excluded.
A System Dynamics Approach to Construction Safety Culture
32
� It has a problem with the time horizon: Two issues relating to this limitation are the
length of time to which the model relates (that is, two years or 50 years), and the
inability to compare the effect of events that occur at different points of time. If the
model relates to a long period of time, it is likely that the system structure will
change during that time, thus making the results invalid.
Despite these limitations, the SD methodology provides a good basis from which to
make decisions. It allows for the interrelationships among important variables, all of
which affect the problem, thus providing a better understanding of the problems, and the
ways in which it can be solved.
2.2.7.1 SD Modelling in Construction
In the construction domain, many researchers have reported SD modelling applications.
Love et al. (2000), for example, developed a SD model to capture the interrelationships
among factors that contribute to design errors and reworks in construction projects. The
developed model helped unravel a series of complex problems into more manageable
interrelated components. It enabled the design and project managers to better understand
the process of design documentation and how design errors occur in construction
projects. The outcome was more effective design management, and an improvement in
the profitability and competitiveness of the design firm.
Chritamara and Ogunlana (2002) developed a SD model to better understand the
interacting nature of the problems inherent in the design-and-build procurement of
construction projects in Thailand. The dynamic model was validated and calibrated for a
typical large design-and-build infrastructure project using the time and cost overrun
problems that were experienced. Extensive simulations, with many policies,
individually and in various combinations, were made via SD modelling to identify the
best policy.
A System Dynamics Approach to Construction Safety Culture
33
Tang and Ogunlana (2003a) used SD modelling to gain insights into the interactions
between a country’s construction market and the organization’s financial, technical, and
managerial capabilities. They also employed the SD methodology to provide a careful
and holistic evaluation of the improvement policies to enhance organizational
performance (Tang and Ogunlana, 2003b).
2.2.7.2 The use of SD Modelling in this Study
The studies referred to above verify the proven usefulness of SD modelling in the
construction industry. Additionally, according to Sterman (1992), SD modelling has a
number of advantages that make it suitable for an organization in developing strategies
and improving performance. Such SD advantages are:
� The methodology was developed to deal with dynamics.
� The model is well suited to representing multiple interdependencies.
� It is the modelling method of choice where there are significant feedback processes.
� More than any other modelling technique, it stresses the importance of non-linearity
in model formulation.
� Both hard (objective oriented, formal, and quantitative) and soft (learning oriented,
intuitive, and qualitative) data can be used in the model.
� The model can be used to deal with future behaviour of the system.
SD modelling was used in this study to capture the interactions and causal relationships
between the five enablers and Goals of CSC, over a period of time. The reasons for its
use are:
� The CSC includes many variables that can give rise to changes, such as changes in
safety policy and strategy, in the workforce, and in the resources. A change,
including its effects, may cause another change (referred to as a dynamic change);
A System Dynamics Approach to Construction Safety Culture
34
this may, in turn, affect the whole system. Thus, SD modelling can be used to deal
with these dynamic changes.
� There is a need to investigate the interactions and causal relationships between the
CSC enablers, and the feedback of Goals on the enablers. SD modelling can be used
to capture these feedback processes.
� Most of the data in this study are soft data, such as top management commitment to
safety, provision of safety resources, risk and hazard assessment, etc. SD modelling
permits the use of soft data in the modelling.
� SD modelling can facilitate testing alternative strategies to improve the CSC in
organizations without actually having to implement them. This saves money by
eliminating costs that may occur from not implementing the best safety strategy.
2.2.7.3 SD Software
A number of software programs have been developed for SD modelling. Eberlein’s
(2007) list, with brief details, is presented below:
� DYNAMO: DYNAMO (Dynamic Models) was the first SD simulation language; for
a long time the language and the field were considered synonymous. Originally
developed by Jack Pugh at Massachusetts Institute of Technology (MIT), the
language was made commercially available from Pugh-Roberts in the early 1960s.
DYNAMO today runs on PC compatibles under DOS/Windows. It provides an
equation based development environment for SD models.
� Powersim (www.powersim.com): In the mid 1980s, the Norwegian government
sponsored research aimed at improving the quality of high school education using
SD models. This project resulted in the development of Mosaic, an object oriented
system aimed primarily at the development of simulation-based games for
education. Powersim was later developed as a Windows based environment for the
development of SD models that also facilitates packaging as interactive games or
learning environments.
A System Dynamics Approach to Construction Safety Culture
35
� Vensim (www.vensim.com): Originally developed in the mid 1980s for use in
consulting projects, Vensim was made commercially available in 1992. It is an
integrated environment for the development and analysis of SD models. Vensim
runs on Windows and Macintosh computers.
� STELLA (Structural Thinking Experimental Learning Laboratory with Animation)
(www.iseesystems.com): Originally introduced on the Macintosh in 1984, the
STELLA software provided a graphically oriented front end for the development of
SD models. The stock and flow diagrams, used in the SD literature, are directly
supported with a series of tools supporting model development. Equation writing is
made through dialog boxes accessible from the stock and flow diagrams. STELLA
is available for Macintosh and Windows computers.
In this study, the CSC dynamic model is formulated using the dynamics software
package ‘STELLA’ (Ithink, 2003) because of its significantly better interface
capabilities. The graphical depictions of the STELLA models, and the ability to quickly
adjust a model and run it on the software, have proven to be an excellent discussion
medium for model enhancement (Morecroft, 1988).
The STELLA software provides the modeller with a menu of symbols for creating a
system diagram on a computer screen (see Figure 2.6 and Table 2.1). Symbols are
selected and moved onto the screen and then connected. Modellers are constrained by
the SD connection rules to produce diagrams, which connect symbols in a set sequence.
In addition to symbols, STELLA provides guidelines for equation formulation. These
guidelines can be thought of as rules for converting symbols, text and words into
algebra (Morecroft, 1988).
A System Dynamics Approach to Construction Safety Culture
36
Stock Flow Converter Connector Decision Process Diamond Paint Brush
Dynamite Ghost
Figure 2.6 STELLA menus of symbols for creating a model (Ithink, 2003)
Table 2.1 Flow diagram model conventions (Morecroft, 1988)
Symbol Name Description safety culture index
Stock A stock can be defined as a structural term for anything
that accumulates, for example, savings in the bank.
enablers
Flow If stocks are bathtubs, then flows are pipes that feed and
drain them.
leadership
Converter If stocks are names of structure and flows are verbs, then
converters are adjectives and adverbs.
enablers
Cloud A cloud is an infinite reservoir representing the map
boundary. The capacity of cloud is so great that it makes
no sense to worry about filling or draining it.
enablers
leadership
Connector A connector is used to link the sectors and converters to
other converters
Cloud
Graph Pad Table Pad
Numeric Display
Text Box Hand
Navigation Arrows
Map/ Model Toggle
Sector Frame
Graphics Frame
Button
A System Dynamics Approach to Construction Safety Culture
37
2.2.7.4 Model Verification and Validation
The development of the CSC dynamic model is described in detail in Chapter 6.
According to Forrester and Senge (1980), the developed dynamic model must be tested
with two important processes: model verification and model validation, to establish
confidence in the soundness and usefulness of the dynamic model. McLucas (2005)
stated that model verification and validation are the two model building processes of
establishing confidence in the dynamic model. Unfortunately, these two terms are often
used interchangeably, which leads to confusion. Verification and validation involve two
distinctly different types of activities, but they are inseparable when it comes to SD
modelling.
According to Rakitin (2001), verification is defined as “the process of determining
whether or not the products of a given phase of SD modelling development cycle fulfil
the requirements established during the previous phase.” Verification involves
designing and applying a sufficiently exhaustive set of tests that measure how models
behave. This behaviour is compared with the modes of behaviour specified for complete
models. Indeed, there is a variety of tests that may be conducted in the process of
verifying a model. Such tests include logical, extreme-value, and mass-balance tests
(McLucas, 2005).
� Logical tests are designed to assure the parametric verification, dimensional
integrity, and unit consistency, correct sequence of calculation, and
stochastic/statistical character.
� Extreme-value tests are designed to assure stability under exposure to extreme
conditions and extreme policies.
� Mass-balance tests are designed to assure that physical flows do not violate the
basic requirement for physical flows into a model to either accumulate or flow out.
Mass-balance must be assured during every time-step of every simulation run.
Rodrigues and Williams (1998) postulated that the primary purpose of model validation
is to ensure that the model captures the general dynamics of the system behaviour, and
A System Dynamics Approach to Construction Safety Culture
38
produces results that are as close as possible to their real occurrences. In order to
validate the model, two validation steps must be undertaken carefully: 1) the feedback
structure must be able to capture the general dynamics of system behaviour; and 2) the
calibration parameters for a specific situation must be as close as possible to their real
occurrences.
A model is considered behaviourally validated if simulation results display similar
behavioural patterns when compared with observed behaviour in a real system.
Structural validity is attained by first evaluating every relationship and feedback loop in
the dynamic hypothesis to ensure that it captures the general dynamic behaviour of a
construction organization. Second, the parameters and equations used in the system
dynamic model are investigated to ensure that the parameters match the effect of
corresponding parts in the real system (Tang and Ogunlana, 2003a). According to
Forrester and Senge (1980), the focus of validating activities are on three types of tests
(model structure, model behaviour, and policy implications), as demonstrated in Tables
2.2 to 2.4, respectively. These tests aim to identify the cause-and-effect mechanisms of
the model.
Table 2.2 Tests of model structure (Forrester and Senge, 1980)
No. Test Explanation
1 Structure Verification Compares the model structure directly with the structure of
the perceived system the model represents.
2 Parameter Verification Compares model parameters to knowledge of a perceived
system to determine if parameters correspond conceptually
and numerically to real life.
3 Extreme Condition Tests Improve the model in the normal operating region by
examining the effects of extreme conditions.
4 Boundary Adequacy Assesses whether a model aggregation is appropriate, and if
the model includes all relevant structure.
5 Dimensional Consistency Uses dimensional analysis to validate model’s rate
equations.
A System Dynamics Approach to Construction Safety Culture
39
Table 2.3 Tests of model behaviour (Forrester and Senge, 1980)
No. Test Explanation
1. Behaviour Reproduction (Focuses on reproducing historical behaviour)
a. Symptom-Generation
Tests
Examine whether a model recreates the symptoms of
difficulty that motivate construction of the model.
b. Frequency Generation and
Relative-Phasing Tests
Focus on periodicities of fluctuations and phase
relationships between variables.
c. Multiple-Model-Tests Consider whether a model is able to generate more than
one mode of observed behaviour.
d. Behaviour-Characteristics Focus on any peculiar shapes in a fluctuating time
series.
2. Behaviour Prediction (Focuses on future behaviour)
a. Pattern-Prediction Tests Examine whether a model generates believable patterns
of future behaviour.
b. Event-Prediction Tests Focus on a particular change in circumstances.
3. Behaviour Anomaly Tests Trace a behavioural anomaly to the elements of the
model structure responsible for the behaviour.
4. Family Member Tests Assess whether a model is generic to the class of
system to which the particular member belongs.
5. Surprise Behaviour Test Highlights behaviour in the real system that has not
been previously recognized.
6. Extreme Policy Involves altering policy statement in an extreme way,
and runs the model to determine dynamic
consequences.
7. Boundary Adequacy Considers whether a model includes the structure
necessary to address the issues for which it is designed.
8. Behaviour Sensitivity
Focuses on the sensitivity of model behaviour to
changes in parameter values. It ascertains whether
plausible shifts in model parameters can cause a model
to fail behaviour tests that are previously passed.
A System Dynamics Approach to Construction Safety Culture
40
Table 2.4 Tests of policy implications (Forrester and Senge, 1980)
No. Test Explanation
1. System Improvement Considers whether policies found beneficial after
working with a model, when implemented, also
improve real-system behaviour. In time, this test
becomes the decisive test, but only as repeated real-
life applications of a model lead overwhelmingly to
the conclusion that the model points the way to
improved policy. In the meantime, confidence in
policy implications of models must be achieved
through other tests.
2. Changed-Behaviour Prediction Asks if a model correctly predicts how behaviour of
the system changes if a governing policy is changed.
3. Boundary Adequacy Examines how modifying the model will alter policy
recommendations arrived at by using the model.
4. Policy Sensitivity Reveals the degree of robustness of model behaviour,
and indicates the degree to which policy
recommendations may be influenced by uncertainty in
parameter values. Such testing can help to show the
risk involved in adopting a model for policymaking.
In this study, the ‘logical’ test had been used for model verification to assure parametric
verification, unit consistency, and correct sequence of calculation. The ‘behaviour
sensitivity’ test, on the other hand, was used to validate the model, as it is widely used in
many research studies (such as Miller, 1990; Duynisveld, 1999; Huy, 2002; and Tang
and Ogunlana, 2003a). The results of the model verification and validation are shown in
Chapter 6. The verified and validated dynamic model was simulated to achieve the CSC
index. This index was then used, together with the CSC maturity levels (see details in
Chapter 3), to assess the current CSC maturity level of the organization, and prioritise
areas for safety improvement. A number of policy analyses were then performed, with
SD modelling, to achieve the best policy the organization could use to enhance its CSC
index. Moreover, the cyclical style of safety management was modelled to reflect real-
A System Dynamics Approach to Construction Safety Culture
41
life situations where management withdraws its attention to safety, which leads to the
decreased CSC index (see details in Chapter 7).
2.3 SUMMARY
This chapter addressed relevant methodological issues, and described particular
appropriate methodologies used in this study. The key elements of the research
methodologies were the research design, and the research activities and expected
outcome frameworks. The research design framework provided the big picture for this
study, and encompassed the research aims and what needed to be done to fulfil those
research aims.
The research activities and expected outcomes framework, on the other hand, described
each step of the study to fulfil the research aims. The research started with a literature
review of performance measurement systems, to achieve a basic measurement system
for the CSC model development. The EFQM Excellence model, which was the selected
measurement system, was examined in detail, including its key constructs and their
associated attributes. The CSC model was then proposed, based on the EFQM
Excellence model. A questionnaire survey was developed for the data collection. The
data collected were screened to increase confidence in the data. A number of data
analyses were performed, including the EFA, the SEM, and the SD modelling, to
achieve a CSC dynamic model. The dynamic model was verified and validated to
ensure its usefulness. The verified and validated model was simulated to ultimately
achieve the CSC index, which was expected to fulfil the research aims.
The next chapter (Chapter 3) describes the critical literature review undertaken in
relation to the development of the CSC model.
A System Dynamics Approach to Construction Safety Culture
42
[THIS PAGE IS INTENTIONALLY LEFT BLANK]
A System Dynamics Approach to Construction Safety Culture
43
33 LLIITTEERRAATTUURREE RREEVVIIEEWW
3.1 GENERAL OVERVIEW
This chapter contains two parts: the literature review and the key constructs. The
literature review covered three performance measurement systems (the MBNQA
framework, the Balanced Scorecard framework, and the EFQM Excellence model).
These measurement systems have the potential to be used as a basic measurement
system for the CSC model development. The advantages and disadvantages of each
measurement system were compared to select the most suitable measurement system for
the CSC.
The second part presents the key constructs of the selected performance measurement
system, the EFQM Excellence model. The attributes associated with each construct are
stated, and the CSC model is proposed. Lastly, the five levels of CSC maturity are
introduced in the end of this chapter.
3.2 PERFORMANCE MEASUREMENT SYSTEMS
There are a number of performance measurement systems that can be used in measuring
an organization’s performance. Wongrassamee et al. (2003) grouped these performance
measurement systems into two broad categories: systems that emphasize self-
assessment and systems designed to help managers in measuring and improving
business processes. Each category is briefly described below:
� Systems that emphasize self-assessment consist of:
o The Deming Prize (www.deming.org);
A System Dynamics Approach to Construction Safety Culture
44
o The Malcolm Baldrige National Quality Award (MBNQA) framework
(www.quality.nist.gov); and
o The European Foundation for Quality Management (EFQM) Excellence model
(www.efqm.org).
� Systems designed to help managers in measuring and improving business processes
include:
o The Capability Maturity Matrices (www.sei.cmu.edu\cmm\);
o The Performance Pyramid;
o The Effective Progress and Performance Measurement (EP2M); and
o The Balanced Scorecard (BSC) framework.
Among the measurement systems listed above, Wongrassamee et al. (2003) observed
that the EFQM Excellence model and the BSC framework have received wide publicity,
and have recently been adopted by many organizations worldwide. Caravatta (1997),
however, argued that the MBNQA framework is one of the best tools for measuring
performance, claiming that there are approximately one million copies of this award’s
criteria in circulation, most of which are being used as a self-assessment tool.
In response to the above statements, this study considers three performance
measurement systems (the MBNQA framework, the BSC framework, and the EFQM
Excellence model), as the most potentially useful model for developing a CSC model.
The details of each system are described below.
3.2.1 The Malcolm Baldrige National Quality Award Framework
The MBNQA was established by the US Congress in 1987, to promote quality
awareness and thus improve the competitiveness of US companies. Since then, it has
become an important catalyst for improving competitiveness, and increasing the
A System Dynamics Approach to Construction Safety Culture
45
awareness of quality improvement methods (Pannirselvam and Ferguson, 2001). Its
criteria have been adopted or used as a model at local, state, and national, as well as
international levels. According to the National Institute of Standards and Technology
(NIST) (1993), the MBNQA criteria have three important purposes in strengthening US
competitiveness, viz�
� To help raise quality performance practices and expectations;
� To facilitate communication and sharing among and within organizations of all
types, based upon a common understanding of key quality and operational
performance requirements; and
� To serve as a working tool for planning, training, assessment, and other uses.
It was expected that the award criteria would help organizations to achieve two
competitive goals, including the delivery of ever-improving value to customers, which
would result in an improved marketplace and operational performances.
The core concepts were embodied in the MBNQA framework, and, according to
Cortada and Woods (1994), they consist of seven categories, namely:
� Leadership;
� Information and analysis;
� Strategic quality planning;
� Human resources development and management;
� Management of process quality;
� Quality and operational results; and
� Customer focus and satisfaction.
Table 3.1 presents these categories and the examination items used under each category.
A System Dynamics Approach to Construction Safety Culture
46
Table 3.1 The MBNQA categories and items (Pannirselvam and Ferguson, 2001)
No. MBNQA Category Examination Item
1. Leadership Senior executive leadership
Management for quality
Public responsibility
2. Information and analysis Management of data
Benchmarks
Company level data
3. Strategic quality planning Performance planning process
Performance plans
4. Human resources development and management Human resource plans
Employee involvement
Employee training
Employee performance
Employee well being
5. Management of process quality Design quality
Process management
Support services management
Supplier quality
Quality assessment
6. Quality and operational results Quality results
Operational results
Business results
Supplier quality results
7. Customer focus and satisfaction Customer expectation
Customer relationship management
Commitment to customers
Satisfaction determination
Satisfaction results
Satisfaction comparison
The categories are grouped into four basic elements (driver, system, measures of
progress, and goals), as shown in Figure 3.1. Each element is described below:
A System Dynamics Approach to Construction Safety Culture
47
� Driver represents the ‘leadership’ category. It focuses on how top management
emphasizes quality at all levels, and communicates this emphasis throughout the
organization.
� System consists of processes for meeting the company’s quality goals and
performance requirements. Those processes are measured by ‘information and
analysis’, ‘strategic quality planning’, ‘human resources development and
management’, and ‘management of process quality’ categories.
� Measures of progress is measured by the ‘quality and operational results’ category.
� Goal includes the ‘customer focus and satisfaction’ category that focuses on
customer expectation, customer relationship, commitment to customer, satisfaction
determination, and satisfaction results and comparisons.
Figure 3.1 Four basic elements of the MBNQA framework (NIST, 1993)
Customer focus and satisfaction
Goal
Quality and operational results
Measures of Progress
Driver
Leadership
System
Strategic quality planning
Information and analysis
Human resources development and
management
Management of process quality
A System Dynamics Approach to Construction Safety Culture
48
The criteria, along with the advantages and disadvantages of the MBNQA framework,
were compared with those of the EFQM Excellence model to select a basic framework
for a CSC model development. The details are discussed in Section 3.3.1.
3.2.2 The Balanced Scorecard Framework
The BSC framework was first introduced by David Norton and Robert Kaplan in 1990
(Wongrassamee et al., 2003). It is a system of linked objectives, measures, targets, and
initiatives, that collectively describe the vision (strategy) of an organization, and how
that vision (strategy) can be achieved. It is a tool designed to enable the implementation
of an organization’s strategy, by translating it into concrete and operational terms,
which can be measured, communicated, and driven to ensure decision-making and
action. According to Lamotte and Carter (2000), the BSC framework is organized
across four key perspectives (financial, customer, internal, and learning and growth), as
shown in Figure 3.2. The details of each perspective are described below:
Figure 3.2 Four key perspectives of the BSC framework (Lamotte and Carter, 2000)
Financial perspective
Customer perspective
Internal perspective
Learning and growth perspective
Vision
A System Dynamics Approach to Construction Safety Culture
49
� Financial (or shareholder) perspective describes the financial objectives that need
to be achieved to meet the expectations of the shareholder. It is normally in the form
of market presence, economic return, or asset utilisation.
� Customer (or external) perspective focuses on describing key attributes of the
products/services offering that represent value for the customer from the customer’s
point of view.
� Internal perspective describes the processes and activities, which, if executed at the
highest level of performance, will drive success in meeting financial and customer
objectives.
� Learning and growth (or innovation) perspective is often referred to as the enablers.
Its objectives may focus on developing specific skills and competencies, knowledge,
and information and culture. It represents the foundation of the company and its
future capability.
An example of a completed BSC template is shown in Table 3.2. The comparisons
between the BSC framework and the EFQM Excellence model, as well as the selection
of a basic model for the CSC, are described in Section 3.3.2.
Table 3.2 An example of a BSC template (Lamotte and Carter, 2000)
Perspective Objective1 Measure2 Target3 Initiative4
Financial
Organic revenue
growth
Revenue from
existing businesses
1998:$800m Re-packaging of
existing products
Customer Consumers first
choice brand
Consumer
satisfaction index
1998:7/10 In-store
presentations
Internal Cross-sell our
products
Percent revenue
from new products
1998:15% Train staff on
new product
offerings
Learning and
Growth
Communicate
strategy
Percent of people touched
by communication
road show
1998:60% Road show
around all
factories
Note: 1 Objective: Statement of what must be achieved if the strategy is to be successful and the vision realised. 2 Measure: How success in achieving the objectives will be measured and tracked. 3 Target: The level of performance or rate of improvement needed over a specific time-scale. 4 Initiative: Key action programmes required to achieve objectives.
A System Dynamics Approach to Construction Safety Culture
50
3.2.3 The European Foundation for Quality Management Excellence Model
The EFQM, a membership based, not-for-profit organization, was created in 1988 by 14
leading European businesses (EFQM, 2000). It has a key role to play in enhancing the
effectiveness and efficiency of European organizations, by reinforcing the importance
of quality in all aspects of their business activities, stimulating, and assisting the
development of quality improvement as a basis for their achievement of organizational
excellence (EFQM, 2000). The EFQM Excellence model has been acknowledged as an
effective way for organizations to improve the quality of their processes. Further, the
model has been used in business generally, as well as in specific industries, such as
hospitality, education, and construction (Camison, 1996; Wright et al., 1999; Sheffield
Hallam University, 2003). Empirical evidence suggests that the application of holistic
management models, such as the EFQM Excellence model, has a positive effect on
organizational performance (Kristensen and Juhl, 1999).
The model recognizes that there are many approaches to achieving sustainable
excellence in all aspects of performance. It is based on the premise that “excellent
results with respect to performance, customers, people, and society are achieved through
leadership driving policy and strategy, that is delivered through people, partnerships and
resources, and processes” (EFQM, 2000). The model consists of nine criteria, five of
which are ‘enablers’ and four of which are ‘results’ (see Figure 3.3). Enablers include
Leadership, Policy and Strategy, People, Partnerships and Resources, and Processes,
and results include People, Customer, Society, and Key Performance results. Put
simply, enablers cover what an organization is doing, while results cover what an
organization aims to achieve. Each criterion, in the context of safety management, is
defined below:
A System Dynamics Approach to Construction Safety Culture
51
Leadership
People
Policy and
Strategy
Partnerships and
Resources
ProcessesKey
Performance Results
PeopleResults
SocietyResults
CustomerResults
Enablers Results
Innovation and Learning
Figure 3.3 The EFQM Excellence model (EFQM, 2000)
� Leadership (Lds) describes how leaders develop and facilitate the achievement of
the mission and vision of health and safety, develop values required for long-term
success, implement them by appropriate actions and behaviours, and personally
involve themselves in ensuring that the organization’s safety management system is
developed and implemented.
� Policy and strategy (Pol) describes how an organization implements its mission and
vision of safety via clear stakeholder focused strategies, which are supported by
relevant policies, plans, objectives, targets, and processes.
� People (Ppl) describes how an organization manages, develops, and releases the
knowledge and full potential of its people at an individual, team-based, and
organization-wide level, and plans these activities to support its policies and
strategies and the effective operation of its processes.
� Partnerships and resources (Prs) describes how an organization plans and manages
its external partnerships with project participants and other stakeholders and
resources to support its safety policies and strategies and the effective operation of
its safety-related processes.
A System Dynamics Approach to Construction Safety Culture
52
� Processes (Pro) describes how an organization designs, manages, and improves its
processes to support its policies and strategies, and fully satisfy and generate
increasing value for its customers, employees, and other stakeholders.
� People results look at what an organization is achieving in relation to its own
employees.
� Customer results look at what an organization is achieving in relation to its external
customers (e.g. clients and project participants), and other stakeholders.
� Society results look at what an organization is achieving in relation to a local
community and society as appropriate.
� Key performance results look at what an organization is achieving in relation to its
planned performance.
The advantages and disadvantages of the EFQM Excellence model, along with those of
two other performance measurement systems (the MBNQA and the BSC frameworks),
are presented in the following section.
3.3 SELECTION OF A BASIC FRAMEWORK FOR THE
CONSTRUCTION SAFETY CULTURE
3.3.1 A Comparison between the MBNQA Framework and the EFQM
Excellence Model
The MBNQA framework and the EFQM Excellence model were both developed to
improve quality management (Tummala and Tang, 1995). Both models are results-
oriented, and give maximum weight to customer satisfaction. The MBNQA criteria,
however, do not include financial performance whereas it is included in the EFQM
Excellence model, thus, making it less broad-based than the EFQM Excellence model.
Further, the EFQM Excellence model, by including the impact on society as one of the
A System Dynamics Approach to Construction Safety Culture
53
nine criteria, covers more aspects (such as preservation of global resources) in a more
detailed fashion than the MBNQA framework.
The similarities and differences of these two models are indicated in Table 3.3. Further,
a more detailed comparison of these two models, against the core concepts of strategic
quality management, is shown in Table 3.4.
Table 3.3 Similarities and differences of the MBNQA framework and the EFQM
Excellence model
Criteria The MBNQA
Framework
The EFQM
Excellence Model
Being results-oriented � �
Giving maximum weight to customer satisfaction � �
Including financial performance �
Including the impact on society �
Table 3.4 Comparison between the core concepts, the MBNQA framework, and the
EFQM Excellence model (Tummala and Tang, 1995)
The MBNQA Framework Core Concepts The EFQM Excellence Model
Leadership Leadership Leadership
Information and Analysis Strategic Quality Planning Policy and Strategy
Strategic Quality Planning Continuous Improvement People
Human Resources Development
and Management
People Participation and
Partnership
Partnerships and Resources
Management of Process Quality
Design Quality, Speed and
Prevention
Processes
Quality and Operational Results Customer Focus People Results
Customer Focus and Satisfaction Fact-Based Management Customer Results
Society Results
Key Performance Results
A System Dynamics Approach to Construction Safety Culture
54
The dashed lines indicate the inclusion of more aspects in Partnerships and Resources,
Society Results, and Key Performance Results criteria of the EFQM Excellence model,
thus showing that all core concepts of strategic quality management are embedded in
the EFQM Excellence model.
The above comparisons confirm that the EFQM Excellence model is a more fitting
basic CSC model, for the following reasons:
� Its inclusion of financial performance is one of the main issues of concern to the
construction industry. Tam et al. (2004) identified that a financial policy for safety
management must be considered during an organization’s goal setting to
appropriately distribute the budgets to aid safety activities.
� A society issue is included, whereas it is not stated in the MBNQA framework. In
the construction industry, organizations should pay attention to the local community
and the environment to: 1) reduce any risks that may affect society and the
environment (Little, 2002); 2) reduce social costs of accidents borne by society and
its institutions (Tang et al., 2003); and 3) improve the organization’s image and
increase the society’s perception of the organization’s competence (Wright et al.,
1999).
3.3.2 A Comparison between the BSC Framework and the EFQM Excellence
Model
Otley (1999) compared the similarities and differences of the BSC framework and the
EFQM Excellence model using five central areas of management control systems. The
results are summarized in Table 3.5. It is important to note that both models provide
broad and non-prescriptive templates, meaning that management can assign their own
measures to suit their corporate situations and environment.
A System Dynamics Approach to Construction Safety Culture
55
Table 3.5 Comparison between the BSC framework and the EFQM Excellence model
(Otley, 1999)
Item The BSC Framework The EFQM Excellence Model
Objectives Multiple objectives based on strategy
and emphasize four key perspectives
Multiple objectives based on TQM
principles and emphasize nine
criteria
Strategies and
Plans
Assign strategic measures by using
strategy map to connect each measure
to strategy
Not particularly addressed but all
weighted criteria and weighted
sub-criteria can be used as
guidance
Targets Not addressed due to non-prescriptive
template. Managers are required to
assign target performance levels
Management can set their expected
performance levels
Rewards Suggests that individual compensation
system should be linked to strategic
measures
Requires an appropriate reward and
recognition system, but no explicit
guidance given
Feedback Requires double-loop learning which is
more complicated than single-loop
feedback
The model itself provides feedback
information as a default of the
assessment method
Lamotte and Carter (2000) showed a comparison of these two models (see Table 3.6).
They concluded that the BSC framework was designed to communicate and assess
strategic performance, whereas the EFQM Excellence model focuses on encouraging
the adoption of good practice across all management activities of an organization.
A System Dynamics Approach to Construction Safety Culture
56
Table 3.6 High-level comparisons of the BSC framework and the EFQM Excellence
model (Lamotte and Carter, 2000)
Item The BSC Framework The EFQM Excellence Model
Origins � Performance measurement, value
creation
� TQM
Aspiration and
Benefits Sought
� Performance improvement
� To translate a company’s strategy
into focused, operational and
measurable terms
� Enabling strategic performance
� Performance improvement
� To identify strengths and areas for
improvement across an
organization’s processes to
encourage best management
practice
� Enabling best management
practice
Deliverables � A set of logically linked strategic
objectives with lead and lag
indicators/targets across four
perspectives
� A benchmark and relative
assessment of the quality of an
organization’s processes and
results by assessing/scoring against
nine criteria
Development
Approach
� Strategy driven, workshop based,
iterative, hypothesis driven,
management team involvement,
macro view, future looking
� Set of objectives and
measurement are unique to every
organization
� Step change in performance
� Process driven, self-assessment
fact gathering, data collection,
scoring based, detail oriented,
present focused
� Set of criteria and measurement
areas are the same for all
organizations
� Continuous improvement
Success Factors � Management team level
sponsorship and commitment
� On-going process embedded in
governance processes
� Management team level
sponsorship and commitment
� On-going process embedded in
day-to-day management
From an assessment of the above two tables (Tables 3.5 and 3.6), it is evident that the
EFQM Excellence model is more suitable to be used in developing a CSC model,
specifically: 1) it is able to identify strengths and areas for improvement across an
organization’s processes; this capability satisfies the research aims; 2) it focuses on
A System Dynamics Approach to Construction Safety Culture
57
continuous improvement; and 3) it adopts sets of criteria and measurement areas that are
the same for all organizations, thus making it easy for comparisons.
Further, Sheffield Hallam University (2003) mapped the BSC framework onto the
EFQM Excellence model, and concluded that the EFQM Excellence model covers more
aspects of Partnerships and Resources, Customer Results, and Society Results, while
they are not explicitly addressed in the BSC framework (see Figure 3.4). Indeed, Wright
et al. (1999), Mohamed (2003), and Tang et al. (2003) suggested that these aspects are
important in the planning of safety improvements in the construction industry. This,
thus, confirms the suitability of the EFQM Excellence model, as an optimum approach,
to use in developing a CSC model.
Internal Business Process Perspective•Process management approach •Process measures
Customer Perspective•Student and customer surveys•Student and customer indicators•Academic outcomes
Funding Provider Perspective•Financial performance•Academic outcomes•Audit outcomes•Institutional review outcomes
•Vision and values •Defining and working with stakeholders•Setting up a management system including process management
•Strategy development and implementation•Balanced Scorecard as approach
Innovation, Learning and Growth •Staff satisfaction•People indicators•People development and involvement•Use of RADAR•Self-assessment scores and outcomes
Leadership
People
Policy and
Strategy
Partnerships and
Resources
ProcessesKey
Performance Results
PeopleResults
SocietyResults
CustomerResults
Enablers Results
Innovation and Learning
Figure 3.4 Mapping the BSC framework onto the EFQM Excellence model
(Sheffield Hallam University, 2003)
A System Dynamics Approach to Construction Safety Culture
58
To further confirm the suitability of the EFQM Excellence model to be used as a basic
model for a CSC model development, Mbuya and Lema (2004) investigated the
relationships between the EFQM Excellence and the safety management system (SMS),
and found links between them (see Figure 3.5). These relationships illustrate that the
enablers of the EFQM Excellence model match with the essential elements of the SMS,
proving the appropriateness of the EFQM Excellence model for the CSC model
development.
Figure 3.5 Links between the safety management system and the EFQM Excellence
model (Adapted from Mbuya and Lema, 2004)
The next section describes the development of the proposed CSC model, based on the
EFQM Excellence model criteria.
Safety Management System Elements Enablers of the EFQM Excellence Model
Policy and Objectives
Organization
Management Review
Practices and Procedure
Implementation and Compliance
Verification and Assessment Processes
Partnerships and Resources
People
Policy and Strategy
Leadership
A System Dynamics Approach to Construction Safety Culture
59
3.4 THE PROPOSED CONSTRUCTION SAFETY CULTURE MODEL
The EFQM Excellence model is suitable to be used as a basic model for the CSC (see
Section 3.3). As discussed earlier, the model consists of five enablers that help to
achieve four results. In this study, however, the focus has been mainly on the
improvements of, and interactions among, the enablers’ criteria to achieve better results.
For this reason the four ‘results’ criteria were combined together into a single construct
(referred to hereinafter as Goals). The proposed CSC model is shown in Figure 3.6.
Enablers Goals
Innovation and Learning
Leadership(Lds)
(100 points)
People (Ppl)
(90 points)
Policy and Strategy
(Pol)(80 points)
Partnerships and Resources
(Prs)(90 points)
Processes(Pro)
(140 points)
Goals(500 points)
Figure 3.6 The proposed CSC model
The proposed CSC model assumes that leadership drives people management, policy
and strategy, as well as resources, and that these three enablers collectively influence
the ability to achieve pre-determined Goals through the implementation and
improvement of suitable processes (see Figure 3.6). The six theoretical constructs (five
enablers and the single set of Goals) represent the basic elements of the proposed
model.
A System Dynamics Approach to Construction Safety Culture
60
In addition to the enablers and Goals, the criterion weights are also an important part of
the model. As shown in Figure 3.6, a total of 1,000 points of the proposed model was
evenly split (500/500) between the enablers and Goals. The 500 points allocated to the
enablers were distributed as follows: 100 points to Leadership, 80 points to Policy and
Strategy, 90 points to People, 90 points to Partnerships and Resources, and 140 points
to Processes (EFQM, 2000). Importantly, this allocation of points among enablers,
reflecting their relative contribution to the achievement of Goals, is an area of much
debate. For practical purposes, however, this study adopted the original enablers’
allocation promoted by the EFQM Excellence model (see Figure 3.6). The Goals
construct, on the other hand, contained 500 points, which represented the aggregate
scores of people, customer, society, and key performance results.
The criterion weights of the above five enablers and Goals were later used as an input
into the development of the CSC dynamic model utilizing the SD modelling technique
(see Chapter 6).
Each construct of the proposed CSC model comprised a number of its associated
attributes, which were carefully selected, with reference to the frequency of citations in
recent construction safety literature. These attributes represented the items used to
operationalise each construct (as described in the questionnaire survey in Chapter 4).
The details of the six constructs, along with their associated attributes, are briefly
described below.
3.4.1 Leadership
Leadership and management commitment to safety is recognized as a fundamental
component of an organization’s occupational health and safety (Lingard and Blismas,
2006). In the area of construction, a number of research papers support leadership as the
main enabler in developing a good safety culture (Little, 2002; Mohamed, 2002;
Molenaar et al., 2002; Teo et al., 2005; Lingard and Blismas, 2006).
A System Dynamics Approach to Construction Safety Culture
61
Leadership can be examined using four associated attributes, namely top management
commitment, effective two-way communication, management accountability, and
management leading by example. A brief description of each is presented below:
1. Top management commitment: Teo et al. (2005) identified that safety culture in
construction organizations was dependent upon the safety commitment of
management and workers towards safety promotions and campaigns.
Organizations, where their top management gave high levels of safety support
and commitment, were found to have better safety performance and safety
records (Hinze and Reboud, 1988; Boonrod et al., 1998; Lingard and Blismas,
2006).
2. Effective two-way communication: Lardner et al. (2001) argued that to progress
through to higher safety culture maturity levels, an organization needed more
face-to-face communications, both formal and informal, between management
and frontline staff. Additionally, Little (2002) stated that two-way
communication was one of the key factors in improving safety culture. Teo and
Fang (2006), likewise, found that to enhance safety performance, safety
information should be passed down from top management to frontline workers.
3. Management accountability: According to Olcott (1997), safety culture was the
responsibility of management, and management held the accountability for
creating an atmosphere where each individual employee understood and
accepted his/her role in preventing accidents. As a safety program cannot be
successful on an individual basic, so that the responsibility to accomplish safety
activities must be transferred from top management to individuals at lower
levels of authority (Aksorn and Hadikusumo, 2007). In addition, all levels of site
management should be evaluated in terms of health and safety to ensure
appropriate accountability (Dias and Coble, 1996).
4. Management leading by example: Management needs to be a role model in how
to behave safely. This approach is an important key to enhancing safety culture;
a lack of proper modelling will lead to employees not taking the development of
a positive safety culture seriously (Dunlap, 2004).
A System Dynamics Approach to Construction Safety Culture
62
3.4.2 Policy and Strategy
The Policy and Strategy enabler refers to how an organization implements its mission
and vision of safety via clear stakeholder focused strategies, which are supported by
relevant policies, plans, objectives, targets, and processes. It consists of five attributes:
1) safety awareness and promotion; 2) alignment of productivity and safety targets; 3)
safety standards, laws, regulations; 4) safety initiatives to improve safety standards; and
5) safety as an integral part of business goal settings. The following briefly describes
these attributes:
1. Safety awareness and promotion (such as rewards, recognitions, and
punishments): Molenaar et al. (2002) included incentives and disincentives as
one of the characteristics of safety culture in construction organizations, and
stated that those incentive rewards could include informational (e.g. feedback),
social (e.g. praise/recognition) and tangible (e.g. bonuses/awards) reinforcement
for desired health and safety behaviour (Lingard and Blismas, 2006). Gibb and
Foster (1996) claimed that construction projects that use safety incentive
schemes demonstrated increased safety performance.
2. Alignment of productivity and safety targets: Potter (2003) proposed that to
enhance a culture of safety, safety should have the same weight as productivity
and profitability when economic decisions are made. Indeed safety rules should
be adhered to even under production pressures (particularly those imposed by
budgetary constraints) (Hinze and Reboud, 1988; Glendon and Litherland,
2001).
3. Safety standards, laws, and regulations: A good safety culture needs realistic
and workable safety rules that are practical in all situations (Glendon and
Litherland, 2001; Aksorn and Hadikusumo, 2006).
4. Safety initiatives to improve safety standards: Safety initiatives should be
proactively planned to continually improve safety standards (Boonrod et al.,
1998; Teo et al., 2005).
5. Safety as an integral part of business goal settings: Ahmed et al. (2004) stated
that safety is a company’s core issue, and it must be given a top priority in the
company’s goals setting. Dunlap (2004), likewise, stated that, in order for an
A System Dynamics Approach to Construction Safety Culture
63
organization to be effective in health and safety performance, leadership must
include health and safety in short and long term business goal settings.
3.4.3 People
Niskanen (1994) stated that it is not just management participation and involvement in
safety activities that is important, but also the extent to which management encourages
the involvement of the workforce. Management must be willing to devolve some
decision-making power to the workforce rather than simply play the more passive role
of recipient. In this way, workers are more likely to take ownership and responsibility
for their safety (Williamson et al., 1997).
The attributes associated with this enabler are shared perceptions about safety, safety
empowerment and responsibilities, supportive environment, workers involvement,
relationships among workers, workload, and work pressure. These attributes are
presented below:
1. Shared perceptions about safety: Employees with good perceptions of safety
tend to participate more in safety activities (Dedobbeleer and Beland, 1991;
Fang et al., 2006).
2. Safety empowerment and responsibilities: Hudson (2001) noted that there are
three main safety cultural developments, one of which is to involve workers in
the task of regulatory compliance, and encourage them to take personal
responsibility. In a positive safety culture environment, leaders provide a
reasonable rationale for a desired outcome, and then empower employees to
customize methods for achieving that outcome (Geller, 2000). According to
O’Dea and Flin (2001), empowering employees may be achieved by involving
them in decision-making and developing safety interventions and safety policy.
3. Supportive environment: Good teamwork is identified as a necessary
characteristic of a good safety culture (Olcott, 1997). In a good supportive
A System Dynamics Approach to Construction Safety Culture
64
environment, workers are responsible for their own safety, as well as for their
fellow workers’ safety (Teo and Fang, 2006).
4. Workers’ involvement: Mohamed (2002) stated that the higher the level of
workers’ involvement in safety matters, the more positive the safety climate.
Aksorn and Hadikusumo (2006) noted that successful safety programs largely
depend on employees’ involvement, as workers tend to support the activities that
they themselves help to create. For this reason, workers should be given the
opportunity to provide input into the design and implementation of safety
programs, such as being a member of the safety committee, reporting hazards
and unsafe practices to supervisors, identifying training needs, and investigating
accidents.
5. Relationships among workers: Cooperation between members and the
coordination of safety systems, particularly on multi-occupied sites, are
important if safety is to be improved (Langford et al., 2000). Construction
workers who continually interact with coworkers also rely on them to a greater
extent to provide a safer work environment (Olcott, 1997). Sites, where
workmates often give suggestions to each other on how to work safely, report
less accident rates and fewer workers’ distress (such as anxiety, frustration, and
job dissatisfaction) (Siu et al., 2004).
6. Workload: Glendon and Litherland (2001) observed that, to develop a positive
safety climate, workload should be reasonably balanced. Siu et al. (2004)
claimed that workers’ perceptions of high-role overloads are associated with an
increased tendency to engage in unsafe acts, thus, as fatigue and over-exertion
set in, workers will lose their focus, which results in unsafe acts (Cohen, 2002).
7. Work pressure: According to Siu et al. (2004), work pressures are caused by
distress, unworkable schedule times, and workforce instability (high turnover).
They proposed that psychological distress (such as anxiety, frustration, and job
dissatisfaction) predicts accident rates. Indeed workers who report more anxiety
report more injuries, and take fewer safety precautions. Time schedules for
completing work projects should be workable and realistic to enhance safety
climate (Glendon and Litherland, 2001; Aksorn and Hadikusumo, 2004).
Workforce stability is another important factor contributing to successful safety
programs (Cohen, 1977). Plants with low accidents usually have a workforce
composition that includes employees who are recruited or retained because they
A System Dynamics Approach to Construction Safety Culture
65
work safely; these work environments also have lower turnover and absenteeism
(Lee, 1998).
3.4.4 Partnerships and Resources
The Partnerships and Resources enabler describes how an organization plans and
manages its external partnerships with project participants and other stakeholders, and
organizes its resources to support its safety policies and strategies, as well as the
effective operation of its safety-related processes. Four attributes associated with this
enabler are project participants and stakeholders’ cooperation, adequacy of financial
resources dedicated to safety, availability of necessary safety-related resources, and
human resources management. These attributes are briefly described below:
1. Project participants and stakeholders’ cooperation: According to Wright et al.
(1999), cultural norms cannot be defined in isolation by management, but must
instead involve all key stakeholders (such as regulators, customers, staff, and
contractors) in decision-makings. This process ensures that those norms are
appropriate and meet the expectations of all parties. An effective safety culture
should be conceived of as an appropriate match between the behaviours, values,
and attitudes of members of the organization with the expectations of
stakeholders.
2. Adequacy of financial resources dedicated to safety: To achieve the
organization’s safety cultural goals, financial resources should be allocated to
aid health and safety policies (such as training, recruiting, and acquiring
information) (Wright et al., 1999).
3. Availability of necessary safety-related resources: Aksorn and Hadikusumo
(2006) proposed that a successful safety implementation could not be
accomplished by lack of safety resources. Sufficient safety resources should be
allocated to carry out day-to-day activities to accomplish short and long term
safety goals (Wangniwetkul, 2007). Satisfactory safety facilities, including tools,
equipment, and information, should be provided to the staff so that they can
implement safety activities safely (Sorensen, 2002).
A System Dynamics Approach to Construction Safety Culture
66
4. Human resources management: An organization should endeavour to have
adequate staff to ensure that the job is performed safely (Oklahoma Department
of Labour, 1998).
3.4.5 Processes
This enabler describes how an organization designs, manages, and improves its
processes to support its policies and strategies, and to fully satisfy and generate
increasing value for its customers, employees, and other stakeholders. It consists of
seven attributes (safety training, risk and hazard assessment, feedback on safety
implementation, adopting a no-blame approach, site layout planning and good
housekeeping, site safety documentations, and having an effective benchmarking
system), as described below:
1. Safety training: Training is a major factor influencing safety levels, as it helps
personnel carry out various activities effectively, establishes a positive safety
attitude, and integrates safety with construction and quality goals (Jaselskis et
al., 1996; Tam et al., 2004; Teo et al., 2005). An organization with a good safety
culture always ensures that its staff are safety aware and properly trained, so that
they understand the consequences of unsafe acts (Lardner et al., 2001; INEEL,
2004).
2. Risk and hazard assessment: Risk assessment, including all potential risks (such
as accidents and injuries, regulatory issues, and environmental releases) should
be included in safety-planned activities (McDougall, 2004; Berg, 2006). As
organizational changes can have major effects on safety performance, an
auditable process of risk identification, analysis, and review is required to
manage those changes, and to maintain safety performance (Taylor, 2003).
3. Feedback on safety implementation: To achieve a good safety culture,
management should foster a climate that encourages feedback, so that
organizations learn from their experiences (ICAO, 1992). Tam et al. (2004)
A System Dynamics Approach to Construction Safety Culture
67
revealed that safety score in an organization, where feedback is given, is higher
than that in organizations where no feedback is given.
4. Adopting a no-blame approach: In a blame-free environment, workers feel that
they are fairly treated and are not blamed when they report safety incidents (as
those incidents are regarded as learning opportunities rather than occasions to
criticize or blame individuals) (Taylor, 2003). Vecchio-Sadus and Griffiths
(2004) supported the approach that a blame-free environment helps enhance the
CSC.
5. Site layout planning and good housekeeping: Cohen (1977) stated that better
housekeeping, more orderly plant operations, and adequate environmental
qualities are expected in the organizations with successful safety experiences.
An inadequate site layout plan may lead to the injury to construction personnel
or the public, along with damage to either property or the environment (Suraji et
al., 2001).
6. Site safety documentations (e.g. documented risk plans, site safety plans, site
accident logbooks, and minutes of site safety meetings): Pasman (2000)
identified the main elements of a safety management system as process
knowledge and documentation, the records of design criteria, and the records of
management decisions. Speirs and Johnson (2002) added that a good safety
culture organization would generate a substantial number of high quality
incident reports.
7. Having an effective benchmarking system: Taylor (2003) proposed a
benchmarking system as one of the key features of a strong safety culture. He
claimed that if the organization stops searching for new ideas of safety
improvements by means of benchmarking and seeking out best practice, there is
a danger that its safety culture will slip backwards.
3.4.6 Goals
Goals, with respect to employees, customers, society, and business performance,
represent the ultimate objectives an organization endeavours to achieve. This construct
is examined under seven attributes, namely level of job satisfaction, safe work
A System Dynamics Approach to Construction Safety Culture
68
behaviour, reduced number of accidents and safety related incidents, exceeded
customers’ expectations, improved industry image and safety standards, higher
workforce morale, and reduced total costs associated with accidents. Each attribute is
briefly described below:
1. Level of job satisfaction: A person with a high level of job satisfaction normally
holds positive attitudes towards the job, which thus, in turn, assists in reducing
work injuries (Paul and Maiti, 2007). To accomplish a higher job satisfaction,
improvement of safety-related issues is required (Grote and Kunzler, 2000).
Those safety improvements may include the encouragement of two-way
communication, adequate provision of safety training, and a supportive
environment in safety matters.
2. Safe work behaviour: Mohamed (2002) proposed that a higher level of safety
climate is positively associated with a higher level of self-reported safe work
behaviours. Further, Fang et al. (2006) highlighted that, by encouraging positive
safety behaviour and reducing negative behaviour, the safety climate of an
organization could be improved.
3. Reduced number of accidents and safety-related incidents: An organization with
positive safety culture usually has an acute awareness of the high-risk, error
prone nature of its work, which may lead to the reduction of accidents
(Sorensen, 2002; McDougall, 2004; Ho and Zeta, 2004). Teo et al. (2005) stated
that when safety aspects are well managed, the frequency of accident
occurrences may be reduced.
4. Exceeded customers’ expectations: Customer perspective represents the product
of safety culture (Mohamed, 2003). This perspective may be perceived by level
of customer satisfaction, customer feedback, and customer’s expectations. To
provide continuous improvement and maintain the standards expected by
customers, this customer perspective is acquired directly through opinions
expressed during meetings, and indirectly through other points of contact
(Wright et al., 1999).
5. Improved the industrial image and safety standards: An organization with a
good safety performance has a better organization image (Tang et al., 2003).
This image may be assessed by measurement of the public trust using attitude
A System Dynamics Approach to Construction Safety Culture
69
surveys, or by public reaction to statements made by an organization on its
safety performance (Wright et al., 1999).
6. Higher workforce morale: Employees are more likely to participate in safety
activities when there is a positive organizational culture within the workplace, as
reflected in good work relations and morale, and level of control over work
(Wallace and Neal, 2000). According to Mohamed (2003), workforce morale
can be enhanced by the recognition of individuals with an excellent safety
performance.
7. Reduced total costs associated with accidents: Little (2002) stated that ignorance
of health and safety commitments leads to economic risks for organizations. The
improvement of safety culture helps reduce the social costs of accidents usually
borne by the society (such as cost of property losses, cost of accidents and
injuries, cost of adverse publicity, and cost of environmental releases), and the
total costs of accidents usually borne by the organization (such as cost of lost
production, plant damage, and lost time through accidents) (Pasman, 2000; Tang
et al., 2003).
The above definitions of the six constructs (five enablers and Goals), and their 34
associated attributes, are summarized in Table 3.7. These constructs, as well as their
attributes, are later used, in this study, to develop a so-called construction safety culture
index (CSC index), which serves as an indicator for assessing the CSC maturity level in
the organization. It is important that an organization be able to assess its current
maturity level, as the type of improvement method, needed to support safety culture
development, differs as safety culture matures (Lardner et al., 2001). Consequently, a
safety improvement method may fail if it is not matched to the maturity of the
organization’s existing safety culture. The next section details the development of a
safety culture maturity model and its five levels of culture maturity.
A System Dynamics Approach to Construction Safety Culture
70
Table 3.7 Six model constructs and their 34 attributes
Construct Attribute
Leadership (Lds) 1. Commitment
2. Communication
3. Accountability
4. Leading by example
Policy and Strategy (Pol) 5. Safety awareness
6. Safety and productivity alignment
7. Safety standards
8. Safety initiatives
9. Safety integration in business goals
People (Ppl) 10. Shared perceptions
11. Safety responsibilities
12. Supportive environment
13. Workers’ involvement
14. Workers’ relationships
15. Workload
16. Work pressure
Partnerships and Resources (Prs) 17. Stakeholders’ cooperation
18. Financial resources
19. Safety resources
20. Human resources
Processes (Pro) 21. Training
22. Risk assessment
23. Feedback
24. No-blame approach
25. Housekeeping
26. Safety documentation
27. Benchmarking system
Goals 28. Job satisfaction
29. Safe work behaviour
30. Number of accidents
31. Customers’ expectations
32. Industrial image
33. Workforce morale
34. Cost of accidents
A System Dynamics Approach to Construction Safety Culture
71
3.5 SAFETY CULTURE MATURITY MODEL
3.5.1 Safety Culture Maturity Levels
Lardner et al. (2001) developed a safety culture maturity model (as shown in Figure 3.7)
based on the capability maturity model, to be used as a tool to assist organizations in
establishing their current level of safety culture maturity, and in identifying actions
required to improve their safety culture. The model consists of 10 elements: 1) visible
management commitment; 2) safety communication; 3) productivity versus safety; 4)
learning organization; 5) health and safety resources; 6) participation in safety; 7)
shared perception about safety; 8) trust between management and staff; 9) industrial
relations and job satisfaction; and 10) safety training. It is likely that an organization
will be at different levels in these 10 elements.
The safety culture maturity model consists of five levels of maturity (emerging,
managing, involving, cooperating, and continually improving). Deciding which level is
most appropriate needs to be based on the average level achieved by the organization or
site being evaluated. It is suggested that organizations progress sequentially through the
five levels, by building on the strengths, and removing the weaknesses of the previous
level. It is, therefore, not advisable for an organization to attempt to jump or skip a
level. For example, it is important for organizations to go through the managing level
before the involving level, as it is important that managers develop their commitment to
safety and understand the need to involve frontline employees.
A System Dynamics Approach to Construction Safety Culture
72
Em erging Level 1
Cooperating Level 4
Involving Level 3
M anaging Level 2
Continually improving
Level 5
Impr
ovin
g sa
fety
cul
ture
Incr
easin
g co
nsist
ency
Engage all staff to develop cooperation and commitment to
improving safety
Develop consistency and fight complacency
Realise the importance of frontline staff and develop personal
responsibility
Develop management commitment
Figure 3.7 Safety culture maturity model (Lardner et al., 2001)
A brief description of each maturity level is described below.
� Level 1 – Emerging level: At this first level, safety is defined in terms of technical
and procedural solutions and compliance with regulations. Safety is not seen as a
key business risk, and the safety department is perceived to have primary
responsibility for safety. Many accidents are seen as an unavoidable, and as a part of
the job. Most frontline staff are uninterested in safety, and may only use safety as
the basis for other arguments, such as changes in shift systems.
� Level 2 – Managing level: At this level, safety is seen as a business risk, and
management time and effort is put into accident prevention. Safety is solely defined
in terms of adherence to rules and procedures, and engineering controls. Accidents
are seen as preventable. Managers perceive that the majority of accidents are solely
caused by unsafe behaviours of frontline staff. Safety performance is measured in
terms of lagging indicators (such as lost time injuries), while safety incentives are
based on reducing those lagging indicator rates. In addition, senior managers are
reactive in their involvement in health and safety, i.e. they use punishment when
accident rates increase.
A System Dynamics Approach to Construction Safety Culture
73
� Level 3 – Involving level: At this level, the organization is convinced that the
involvement of frontline staff in health and safety is critical if future improvements
are going to be achieved. Managers recognize that wide ranges of factors cause
accidents, and that the root causes often originate from management decisions. A
significant proportion of frontline staff are willing to work with management to
improve health and safety. Further, majority of staff accept personal responsibility
for their own health and safety. Safety performance is actively monitored, and the
data is used effectively.
� Level 4 – Cooperating level: At this level, the majority of staff in the organization is
convinced that health and safety is important from both a moral and economic point
of view. Managers and frontline staff recognize that wide ranges of factors cause
accidents, and that the root causes are likely to come back to management decisions.
Frontline staff accept personal responsibility for their own, and others, health and
safety. The importance of all employees feeling valued and treated fairly is
recognized. The organization puts a significant effort into proactive measures to
prevent accidents. Additionally, safety performance is actively monitored using all
data available. Non-work accidents are also monitored, and a healthy lifestyle is
promoted.
� Level 5 – Continually improving level: At this final level, the prevention of all
injuries or harm to employees (both at work and at home) is a core company value.
The organization uses a range of indicators to monitor performance, but it is not
performance-driven, as it has confidence in its safety processes. The organization is
constantly striving to improve, and find better ways of improving hazard control
mechanisms. All employees share the belief that health and safety is a critical aspect
of their job, and accept that the prevention of non-work injuries is important. The
company invests considerable effort in promoting health and safety at home.
The safety culture maturity model helps senior management in planning and designing a
safety culture improvement initiative appropriated to their local needs and
circumstances. It provides a systematic process to help senior managers understand key
organizational and behavioural aspects of safety, prioritize areas for safety
improvement, and plan how to make those improvements.
A System Dynamics Approach to Construction Safety Culture
74
3.5.2 Scoring Each Maturity Level
To be able to assess the level of CSC maturity, each maturity level needs a score-range
(zero to 1,000 points). According to the EFQM (1998), the total score of 1,000 points
could be divided into five levels, as follows:
� Uncommitted: The score ranges from 0 to 249 points.
� Drifters: The score ranges from 250 to 499 points.
� Improvers: The score ranges from 500 to 749 points.
� Award winners: The score ranges from 750 to 999 points.
� World-class: This level has a single score of 1,000 points.
Many researchers, however, report the use of the EFQM Excellence model with a
number of different levels and respective score ranges. Dale and Smith (1997), for
example, divided the total of 1,000 points into six levels with the score-ranges, as
presented below. They suggested that the levels were a useful way of characterizing
organizations, and helping them to recognize symptoms and develop plans for the
future.
� Unaware or uncommitted: The score-range is between zero - 99 points.
� Initiators: The score-range is between zero - 299 points (covers the immediate
previous and next levels).
� Drifters: The score-range is between 100 - 299 points.
� Improvers: The score-range is between 300 - 649 points.
� Award winners: The score-range is between 650 - 749 points.
� World-class: The score-range is between 750 - 1,000 points.
Ahmed et al. (2003), on the other hand, allocated the 1,000 points, based on the EFQM
Excellence model and the interviews with senior managers and consultants, into seven
levels, to be used as a quality self-assessment. Those levels are:
A System Dynamics Approach to Construction Safety Culture
75
� Uncommitted: The score is between zero - 149 points.
� Drifter: The score is between 150 - 299 points.
� Tool pusher: The score is between 300 - 499 points.
� Improver: The score is between 500 - 649 points.
� Award winner: The score is between 650 and 849 points.
� World-class: The score is between 850 and 999 points.
� Supertitive: This level has a single score of 1,000 points.
In view of the score-range diversity listed above, the author decided to use five levels of
safety culture maturity, as represented in Figure 3.7, with each level having a score-
range of 200 points (as shown below):
� Emerging level: A score ranges between zero - 200 points.
� Managing level: A score ranges between 201 - 400 points.
� Involving level: A score ranges between 401 - 600 points.
� Cooperating level: A score ranges between 601 - 800 points.
� Continually improving level: A score ranges between 801 - 1,000 points.
These score ranges are later used, together with the CSC index developed through SD
modelling (described in Chapter 6), to identify the CSC maturity level in the
organization.
3.6 SUMMARY
In this chapter, three performance measurement systems, including the MBNQA
framework, the BSC framework, and the EFQM Excellence model were considered as
they were likely to be used in developing the CSC model. The advantages and
disadvantages of each performance measurement system were compared, and the
EFQM Excellence model was selected for the CSC model development. Using the
EFQM Excellence model as the basic framework, a CSC model was proposed. It
A System Dynamics Approach to Construction Safety Culture
76
consisted of six constructs, including five enablers (Leadership, Policy and Strategy,
People, Partnerships and Resources, and Processes), and a single set of Goals, to
represent the basic elements of the proposed model.
The proposed CSC model comprised 34 attributes to operationally define its six
constructs (five enablers and Goals). These attributes were used in developing a
questionnaire survey to elicit respondents’ opinions on the different attributes in the
context of their current safety practices and performance. The details of questionnaire
survey development, data collection, as well as the preliminary analyses, are explained
in the next chapter.
A System Dynamics Approach to Construction Safety Culture
77
44 DDAATTAA CCOOLLLLEECCTTIIOONN AANNDD PPRREELLIIMMIINNAARRYY AANNAALLYYSSEESS
4.1 GENERAL OVERVIEW
The 34 attributes derived in Chapter 3 were used in developing a questionnaire survey,
which is described in this chapter. Response rates, as well as the sample characteristics
are demonstrated. Then the preliminary analyses, including the handling of missing
data, the normality test, the outliers test, and the reliability test were performed to
increase confidence in the data collected.
4.2 QUESTIONNAIRE SURVEY
As stated in Section 2.2.3, this study uses the questionnaire survey to facilitate the
collection of information from construction organizations. Selvanathan and Selvanathan
(2005) stated that a good questionnaire survey could influence a high response rate. The
longer the questionnaire, the lower both the response rate and the quality of the data
collected. They suggested that, in developing the questionnaire, the researcher should:
� Keep the questionnaire as short as possible. This approach encourages respondents
to complete it.
� Ask short, simple, and clearly worded questions, to enable respondents to answer
quickly, correctly, and without ambiguity.
� Start with simple questions to help respondents get started comfortably.
� Arrange the questions in logical order.
� Avoid using leading questions.
� Include a covering letter, which explains the purpose of the survey.
A System Dynamics Approach to Construction Safety Culture
78
� Ensure that the questionnaire does not violate any ethical issues.
The questionnaire survey in this study contained a total of six pages, which could be
considered appropriate (not too long). It comprised two parts. The first part was devoted
to gathering demographical information about the respondents and their respective
organizations to ensure that the respondents have the appropriate backgrounds, and to
determine fundamental organizational safety issues including safety performance
records compared to their peers, safety policy and initiatives, reporting systems, and
devoted safety resources. This part contained nine questions: four open-ended questions
and five partially open-ended questions. Consequently, it was useful in identifying
discrepancies in the received responses.
The second part of the questionnaire covered 34 statements (representing the 34 CSC
attributes, see Table 3.7), to operationally define the six constructs (five enablers and
Goals) of the proposed CSC model. Each statement was designed to elicit respondents’
opinions on the different attributes in the context of their current safety practices and
performance using a five-point Likert scale, with point 1 representing ‘strongly
disagree’ and point 5 representing ‘strongly agree’. This approach enabled the
evaluation of the organization’s perception of, and commitment towards, each construct
to be carried out (see the questionnaire survey in Appendix 1).
With help from the Department of Labour Protection and Welfare, Ministry of Labour,
Thailand, a list of more than 150 medium to large construction organizations, with more
than 100 staff, was prepared and used as the sampling frame. The targeted respondents
were selected on the assumption that they held senior appointments (such as executive
directors, managing directors, and senior project managers) within their respective
organizations to capture a macro-level perspective of safety culture. The questionnaire
survey was both mailed and handed directly to targeted organizations.
A System Dynamics Approach to Construction Safety Culture
79
4.3 SAMPLE CHARACTERISTICS
Two hundred and twenty questionnaires were distributed, with 118 responses
represented a response rate of 53.6% (as research experts have argued that mail surveys
may not be reliable unless they either achieve a minimum of 50% response, or
demonstrate with some form of verification that the nonrespondents are similar to the
respondents) (Erdos, 1970). Up to three usable feedback questionnaires were chosen
from each organization to avoid bias in the data. From the returned responses, only three
were deemed unusable, due to unanswered items (data incompleteness or response
discrepancy), and were subsequently dropped from the data set. As a result, 115 usable
questionnaires provided data for 101 companies for the analyses (Appendix 2 contains
all raw data pertaining to questionnaire results).
�
As shown in Figures 4.1 and 4.2, 75.7% of the respondents had more than five years
working experience in the local Thai construction industry, and 59.1% had been
working for their present organization for at least five years. This result indicates the
reasonably high work experience rate of the respondents.
24.318.3
57.4
0.0
10.0
20.0
30.0
40.0
50.0
60.0
70.0
80.0
90.0
100.0
0-5 6-10 >10
Years in the construction industry
Res
pond
ents
(%)
Figure 4.1 Years of experience in the Thai construction industry
A System Dynamics Approach to Construction Safety Culture
80
40.9
19.1
40.0
0.0
10.0
20.0
30.0
40.0
50.0
60.0
70.0
80.0
90.0
100.0
0-5 6-10 >10
Years in the present organization
Res
pond
ents
(%)
Figure 4.2 Years of experience in the present organization
As shown in Figure 4.3, all respondents held senior positions in their organizations.
Most (83%) were also involved actively in site operations as they worked as project
managers, site managers, and safety managers.
Safety manager37%
Project manager26%
Site manager20%
Director17%
Figure 4.3 Job titles of the respondents
A System Dynamics Approach to Construction Safety Culture
81
Most of the respondents had safety responsibilities in planning and auditing safety on
site (see Figure 4.4). In addition, more than 50% of the respondents engaged in safety
related activities (such as training, and meetings reporting) at least once a month (see
Figure 4.5). These figures indicate the positive involvement of the respondents in a wide
range of safety activities.
Planning37%
Supervising23%
Auditing32%
Others8%
Figure 4.4 Safety responsibilities
Every month51%
Irregular23%
Every year17%
Every six months9%
Figure 4.5 Safety activities engagement
A System Dynamics Approach to Construction Safety Culture
82
Almost all of the respondents (93%) reported that their organizations had a formal
safety policy (see Figure 4.6), thus proving the appropriateness of the sampled
organizations involved in the survey. Also, more than 80% of the respondents believed
that their organization’s safety performance was at least as good as the national average
safety record (see Figure 4.7). These results (Figures 4.6 and 4.7) give confidence in the
suitability of the sampled companies to reflect the correct practices of the Thai
construction industry.
No7%
Yes93%
Figure 4.6 Formal safety policy in the organization
Better29%
Same52%
Worse19%
Figure 4.7 Safety performance compared to the national average record
A System Dynamics Approach to Construction Safety Culture
83
The majority of respondents ranked People as the most influential enabler for
significantly improving safety culture; both in the organization and in the construction
industry (see Figure 4.8). This result was consistent with the research findings of
Pipitsupaphol and Watanabe (2000), who investigated the root causes of labour
accidents in the Thai construction industry. They concluded that the major immediate
causes of accidents relate to unsafe acts of workers (such as not wearing personal
protective equipment). Further, they maintained that improvements in these aspects
could reduce more than one-third of the accidents.
14.3
24.1
33.9
18.8
8.9
22.328.6
32.1
7.1 9.8
0.0
10.0
20.0
30.0
40.0
50.0
60.0
70.0
80.0
90.0
100.0
Leadership Policy andStrategy
People Partnershipsand Resources
Processes
CSC Enablers
Res
pond
ents
(%)
The Industrial Level The Organizational Level
Figure 4.8 The most influential enablers in improving safety culture
In addition to the People enabler, the respondents ranked Policy and Strategy and
Leadership as the other two important enablers for improving safety culture. Leaders,
therefore, should motivate their team members to achieve safety goals (Northouse,
1997). The survey respondents also considered Partnerships and Resources as a
significant factor in improving safety culture at the industrial level, but not at the
organizational level. This outcome may result from the organization not being able to
provide adequate, and necessary, safety resources (such as personal protective
equipment) to all workers because of employee turnover, and the addition of and
released from the project team in response to the work schedules (workforce instability).
A System Dynamics Approach to Construction Safety Culture
84
In conclusion, organizations that participated in the survey claimed that they had safety
policies and safety implementations in place. The respondents had a high level of
experience in the construction industry, were in senior positions, and were thus able to
drive safety improvements.
4.4 DATA SCREENING AND PRELIMINARY ANALYSES
After the data was collected, a number of data examination techniques, ranging from the
simple process of visual inspection of graphical displays to statistical methods. Thus,
statistical methods of the handling of missing data, the normality test, the outliers test,
and the reliability test needed to be performed to increase confidence in the data. Each
statistical method is described in detail below.
4.4.1 Handling Missing Data
Missing data is one of the most pervasive problems in data analysis. Its seriousness
depends on the pattern of missing data, how much is missing, and why it is missing.
According to Tabachnick and Fidell (2007), the pattern of missing data is more
important than the amount missing. Thus, missing values scattered randomly through a
data matrix pose less serious problems than non-randomly missing values, which are
serious, no matter how few they are, because they affect the generalizability of the
results. There are a number of methods used to handle missing data, such as deleting
cases, using mean substitution, using a missing data correlation matrix, and treating
missing data as data. Tabachnick and Fidell (2007), however, claimed that if only 5% or
less of the data points are missing in a random pattern from a large data set, the
problems are less serious, and that almost any procedures for handling missing data
yield similar results. Table 4.1 illustrates the percentage of the missing values for the 34
items (attributes) within the six CSC constructs.
A System Dynamics Approach to Construction Safety Culture
85
Table 4.1 Missing values
No. Attribute (Item) Usable Case(s) Missing Data (Total Cases) Count Percent
1. Commitment 114 1 0.9
2. Communication 114 1 0.9
3. Accountability 115 - -
4. Leading by example 115 - -
5. Safety awareness 113 2 1.7
6. Safety and productivity alignment 114 1 0.9
7. Safety standards 112 3 2.6
8. Safety initiatives 114 1 0.9
9. Safety integration in business goals 114 1 0.9
10. Shared perceptions 115 - -
11. Safety responsibilities 115 - -
12. Supportive environment 114 1 0.9
13. Workers’ involvement 113 2 1.7
14. Workers’ relationships 115 - -
15. Workload 114 1 0.9
16. Work pressure 115 - -
17. Stakeholders’ cooperation 113 2 1.7
18. Financial resources 113 2 1.7
19. Safety resources 114 1 0.9
20. Human resources 115 - -
21. Training 113 2 1.7
22. Risk assessment 113 2 1.7
23. Feedback 115 - -
24. No-blame approach 113 2 1.7
25. Housekeeping 112 3 2.6
26. Safety documentation 114 1 0.9
27. Benchmarking system 115 - -
28. Job satisfaction 112 3 2.6
29. Safe work behaviour 115 - -
30. Number of accidents 115 - -
31. Customers’ expectations 114 1 0.9
32. Industrial image 111 4 3.5
33. Workforce morale 114 1 0.9
34. Cost of accidents 115 - -
A System Dynamics Approach to Construction Safety Culture
86
None of the 34 items (attributes) had more than 5% of their values missing (with the
highest percentage is of the ‘industrial image’ item, see Table 4.1). Thus, any ‘handling
missing data’ method could be employed. For this study, the ‘mean substitution’ method
was chosen.
4.4.2 Test of Normality
The screening of continuous variables for normality is an important early step in almost
every multivariate analysis (Tabachnick and Fidell, 2007). Although the normality of
the variables is not always required for an analysis, the solution is usually more
appropriate if the variables are all normally distributed. For this reason, the normality of
the variables is assessed by either statistical or graphical methods.
Two important components of normality are skewness and kurtosis (Tabachnick and
Fidell, 2007). Skewness relates to the symmetry of the distribution; a skewed variable is
a variable whose mean is not in the centre of the distribution. Kurtosis, on the other
hand, relates to the peakedness of a distribution; a distribution is either too peaked (with
short, thick tails), or too flat (with long, thin tails). When a distribution is normal, the
values of skewness and kurtosis are zero (Pallant, 2005). If there is a positive skewness,
there is a pileup of cases to the left, and the right tail is too long; with negative
skewness, the result is reversed. Kurtosis values above zero indicate a distribution that
is too peaked, while kurtosis values below zero are reversed. Non-normal kurtosis
produces an underestimate of the variance of a variable.
According to Morgan and Griego (1998), if the division of statistics values (Stat.) of
skewness (or kurtosis) and its standard error (S.E.) are not above 5.5, then that skewness
(or kurtosis) is not significantly different from normal. Curran et al. (1996), however,
recommend that the values of skewness < 2.0 and kurtosis < 7.0 are acceptable. Table
4.2 shows the skewness and kurtosis values of the 34 attributes. The results demonstrate
that all 34 attributes show normal distribution, thus increasing confidence in the data.
A System Dynamics Approach to Construction Safety Culture
87
Table 4.2 Skewness and kurtosis of the 34 attributes
No. Attribute (Item) Skewness Kurtosis Stat. S.E. Stat./S.E. Stat. S.E. Stat./S.E.
1. Commitment -1.16 .226 -5.13 1.81 .449 4.03
2. Communication -0.59 .226 -2.61 0.25 .449 0.56
3. Accountability -0.95 .226 -4.20 1.25 .447 2.80
4. Leading by example -1.47 .226 -6.50 2.15 .447 4.81
5. Safety awareness -0.76 .227 -3.35 0.23 .451 0.51
6. Safety and productivity alignment -0.89 .226 -3.94 0.64 .449 1.43
7. Safety standards -0.69 .228 -3.03 0.46 .453 1.02
8. Safety initiatives -1.00 .226 -4.42 1.42 .449 3.16
9. Safety integration in business goals -0.75 .226 -3.32 0.35 .449 0.78
10. Shared perceptions -0.50 .226 -2.21 0.27 .447 0.60
11. Safety responsibilities -0.19 .226 -0.84 -0.14 .447 -0.31
12. Supportive environment -0.78 .226 -3.45 1.31 .449 2.92
13. Workers’ involvement -0.49 .227 -2.16 -0.76 .451 -1.69
14. Workers’ relationships -0.42 .226 -1.86 0.17 .447 0.38
15. Workload -0.77 .226 -3.41 0.10 .449 0.22
16. Work pressure -0.54 .226 -2.39 -0.12 .447 -0.27
17. Stakeholders’ cooperation -0.76 .227 -3.35 0.46 .451 1.02
18. Financial resources -0.32 .227 -1.41 -0.45 .451 -1.00
19. Safety resources -0.72 .226 -3.19 0.05 .449 0.11
20. Human resources -0.79 .226 -3.50 0.81 .447 1.81
21. Training -0.39 .227 -1.72 -0.33 .451 -0.73
22. Risk assessment -0.51 .227 -2.25 -0.42 .451 -0.93
23. Feedback -0.59 .226 -2.61 0.00 .447 0.00
24. No-blame approach -0.43 .227 -1.89 -0.35 .451 -0.78
25. Housekeeping -0.48 .228 -2.11 -0.02 .453 -0.04
26. Safety documentation -1.16 .226 -5.13 1.34 .449 2.98
27. Benchmarking system -0.69 .226 -3.05 -0.28 .447 -0.63
28. Job satisfaction -1.02 .228 -4.47 1.81 .453 4.00
29. Safe work behaviour -0.88 .226 -3.89 1.06 .447 2.37
30. Number of accidents -0.93 .226 -4.12 1.40 .447 3.13
31. Customers’ expectations -0.66 .226 -2.92 0.51 .449 1.14
32. Industrial image -1.04 .229 -4.54 1.34 .455 2.95
33. Workforce morale -1.01 .226 -4.47 1.73 .449 3.85
34. Cost of accidents -1.28 .226 -5.66 2.03 .447 4.54
Note: Stat. = Statistics values, S.E. = Standard error values
A System Dynamics Approach to Construction Safety Culture
88
4.4.3 Outliers Test
An outlier is a case with such an extreme value on one variable (a univariate outlier), or
such a strange combination of scores on two or more variables (multivariate outlier),
that distorts the statistical results (Tabachnick and Fidell, 2007). There are many ways
to test for outliers, such as the use of the 5% trimmed mean, the use of standardized
scores (z-scores), and the use of boxplots (Pallant, 2005; Tabachnick and Fidell, 2007).
In this study, however, the mean, the 5% trimmed mean, and the z-score tests were used
to detect outliers.
4.4.3.1 5% Trimmed Mean
The 5% trimmed mean is a mean calculated from the cases in which 5% of the top and
the bottom of the cases are removed (Pallant, 2005). According to Pallant (2005), the
big difference (> 0.2) between a mean and its 5% trimmed mean may indicate a problem
with an outlier. Table 4.3 illustrates the means, the 5% trimmed means, and the standard
deviations (S.D.) of the 34 CSC attributes. The results show that the mean differences
(�mean) of all attributes are small, providing support for the absence of outliers.
4.4.3.2 Z-Score
To further detect outliers, a standardized score (z-score) test was performed. According
to Tabachnick and Fidell (2007), the cases with z-scores that exceed 3.29, at p < 0.01,
two-tailed test, are the potential outliers. There were 12 z-scores exceeding 3.29, in
which most were from case number ‘76’ (data file number 76, see Appendix 3). As a
result, case number ‘76’ was deleted from the data file, leading to a total of 114 data
retained for further analyses.
A System Dynamics Approach to Construction Safety Culture
89
Table 4.3 The mean, the 5% trimmed mean, and the standard deviation of the 34
attributes
Attribute (Item) Mean 5% Trimmed Mean ����Mean S.D. Commitment 4.19 4.27 0.08 0.819
Communication 3.66 3.70 0.04 0.948
Accountability 4.01 4.07 0.06 0.832
Leading by example 4.25 4.36 0.11 0.926
Safety awareness 3.61 3.68 0.07 1.073
Safety and productivity alignment 3.80 3.87 0.07 0.997
Safety standards 3.81 3.87 0.06 0.916
Safety initiatives 3.82 3.89 0.07 0.895
Safety integration in business goals 3.78 3.85 0.07 1.011
Shared perceptions 3.92 3.96 0.04 0.751
Safety responsibilities 3.72 3.74 0.02 0.732
Supportive environment 3.99 4.04 0.05 0.781
Workers’ involvement 3.73 3.76 0.03 0.813
Workers’ relationships 3.67 3.70 0.03 0.835
Workload 3.83 3.89 0.06 0.861
Work pressure 3.77 3.80 0.03 0.921
Stakeholders’ cooperation 3.63 3.67 0.04 0.928
Financial resources 3.72 3.75 0.03 0.940
Safety resources 3.88 3.93 0.05 0.942
Human resources 3.94 4.01 0.07 0.891
Training 3.81 3.86 0.05 0.921
Risk assessment 3.47 3.50 0.03 1.036
Feedback 3.80 3.84 0.04 0.910
No-blame approach 3.21 3.24 0.03 1.097
Housekeeping 3.75 3.78 0.03 0.800
Safety documentation 3.93 4.02 0.09 0.984
Benchmarking system 3.30 3.34 0.04 1.036
Job satisfaction 3.65 3.70 0.05 0.813
Safe work behaviour 3.86 3.91 0.05 0.826
Number of accidents 4.11 4.17 0.06 0.803
Customers’ expectations 3.89 3.94 0.05 0.849
Industrial image 3.79 3.85 0.06 0.865
Workforce morale 3.81 3.86 0.05 0.819
Cost of accidents 4.02 4.10 0.08 0.917
A System Dynamics Approach to Construction Safety Culture
90
4.4.4 Scale Reliability (Cronbach’s Alpha)
When selecting scales to include in the study, it was important to find the scales that
were statistically reliable. The scale reliability, the proportion of variance attributable to
the true score of latent variable (Pallant, 2005), can be defined as the extent to which a
measure produces similar results over different occasions of the data collection (Seo et
al., 2004). It was thus essential to examine the reliability of each scale whenever a
measurement was involved.
One of the main issues in scale reliability concerns the scale’s internal consistency
(Cronbach’s alpha, �). In a good solution, Cronbach’s alpha (�) ranges between zero
and one - the larger the value, the more stable the factors. A high value means that the
observed variables account for substantial variance in the factor scores, while a low
value means the factors are poorly defined by the observed variables. Generally, the
value of 0.70 is accepted as the minimum desired value of reliability (Pallant, 2005).
In this study, the 34 attributes within the six CSC constructs were tested for internal
consistency, using the retained 114 data. The results, shown in Table 4.4, had values
ranging from 0.83 to 0.89, all of which were considered acceptable. This thus increases
confidence in the contribution of the 34 attributes to the measurement of their respective
constructs.
A System Dynamics Approach to Construction Safety Culture
91
Table 4.4 Internal consistency of the five enablers and Goals
Construct ����
1. Leadership Enabler
(Attributes: commitment, communication, accountability, and leading by
example)
0.837
2. Policy and Strategy Enabler
(Attributes: safety awareness, safety and productivity alignment, safety
standards, safety initiatives, and safety integration in business goals)
0.881
3. People Enabler
(Attributes: shared perceptions, safety responsibilities, supportive environment,
workers’ involvement, workers’ relationships, workload, and work pressure)
0.858
4. Partnerships and Resources Enabler
(Attributes: stakeholders’ cooperation, financial resources, safety resources, and
human resources)
0.874
5. Processes Enabler
(Attributes: training, risk assessment, feedback, no-blame approach,
housekeeping, safety documentation, and benchmarking system)
0.852
6. Goals
(Attributes: job satisfaction, safe work behaviour, number of accidents,
customers’ expectations, industrial image, workforce morale, and cost of
accidents)
0.893
To further confirm this finding, as well as to gain a better understanding of the factor
structure of the CSC scale as a preliminary step toward the SEM, the EFA was
conducted (see following chapter).
A System Dynamics Approach to Construction Safety Culture
92
[THIS PAGE IS INTENTIONALLY LEFT BLANK]
A System Dynamics Approach to Construction Safety Culture
93
55 EEXXPPLLOORRAATTOORRYY FFAACCTTOORR AANNAALLYYSSIISS AANNDD
SSTTRRUUCCTTUURRAALL EEQQUUAATTIIOONN MMOODDEELLLLIINNGG
5.1 GENERAL OVERVIEW
Preliminary analyses were conducted to increase confidence in the data (see Chapter 4).
Following on from that work, an exploratory factor analysis (EFA) was performed to
extract attributes into a number of factors that represent the interrelations among the set
of those attributes (as discussed in the following section). The attributes associated with
the five enablers of the proposed CSC model were analysed with the EFA to confirm
the construct validity of those five enablers. To further confirm the construct validity,
and to examine the causal relationships between the six constructs (five enablers and
Goals), structural equation modelling (SEM) was performed, using the AMOS program.
The final CSC model was achieved, as reported at the end of this chapter
The exploratory factor analysis, along with its details, is presented in the next section.
5.2 THE EXPLORATORY FACTOR ANALYSIS
The exploratory factor analysis (EFA) method is often used in the early stages of data
analysis to gather information about interrelationships among a set of variables.
According to Seo et al. (2004), the EFA is a precursor to the SEM. When conducting an
EFA, three main steps are followed: 1) the assessment of the suitability of the data; 2)
the factor extraction; and 3) the factor rotation (Pallant, 2005). The details of each step
are described below.
A System Dynamics Approach to Construction Safety Culture
94
5.2.1 Assessment of the Suitability of the Data for the Analysis
Two main issues facilitate the determination of whether a particular data set is suitable
for factor analysis. The first issue is the sample size. Tabachnick and Fidell (2007)
noted that it is comforting to have at least 300 cases for factor analysis. However, they
conceded that a smaller sample size (e.g. 150 cases) should be sufficient, if the solutions
have several high loading marker variables (above 0.80). Pallant (2005), on the other
hand, recommended that five cases for each item are adequate in most cases. Coakes
and Steed (2003) took a stance in between, arguing that a sample of 100 cases is
acceptable, with a sample of 200 or more cases being preferable. A total of (usable) 114
cases (with case number ‘76’ deleted) were considered acceptable for the analysis in
this study.
The second issue concerns the strength of the inter-correlations among the items.
Bartlett’s test of sphericity and the Kaiser-Meyer-Olkin (KMO) test are normally
applied to assess the factorability of the data (Pallant, 2005). Bartlett’s test of sphericity
should be significant (p < 0.05) for factor analysis to be considered appropriate, while
the KMO index should range from zero to one, with 0.6 suggested as the minimum
value for a good factor analysis (Tabachnick and Fidell, 2007).
In this study, Bartlett’s test of sphericity was significant (see Table 5.1), with the KMO
index being 0.91, thus indicating that the data was suitable for factor analysis.
Table 5.1 Bartlett’s test of sphericity and the KMO index
Test Recommended
Value
Calculated
Value
Bartlett's test of sphericity (significant) < 0.05 0.00
Kaiser-Meyer-Olkin measure of sampling adequacy (KMO) > 0.60 0.91
A System Dynamics Approach to Construction Safety Culture
95
5.2.2 Factor Extraction
Factor extraction, the second step in conducting the EFA, involves determining the
smallest number of factors that can be used to best represent the interrelations among
the set of variables (Tabachnick and Fidell, 2007). There are a variety of approaches
that can be used to extract (or identify) the number of underlying factors. Some of the
most commonly available extraction techniques are principal components, principal axis
factoring, and general least square (Pallant, 2005).
According to Coakes and Steed (2003), the most frequently used techniques are
principal components and principal axis factoring. The goal of both techniques is to
extract the maximum variance from the data set with each component. The principal
axis factoring is, however, widely used, and conforms to the factor analytic model in
which common variance is analysed, with the unique and error variance removed
(Tabachnick and Fidell, 2007). For this reason, the principal axis factoring method was
chosen for the analysis.
It is also important to determine the number of factors requested for factor analysis.
Thus, according to Pallant (2005), three techniques are generally used to assist in
decisions concerning the number of factors to retain. These techniques are: 1) the
Kaiser’s criterion or the eigenvalue rule; 2) the Catell’s screed test; and 3) the Horn’s
parallel analysis. The eigenvalue is the value that represents the amount of total variance
explained by that factor. In this study, the eigenvalue over one (> 1.0) was used as the
criterion for extracting the factors of the CSC model (Tabachnick and Fidell, 2007).
5.2.3 Factor Rotation and Interpretation
After the extraction, factor rotation is ordinarily used to present the pattern of loadings
in a manner that is easy to interpret. Numerous methods of factor rotation are available,
but the most commonly used is ‘varimax’ (Coakes and Steed, 2003; Pallant, 2005).
A System Dynamics Approach to Construction Safety Culture
96
Varimax is a variance maximizing procedure. The goal of varimax rotation is to
maximize the variance of factor loadings by making high loadings higher, and low
loadings lower, for each factor (Tabachnick and Fidell, 2007). The varimax method was
used for the factor analysis in this study.
In summary, principal axis factoring, with an eigenvalue over one, together with the
varimax rotation method, were used to examine the dimensionality of the 27 attributes
of the CSC’s five enablers, and to achieve better interpretability of the factor loadings.
As the remaining seven attributes of the Goals construct were grouped together as one
factor, they were not included in the analysis. The results of the EFA are shown below.
5.2.4 The EFA Results
The 27 attributes of the CSC’s enablers were analysed for factor extraction. Due to the
sample size (114 data points), a cut-off factor loading of 0.45 was used to screen out the
attributes (or items) that were weak indicators of the constructs (as suggested by Hair et
al. (1998)). The first run thus led to dropping of the ‘no-blame approach’ item as it
failed to make the cut-off. Consequently, performing the factor analysis on the
remaining 26 items highlighted another problematic item, namely ‘shared perceptions’.
Based on an eigenvalue greater than one (> 1.0), the remaining 25 attributes were
extracted into three factors, which accounted for 59.73% of the total variance, as shown
in Table 5.2. Factor 1 was predominantly accounted for by nine items, initially
measuring Processes (Pro) and Partnerships and Resources (Prs); Factor 2 by 10 items,
initially measuring Leadership (Lds) and Policy and Strategy (Pol); and Factor 3 by six
items, initially measuring People (Ppl) and Partnerships and Resources (Prs).
A System Dynamics Approach to Construction Safety Culture
97
Table 5.2 Three factors extracted from the remaining 25 items
Item Factor Extracted
1 2 3
Safety resources .791
Risk assessment .759
Benchmarking system .670
Financial resources .651
Workers’ involvement .641
Safety documentation .587
Training .585
Feedback .536
Safety integration in business goals .535
Accountability .731
Safety and productivity alignment .703
Commitment .697
Communication .676
Safety initiatives .609
Leading by example .575
Safety awareness .550
Safety standards .494
Supportive environment .476
Workers’ relationships .465
Workload .775
Human resources .673
Housekeeping .597
Stakeholders’ cooperation .578
Work pressure .568
Safety responsibilities .481
Extraction Method: Principal Axis Factoring. Rotation Method: Varimax with Kaiser Normalization. Rotation converged in seven iterations.
A closer examination of the identified factors revealed the potential for further analysis
to extract the independent factors in line with those of the proposed CSC model (see
Figure 3.6). For this reason, it was decided to further factor-analyse the three factors,
setting their required extraction limit to two new factors each. The nine items of Factor
A System Dynamics Approach to Construction Safety Culture
98
1 (in Table 5.2) thus gave rise to two new factors, accounting for 62.58% of the total
variance, as shown in Table 5.3.
Table 5.3 Two factors extracted from nine items of Factor 1 of Table 5.2
Item Factor Extracted
1 2
Financial resources .792
Safety resources .761
Risk assessment .689
Workers’ involvement .535
Training .524
Benchmarking system .625
Safety integration in business goals .614
Feedback .610
Safety documentation .576
Extraction Method: Principal Axis Factoring. Rotation Method: Varimax with Kaiser Normalization. Rotation converged in three iterations.
Similarly, two factors were extracted from the 10 items of Factor 2 (in Table 5.2).
However, as the ‘supportive environment’ item failed to make the cut-off factor loading,
it was removed from the data file. The remaining nine items of Factor 2 were
reanalysed, and produced two new factors, accounting for 59.10% of the total variance
(see Tables 5.4). No new factors were extracted from Factor 3, thus this factor was
considered as a single construct.
A System Dynamics Approach to Construction Safety Culture
99
Table 5.4 Two factors extracted from nine items of Factor 2 of Table 5.2
Item Factor Extracted
1 2
Safety awareness .758
Safety standards .719
Safety initiatives .706
Workers’ relationships .515
Commitment .770
Communication .644
Accountability .624
Safety and productivity alignment .551
Leading by example .494
Extraction Method: Principal Axis Factoring. Rotation Method: Varimax with Kaiser Normalization. Rotation converged in three iterations.
In this study, five factors, within the remaining 24 items, were extracted from the EFA
(as summarized in Table 5.5), with Leadership (five items), Policy and Strategy (four
items), People (six items), Partnerships and Resources (five items), and Processes (four
items).
Interestingly, the above analysis led to nine items1, initially assumed to be associated
with a certain enabler, to strongly correlate with another enabler. To illustrate, the
‘safety and productivity alignment’ item appeared to be loading on the Leadership
enabler not the Policy and Strategy enabler, as was initially hypothesized.
1The nine relocated items were: 1) the ‘safety and productivity alignment’; 2) the ‘workers relationships’; 3) the
‘human resources’; 4) the ‘housekeeping’; 5) the ‘stakeholders cooperation’; 6) the ‘risk assessment’; 7) the ‘workers involvement’; 8) the ‘training’; and 9) the ‘safety integration in business goals’ items.
A System Dynamics Approach to Construction Safety Culture
100
Table 5.5 Five factors extracted from the EFA
Item Factors Extracted Lds Pol Ppl Prs Pro
Commitment .770
Communication .644
Accountability .624
Safety and productivity alignment* .551
Leading by example .494
Safety awareness .758
Safety standards .719
Safety initiatives .706
Workers’ relationships* .515
Workload .775
Human resources* .673
Housekeeping * .597
Stakeholders’ cooperation* .578
Work pressure .568
Safety responsibilities .481
Financial resources .792
Safety resources .761
Risk assessment* .689
Workers’ involvement* .535
Training * .524
Benchmarking system .625
Safety integration in business goals* .614
Feedback .610
Safety documentation .576
Note: * Items relocated to another enabler, Lds = Leadership, Pol = Policy and Strategy, Ppl = People, Prs = Partnerships and Resources, Pro = Processes
Following the re-allocation of the nine items, the Cronbach’s alpha (�) test was re-
applied to ensure the appropriateness of the groupings of the five factors extracted. As
shown in Table 5.6, the alpha coefficients ranged from 0.85 to 0.90, all of which were
considered highly reliable.
A System Dynamics Approach to Construction Safety Culture
101
Table 5.6 Internal consistency of five factors extracted from the EFA
Construct Cronbach’s Alpha (����)
1. Leadership Enabler
(Attributes: Commitment, Communication, Accountability, Safety
and productivity alignment, and Leading by example)
0.856
2. Policy and Strategy Enabler
(Attributes: Safety awareness, Safety standards, Safety initiatives,
and Workers’ relationships)
0.849
3. People Enabler
(Attributes: Workload, Human resources, Housekeeping,
Stakeholders’ cooperation, Work pressure, and Safety
responsibilities)
0.896
4. Partnerships and Resources Enabler
(Attributes: Financial resources, Safety resources, Risk assessment,
Workers’ involvement, and Training)
0.886
5. Processes Enabler
(Attributes: Benchmarking system, Safety integration in business
goals, Feedback, and Safety documentation)
0.873
6. Goals
(Attributes: Job satisfaction, Safe work behaviour, Number of
accidents, Customers’ expectations, Industrial image, Workforce
morale, and Cost of accidents)
0.893
5.2.5 Conclusion of the EFA
The EFA gave rise to a total of 24 attributes grouped to explain five latent factors
(enablers), whereas a total of seven attributes were grouped to explain the sixth latent
factor (Goals), leading to the so-called baseline model of the CSC (shown in Figure
5.1).
A System Dynamics Approach to Construction Safety Culture
102
PolSafety initiativesSafety standards
Safety and productivity alignment
Safety awareness
Safety integration in business goals
Ppl
Safety responsibilities
Workers' involvement
Workers' relationships
Prs
Human resources
Safety resourcesFinancial resources
Stakeholders' cooperation
Pro
Feedback
Training
Housekeeping
Safety documentation
GoalsCustomers' expectationsNumber of accidentsSafe work behaviour
Job satisfaction
Industrial imageWorkforce moraleCost of accidents
LdsLeading by example
AccountabilityCommunication
Commitment
WorkloadWork pressure
Benchmarking system
Risk assessment
Note: Lds = Leadership, Pol = Policy and Strategy, Ppl = People, Prs = Partnerships and Resources, Pro = Processes
Figure 5.1 Baseline model of the CSC
A System Dynamics Approach to Construction Safety Culture
103
The ovals and rectangles, shown in the baseline model, symbolise latent and observed
variables, respectively. The former represent the six constructs of the baseline model,
whereas the latter represent their respective attributes.
The arrows connecting the two sets of variables indicate the direction of the
hypothesized influence. For example, it is hypothesized that Leadership is manifested
by the achievement of its five attributes, namely: ‘commitment’, ‘communication’,
‘accountability’, ‘leading by example’, and ‘safety and productivity alignment’; the
arrows are thus shown originating from Leadership to each one of the five attributes.
To provide further evidence of the construct validity of the CSC model, and to
investigate the causal relationships between the five enablers and Goals of the CSC, the
structural equation modelling was conducted next.
5.3 THE STRUCTURAL EQUATION MODELLING
Theoretically, the SEM comprises two types of models: measurement and structural
models. The former is concerned with how well the observed variables measure the
latent factors, addressing their reliability and validity. The latter is concerned with
modelling the relationships between the latent factors, by describing the amount of
explained and unexplained variance, which is akin to the system of simultaneous
regression models (Wong and Cheung, 2005). The details of the measurement and
structural models are explained below.
A System Dynamics Approach to Construction Safety Culture
104
5.3.1 Measurement Model
Testing the structural model would have been meaningless until it was established as a
good measurement model. In this study, a confirmatory factor analysis (CFA) was
undertaken to establish confidence in the measurement model. The CFA specifies the
posited relations of the observed variables to the underlying constructs. It belongs to the
family of SEM techniques, as it allows for the assessment of fit between the observed
data and a priori conceptualized, theoretically grounded model that specifies the
hypothesized causal relationships between latent factors and their observed indicator
variables (Mueller and Hancock, 2004).
According to Byrne (2001), the measurement model should be assessed by five
methods: 1) the feasibility of parameter estimates; 2) the appropriateness of standard
errors (S.E.); 3) the statistical significance of parameter estimates; 4) model fit as a
whole (using goodness of fit, GOF, indices); and 5) square multiple correlation (SMC,
R2). Each method is described in detail below.
5.3.1.1 Feasibility of Parameter Estimates
Further, Byrne (2001) noted that the parameter estimates of a good measurement model
must exhibit the correct sign and size, and be consistent with the underlying theory. Any
estimates falling outside the admissible range signal a clear indication that either the
model is wrong or the input matrix lacks sufficient information. Examples of parameters
exhibiting unreasonable estimates are: 1) correlation values more than one; 2) negative
variances; and 3) covariance or correlation matrices that are not positive definite.
The analysis results of the measurement model, shown in Appendix 4, revealed that all
the parameter estimates were both reasonable and statistically significant, thus
confirming the construct reliability and validity of the baseline model.
A System Dynamics Approach to Construction Safety Culture
105
5.3.1.2 Appropriateness of Standard Errors (S.E.)
Joreskog and Sorbom (1993) confirmed that the standard errors that are excessively
large or small indicate a poor model fit. For instance, if a standard error approaches
zero, the test statistic for its related parameter cannot be defined. Likewise, standard
errors that are extremely large indicate parameters that cannot be determined.
Because the standard errors are influenced by the units of measurement in observed
and/or latent variables, as well as the magnitude of the parameter estimate itself, no
definitive criterion for small and large has been established (Joreskog and Sorbom,
1993). The results from this study demonstrated that all standard errors appeared to be
in good order, proving the construct validity of the baseline model (see Appendix 3).
5.3.1.3 Statistical Significance of Parameter Estimates
The critical ratio (C.R.) was used to test the statistical significance of the parameter
estimates. The C.R. represents the parameter estimate divided by its standard error. It
operates as a z-statistic in testing that the estimate is statistically different from zero.
Based on a probability level of 0.05, the test statistic needed to be > �1.96 before the
hypothesis (that the estimate equals to zero) could be rejected (Byrne, 2001). The results
showed that all the C.R. values of the latent and observed variables of the baseline
model were more than 1.96, therefore, the hypothesis (that the estimate equals to zero)
could be rejected (see Appendix 3).
5.3.1.4 Model Fit as a Whole
The assessment of the overall model fit is considered as a critical issue in relation to any
SEM. There are a number of indices that may be used in assessing the model fit. One of
A System Dynamics Approach to Construction Safety Culture
106
the most widely used fit indices is a model chi square (�2), which tests the closeness of
the fit between the sample covariance matrix and the fitted covariance matrix (Kline,
2005). However, since the formula for computing �2 is directly related to the sample
size, almost all the models are evaluated as incorrect as the sample size increases. For
this reason, the ratio of �2 to the degrees of freedom (�2/DF) has been commonly used
as an alternative fit index. Normally, the value of �2/DF less than two represents the
model as a good fit (Kline, 2005). Garson (2006), however, proposed that the value of
less than three is acceptable.
Another widely used fit index is the Root Mean Square Error of Approximation
(RMSEA) (Byrne, 2001). It has been recognized as one of the most informative criteria
in covariance structure modelling. It is the best fit index where models are very
parsimonious, as it measures the lack of fit per degree of freedom. According to
Tabachnick and Fidell (2007), the value of RMSEA up to 0.05 indicates a good model
fit, with a value up to 0.10 indicating a reasonable error of approximation.
Apart from the �2/DF and the RMSEA, an additional group of fit indices are commonly
used, including a Bentler-Bonett normed fit index (NFI), a comparative fit index (CFI),
an incremental fit index (IFI), a Tucker-Lewis index (TLI), and a relative fit index
(RFI)). Kline (2005) suggested that these indices’ value should be at least 0.90 to
indicate a model fit. Garson (2006), however, used the cut-off value of 0.80 as the
criterion. The GOF indices of the baseline model revealed a need to modify the model
in order to improve the model fit (see Table 5.7).
According to Kline (2005), there are potentially two possible options involved in the
process of model refinement. The first option is to eliminate the links or ‘paths’ with
very low correlations. This option was found not to be applicable to the baseline model
due to the high correlations of all paths (see Appendix 3). The second option is to
remove the observed variables shown by the computed modification indices as having
multicollinearity (e.g. the observed variables/their error variances that correlated to
A System Dynamics Approach to Construction Safety Culture
107
more than two other observed variables/error variances). Garson’s (2006) observations
are that signs of high multicollinearity are indicated by:
Table 5.7 The GOF indices of the baseline and the best-fit measurement models
GOF Index Recommended Level Baseline
Model
Best-Fit Measurement
Model
�2/DF < 2.00 (Byrne, 2001) 2.10 1.65
RMSEA � 0.10 (Tabachnick and Fidell, 2007) 0.10 0.07
NFI > 0.80 (Ullman, 2001) 0.71 0.81
CFI > 0.90 (Kline, 2005) 0.82 0.91
IFI > 0.80 (Garson, 2006) 0.82 0.91
TLI > 0.90 (Kline, 2005) 0.80 0.90
RFI > 0.80 (Garson, 2006) 0.68 0.78
Note: The baseline model is shown in Figure 5.1, while the best-fit measurement model is shown in Figure 5.2.
� Standardized regression weights: Since all latent variables in a SEM model have
been assigned a metric of one, all standardized regression weights should be within
the range of plus or minus one. When there is a multicollinearity problem, the
standardized regression weights may show the values of greater than one and/or less
than minus one.
� Standard errors of the unstandardized regression weights: As with the standardized
regression weights, the unstandardized values of greater than one and/or less than
minus one may indicate a sign of high multicollinearity.
� Covariances of the parameter estimates: The variables with high multicollinearity
may well be reflected in high covariances of the parameter estimates.
� Variance estimates: Negative error variance estimates may also be another effect of
the multicollinearity
According to the second option, seven observed variables that showed signs of high
multicollinearity were removed (see the modification indices in Appendix 3):
A System Dynamics Approach to Construction Safety Culture
108
� Three of those variables (‘workload’, ‘work pressure’, and ‘housekeeping’) were
from the People enabler.
� One variable, which was ‘risk assessment’, came from the Partnerships and
Resources enabler.
� Three variables (‘job satisfaction’, ‘safe work behaviour’, and ‘customers’
expectations’) were from Goals.
Further modifications appeared not to improve the model fit, thus leading to the best-fit
measurement model with the best GOF indices, as shown in Figure 5.2 and Table 5.7,
respectively. Table 5.7 highlights the significant improvement of the GOF indices’
values compared to those obtained for the baseline model.
5.3.1.5 Square Multiple Correlation (SMC, R2)
SMC (R2) is the extent to which a measurement model is adequately represented by the
observed measures. Each R2 is interpreted as the proportion of variance in the indicator
that is explained by the respective latent variable; this is a similar concept to the
communality estimate in the EFA (Ullman, 2001). Normally, the R2 0.5 is used as an
indicator of a reasonably good convergent validity for the model (Kline, 2005).
A summary of R2, together with the standardized path coefficients, of the best-fit
measurement model is shown in Table 5.8. The results showed that most of the R2 of the
observed variables were greater than 0.50, indicating a reasonably good convergent
validity of the fit-measurement model. Moreover, all path coefficients were positive and
statistically significant at p < 0.05, thus their significance to the model was augmented.
A System Dynamics Approach to Construction Safety Culture
109
PolSafety initiativesSafety standards
Safety and productivity alignment
Safety awareness
Safety integration in business goals
PplSafety responsibilities
Workers' involvement
Workers' relationships
Prs
Human resources
Safety resourcesFinancial resources
Stakeholders' cooperation
Pro
Feedback
Training
Safety documentation
Goals
Number of accidentsIndustrial image
Workforce moraleCost of accidents
LdsLeading by example
AccountabilityCommunication
Commitment
Benchmarking system
Note: Lds = Leadership, Pol = Policy and Strategy, Ppl = People, Prs = Partnerships and Resources, Pro = Processes
Figure 5.2 The best-fit measurement model
A System Dynamics Approach to Construction Safety Culture
110
Table 5.8 Square multiple correlations and standardized coefficients of observed
variables
Regression Path R2 Standardized Coefficient
Leadership
Commitment 0.53 0.73
Communication 0.61 0.78
Accountability 0.70 0.84
Leading by example 0.38 0.61
Safety and productivity alignment 0.40 0.63
Policy and Strategy
Safety awareness 0.59 0.77
Safety standards 0.62 0.79
Safety initiatives 0.75 0.87
Workers’ relationships 0.35 0.59
People
Safety responsibilities 0.37 0.61
Stakeholders’ cooperation 0.76 0.87
Human resources 0.72 0.85
Partnerships and Resources
Financial resources 0.81 0.90
Safety resources 0.76 0.87
Workers’ involvement 0.35 0.59
Training 0.48 0.69
Processes
Feedback 0.59 0.77
Safety documentation 0.66 0.81
Benchmarking 0.45 0.67
Safety integration in business goals 0.59 0.77
Goals
Number of accidents 0.55 0.74
Industrial image 0.42 0.65
Workforce morale 0.46 0.68
Cost of accidents 0.36 0.60
Note: All path coefficients were significant at 0.05 probability level
A System Dynamics Approach to Construction Safety Culture
111
5.3.2 Structural Model
Having established confidence in the measurement model, a structural equation model
was developed and tested to examine the direction of the assumed relationships between
the six latent variables (constructs), as reflected by the arrows connecting them (see
Figure 3.6). A fundamental feature of any SEM is the direction of the arrows denoting
the direction of the assumed relationships between the variables as explained below.
In the CSC model, the arrows were assumed to support the argument that Leadership
drives (influences) three enablers (Policy and Strategy, People, and Partnerships and
Resources), and that these enablers collectively influence the ability to achieve pre-
determined Goals through the implementation and improvement of suitable Processes.
As a starting point, bi-directional arrows were used to represent the relationships among
the three enablers (Policy and Strategy, People, and Partnerships and Resources),
without an explicitly defined causal direction. This is because of the variables’ potential
to affect each other. For example, Policy and Strategy might influence People, and/or
vice versa.
To explore this relationship further, and to improve the overall model fit, a number of
model runs (with different arrow directions connecting the enablers) were carried out.
Any links with very low correlations, or items showing signs of multicollinearity, were
deleted. For each run, the GOF indices were computed and compared. According to
Clissold (2004), the model with the best fit would prove the directional influences.
As a result, the ‘leading by example’ item was deleted because of its high
multicollinearity, leading to four observed variables representing the Leadership
construct. The fitted structural model (shown in Figure 5.3), with the best GOF indices
(listed in Table 5.9), was deemed to be the final CSC model (see Figure 5.4) (the full
results of the structural model are presented in Appendix 5).
A System Dynamics Approach to Construction Safety Culture
112
PolSafety initiativesSafety standards
Safety and productivity alignment
Safety awareness
Safety integration in business goals
Ppl
Safety responsibilities
Workers' relationships
Prs
Human resources
Safety resources
Financial resources
Stakeholders' cooperation
Pro
Feedback
Training
GoalsIndustrial imageWorkforce moraleCost of accidents
LdsAccountability
CommunicationCommitment
Safety documentation
Number of accidents
Benchmarking system
Workers' involvement
Note: Lds = Leadership, Ppl = People, Prs = Partnerships and Resources, Pol = Policy and Strategy, Pro = Processes
Figure 5.3 The best-fit structural model
A System Dynamics Approach to Construction Safety Culture
113
Table 5.9 The GOF indices of the best-fit structural model
GOF Index Recommended Level Best-Fit
Measurement Model
Best-fit
Structural Model
�2/DF < 2.00 (Byrne, 2001) 1.65 1.68
CFI > 0.90 (Kline, 2005) 0.91 0.91
NFI > 0.80 (Ullman, 2001) 0.81 0.81
TLI > 0.90 (Kline, 2005) 0.90 0.90
IFI > 0.80 (Garson, 2006) 0.91 0.91
RFI > 0.80 (Garson, 2006) 0.78 0.78
RMSEA � 0.10 (Tabachnick and Fidell, 2007) 0.07 0.07
Note: The best-fit measurement model is shown in Figure 5.2, and the best-fit structural model is shown in Figure 5.3
The final CSC model (see Figure 5.4) confirmed that Processes had a significant direct
relationship with Goals (with a path coefficient = 0.90), and that Processes explains (or
influences) 82% of the variance in Goals.
Enablers Goals
Innovation and Learning
Leadership(Lds)
People (Ppl)
Policy and Strategy
(Pol)
Partnerships and Resources
(Prs)
Processes(Pro) Goals
0.64
0.16
0.59
0.86
0.36
0.46
0.62
0.90
0.41
0.94
0.77
0.99 0.82
(100 points)
(90 points)
(90 points)
(80 points)
(140 points)(500 points)
Note: All path coefficients were significant at 0.05 probability level. The italised numbers show the variance values (R2) of the factors
Figure 5.4 The final CSC model
A System Dynamics Approach to Construction Safety Culture
114
As previously mentioned, the value of SEM lies in its ability to depict both direct and
indirect effects between the variables. In light of this, the final CSC model (see Figure
5.4) appears to indicate that People strongly influences Partnerships and Resources,
whereas Partnerships and Resources moderately affects Policy and Strategy. Both
People and Policy and Strategy were found to have significant direct relationships with
Processes (with path coefficients of 0.46 and 0.62, respectively) at 0.05 probability
level.
No statistically significant relationship, however, was found between Partnerships and
Resources and Processes, indicating the absence of any direct effect. An indirect effect
existed, though, through Policy and Strategy. Further, Partnerships and Resources was
found to have a positive impact on Policy and Strategy (path coefficient = 0.36), which,
in turn, influenced Processes. It was worth noting that the R2 for Processes was 0.99,
demonstrating that 99% of the variance associated with this particular enabler was
accounted for by its two predictors, i.e. the People and Policy and Strategy enablers.
Leadership showed a significant direct relationship with People (path coefficient =
0.64) and Policy and Strategy (path coefficient = 0.59), but surprisingly bore no
statistically significant relationship with Partnerships and Resources (path coefficient =
0.16). The relatively strong influence People had on Partnerships and Resources (path
coefficient = 0.86) suggests that Leadership indirectly influences Partnerships and
Resources through People.
A summary of the direct and indirect path coefficients, together with the values of R2
between the five enablers and Goals, is shown in Table 5.10. Indeed, most of the R2 of
the latent variables were greater than 0.50 indicating a good convergent validity of the
model.
A System Dynamics Approach to Construction Safety Culture
115
Table 5.10 The direct and indirect path coefficients between the five enablers and Goals
Latent Factor Correlation Coefficient R2
Ppl 0.64*Lds 0.41
Prs (0.16*Lds)+(0.86*Ppl)+(0.55*Lds*Ppl) 0.94
Pol (0.59*Lds)+(0.36*Prs)+(0.31*Ppl*Prs) 0.77
Pro (0.46*Ppl)+(0.62*Pol)+(0.29*Lds*Ppl)+(0.37*Lds*Pol)+
(0.22*Prs*Pol)
0.99
Goals (0.90*Pro)+(0.41*Ppl*Pro)+(0.56*Pol*Pro) 0.82
Note: All path coefficients were significant at 0.05 probability level, Ppl = People, Prs = Partnerships and Resources, Pol = Policy and Strategy, Pro = Processes
5.3.3 Conclusion of the SEM
The final CSC model reveals that Leadership strongly influences People and Policy and
Strategy. Leaders should, therefore, be a role model in promoting healthy and safe work
behaviour, ensure that workers accept their safety responsibilities, and set a realistic
safety policy and communicate this policy throughout organizations. It is clear that
Leadership is the main driver to effective safety culture, and the strong commitment of
leaders is crucial in promoting safety culture.
Leadership also has an influence on Partnerships and Resources; however, it appears to
be a relatively weak direct effect. It appears that most of its influence on this particular
enabler is mediated through the People enabler. This indirect effect corroborates well
with the overall perception of Thai construction managers, where human resources and
teamwork are believed to be more crucial to successful safety implementation than the
provision of safety resources (Aksorn and Hadikusumo, 2007). These relationships are
reflected by a strong correlation between People and Partnerships and Resources, and a
weak correlation between Leadership and Partnerships and Resources.
A System Dynamics Approach to Construction Safety Culture
116
Consequently, safety policy and strategy should reflect the need of safety resources, as
requested by the workers, because of the direct and indirect relationships that exist
between Partnerships and Resources and Policy and Strategy, and between People and
Policy and Strategy, respectively.
In conclusion, it can be stated that People and Policy and Strategy play a key role in
successful safety implementation, as verified by the strong links from these two
enablers to Processes. Partnerships and Resources, on the other hand, shows no
significant direct, but indirect, effect on Process through Policy and Strategy.
Therefore, an effective safety policy and strategy will influence an effective safety
implementation, which, in turn, will enhance Goals achievement in organizations.
The final CSC model, as well as the correlation coefficients between its six constructs
(five enablers and Goals), were used in developing the CSC dynamic model (described
in the next chapter).
A System Dynamics Approach to Construction Safety Culture
117
66 SSYYSSTTEEMM DDYYNNAAMMIICCSS MMOODDEELLLLIINNGG OOFF
CCOONNSSTTRRUUCCTTIIOONN SSAAFFEETTYY CCUULLTTUURREE -- MMOODDEELL BBUUIILLDDIINNGG
6.1 GENERAL OVERVIEW
This chapter outlines the development of the CSC dynamic model. System dynamics
(SD) modelling technique was used to capture the interactions and causal relationships
between the five enablers and Goals of the CSC, over time. The developed dynamic
model was verified and validated to increase confidence in the model.
6.2 SYSTEM DYNAMICS MODELLING
System dynamics (SD) was used to examine the various social, economic, and
environmental systems, where a holistic view is important, and feedback loops are
critical to understanding the interrelationships (Rodrigues and Bowers, 1996).
According to Simonovic (2005), this understanding of interrelationships is achieved by
developing a model that can simulate and quantify the behaviour of the system, once
again over time. The simulation of the model is considered essential to understand the
dynamics of the system. An overview of SD, its applications in the construction, and the
available softwares were presented in Section 2.2.7.
To develop a SD simulation model, five steps must be carried out: 1) understanding the
system and its boundaries; 2) identifying the key variables; 3) representing the physical
processes or variables through mathematical relationships; 4) mapping the structure of
the model; and 5) simulating the model (Simonovic, 2005). Vennix (1996) has
identified the systematic procedural steps in SD modelling as follows:
A System Dynamics Approach to Construction Safety Culture
118
� Problem identification and model purpose;
� System conceptualization;
� Model formulation and parameter estimation;
� Analysis of model behaviour including sensitivity analysis;
� Model evaluation, including model validity and verification;
� Policy analysis testing; and
� Model use or implementation.
In SD modelling, feedback loops are considered critical in understanding the
interrelationships between key elements of the model. These feedback loop structures,
once identified, are translated to so-called stock-flow diagrams, to enable the
simulations (Ford, 1999). The basic components of a typical SD model are shown in
Figure 6.1.
Stock: State or condition of the sy stem
Flow
Conv erter: Inf ormation about lev el of the sy stem
Figure 6.1 Basic components of a SD model
The basic CSC dynamic model (Figure 6.2) in this study showed that the CSC index
was represented by the sum of the five enablers and Goals values (with a maximum
score of 1,000 points). For simplicity, and keeping the basic 50/50 for the
Enablers/Goals ratio, as indicated by the EFQM Excellence model in Figure 3.6, any
increase in the value of the CSC index was assumed to have contributed from both
Enablers and Goals, evenly.
Source
Action
A System Dynamics Approach to Construction Safety Culture
119
CSC INDEX
Enablers Goals
Leadership
Policy and Strategy
People
Partnerships and Resources
Processes
Figure 6.2 Basic CSC dynamic model
The following section sheds more light on the relationships between the five enablers
and Goals of the CSC, as modelled by the SD.
6.3 CAUSAL LOOP DIAGRAMS OF CONSTRUCTION SAFETY
CULTURE
6.3.1 Causal Loop Diagram
To conceptualize a real world system under investigation, the SD focuses on the
structure and behaviour (over time) of the system using multiple feedback loops (closed
chains of cause-and-effect links, in which information about the result of actions is fed
back to generate further action). These feedback loops are presented graphically using a
causal loop diagram (see Figure 6.3).
A System Dynamics Approach to Construction Safety Culture
120
-
+ + +
+
+ -
+
-
Arboleda et al., 2003
Arboleda et al., 2003
Dedobbeleer and Beland, 1991
Pipitsupaphol and Watanabe, 2000
Speirs and Johnson, 2002
Siu et al., 2004 Siu et al., 2004
Siu et al., 2004
Gillen et al., 2002
Figure 6.3 An example of a causal loop diagram
The causal loop diagram is a SD tool used to portray a feedback loop in an easy
understanding diagram. A loop is a closed system, comprising a number of elements
and causal relationships. The arrows (as shown in Figure 6.3) indicate the direction of
influence, and plus/minus (+, -) signs indicate the type of the influence (Khanna et al.,
2004). These plus/minus signs have the following meanings (Forrester, 1985):
� A causal link from one element ‘A’ to another element ‘B’ is positive (that is, +), if
either: a) ‘A’ adds to ‘B’; or b) a change in ‘A’ produces a change in ‘B’ in the same
direction. For example, a better ‘perception of safety’ will enhance more
‘participation in safety activities’; this relationship is represented by a plus (+) sign
(see Figure 6.3).
� A causal link from one element ‘A’ to another element ‘B’ is negative (that is, -), if
either: a) ‘A’ subtracts from ‘B’; or b) a change in ‘A’ produces a change in ‘B’ in
the opposite direction. For example, a lesser ‘accident rate’ leads to higher ‘job
satisfaction’; this is represented by a minus (-) sign (see Figure 6.3).
In addition to the signs on each link, the complete feedback loop also is given a sign. If
a particular element starts the loop by changing its value in one direction (e.g. by
Management Commitment
Safety Resources
Safety Related Activities
Perception of Safety
Participation in Safety Activities
Distress
Accident Rate
Job Satisfaction
- - +
Feedback Loop ‘A’ Feedback Loop ‘B’
A System Dynamics Approach to Construction Safety Culture
121
increasing its value), and closes the loop with the value changed in the same direction
(e.g. closes the loop by increasing the value), then the loop is called a positive loop. A
negative loop is the reverse.
According to Ventana Systems, Inc. (2001), the positive and negative loops can also be
determined by counting the number of minus (-) signs appearing on all the links that
make up the loop. Specifically,
� A feedback loop is called positive (indicated by or sign), if it contains an
even number of negative causal links, as shown in feedback loop ‘A’ in Figure 6.3.
� A feedback loop is called negative (indicated by or sign), if it contains an
odd number of negative causal links, as shown in feedback loop ‘B’ in Figure 6.3.
Thus, the sign of a loop is the algebraic product of the signs of its links. To better
understand the feedback loop, an explanation of a positive feedback loop ‘A’, between
‘perception of safety’, ‘participation in safety activities’, ‘distress’, ‘accident rate’, and
‘job satisfaction’, is described in detail below (see Figure 6.3).
An increased ‘perception of safety’ has the potential to increase the level of
‘participation in safety activities’ (representing a positive influence, ‘+’ sign)
(Dedobbeleer and Beland, 1991). A higher level of ‘participation in safety activities’
will tend to decrease the ‘distress’ (representing a negative influence, ‘-’sign), leading to
a reduced ‘accident rate’ (the decrease in distress reduces the accident rate; this
represents a positive influence, ‘+’ sign) (Siu et al., 2004).
A reduced ‘accident rate’ will lead to a higher ‘job satisfaction’ (representing a negative
influence, ‘-’ sign) (Siu et al., 2004), which, in turn, will enhance the people’s
‘perception of safety’ (the higher job satisfaction, the better perception of safety; this
represents a positive influence, ‘+’ sign) (Gillen et al., 2002). Positively enhancing the
‘perception of safety’ closes the loop. Thus, the feedback loop linking ‘perception of
+ +
- -
A System Dynamics Approach to Construction Safety Culture
122
safety’, ‘participation in safety activities’, ‘distress’, ‘accident rate’ and ‘job
satisfaction’ is a positive loop, as shown by the sign (see Figure 6.3).
A negative feedback loop ‘B’ between ‘management commitment’, ‘safety related
activities’, ‘perception of safety’, ‘participation in safety activities’, and ‘safety
resources’ can be described as follows (see Figure 6.3). An increased ‘management
commitment’ towards safety will tend to increase the number and intensity of ‘safety
related activities’ (representing a positive influence, ‘+’ sign), which, in turn, will
significantly enhance the ‘perception of safety’ (representing a positive influence, ‘+’
sign) (Arboleda et al., 2003). A notable increase in the ‘perception of safety’ has the
potential to increase the level of ‘participation in safety activities’ (representing a
positive influence, ‘+’ sign) (Dedobbeleer and Beland, 1991). Thus, more ‘safety
resources’ are required as a result of more people participating in safety activities
(representing a positive influence, ‘+’ sign) (Pipitsupaphol and Watanabe, 2000). This
will, unfortunately, tend to put more pressure on ‘management commitment’ towards
safety (the requirement of safety resources negatively affects management’s
commitment towards safety; this is represented by a ‘-’ sign) (Speirs and Johnson,
2002). This, thus, closes a negative feedback loop (as shown by the sign) between
‘management commitment’ towards safety, ‘safety related activities’, ‘perception of
safety’, ‘participation in safety activities’, and ‘safety resources’ (see Figure 6.3).
6.3.2 A Causal Loop Diagram of the CSC Index
Figure 6.4 shows a causal loop diagram of the proposed CSC index. The loop consists
of seven elements to explain the relationships between the Enablers, Goals, and the
CSC index. These seven elements are:
+
-
A System Dynamics Approach to Construction Safety Culture
123
Gap of CSC Index
Enablers Score
CSC Index Score
Goals Score
Gap of Goals
Desired CSC Index
Desired Goals
+
+
+
-
+
-
-
-
+
+
+
Figure 6.4 A causal loop diagram of the CSC index
1. Enablers score at point (t) in time (maximum 500 points): This score is equal to
the sum of the Leadership (Lds) score (maximum 100 points), the People (Ppl)
score (maximum 90 points), the Partnerships and Resources (Prs) score
(maximum 90 points), the Policy and Strategy (Pol) score (maximum 80 points),
and the Processes (Pro) score (maximum 140 points) (each enabler score is
assigned based on the EFQM Excellence model, see Figure 3.6).
Enablers score = Lds score + Ppl score + Prs score + Pol score + Pro score
2. Goals score at point (t) in time (maximum 500 points)
3. CSC index score at point (t) in time (maximum 1,000 points): This score is equal
to the sum of the Enablers score and the Goals score.
CSC index score = Enablers score + Goals score at point (t) in time
4. Desired goals score: This is the ultimate score that each and every organization
aspires to achieve. The score is set as 500 points.
A System Dynamics Approach to Construction Safety Culture
124
5. Gap of goals at point (t) in time: It is equal to the difference between the Desired
goals score and Goals score at point (t) in time.
Gap of goals = Desired goals score – Goals score
6. Desired CSC index score: This score contains five values: 200, 400, 600, 800,
and 1,000 points, to match the five CSC maturity levels (see Figure 3.7 and
Section 3.5.2). Deciding which value to be used depends on the CSC index score
at that point of time. For instance, if the CSC index score at point (t) in time
equals 100, meaning that the organization is in the first maturity level, and thus
has a maximum score of 200 points. Then, the Desired CSC Index at this point
of time is set as 200 points (representing the threshold for the immediately
following maturity level).
At the first CSC maturity level, the Desired CSC index score = 200
At the second CSC maturity level, the Desired CSC index score = 400
At the third CSC maturity level, the Desired CSC index score = 600
At the fourth CSC maturity level, the Desired CSC index score = 800
At the fifth CSC maturity level, the Desired CSC index score = 1,000
7. Gap of CSC index at point (t) in time: This index score is equal to the difference
between the Desired CSC index score and CSC index at point (t) in time.
Gap of CSC index = Desired CSC index score – CSC index score
The relationships between these seven elements (see Figure 6.4) are as follows. At any
point of time, the CSC index score represents the sum of the Enablers score and the
Goals score. This score is compared with the Desired CSC index score, resulting in a
Gap of CSC index that reflects the difference between these two values. As the CSC
A System Dynamics Approach to Construction Safety Culture
125
index score increases (as a result of an improvement in the Enablers score and the
Goals score), the Gap of CSC index decreases, forming a negative (-) relationship.
Following the continuous improvement cycle, the resulting decrease of the Gap of CSC
index tends to increase the Enablers score (for example, the perception of a smaller Gap
of CSC index results in a better perception of safety, which, in turn, increases the
participation in safety activities; they are then reflected as a higher Enablers score).
Thus, the relationship between the Gap of CSC index and the Enablers score is negative
(-) because a change in the Gap of CSC index results in a change in the opposite
direction of the Enablers score. The increased Enablers score, undoubtedly, enhances
the CSC index score, representing a positive (+) relationship (the changes of the two
elements are in the same direction). This then closes a positive loop ( ) between the
CSC index score, the Gap of CSC index, and the Enablers score (the loop starts and
ends by increasing the CSC index score).
Continuing with the Enablers score, the increased Enablers score improves the Goals
score (represented by a ‘+’ sign). This may be seen, for example, as proper safety
training leads to lower accidents, which, in turn, increases job satisfaction (Teo et al.,
2005). The higher Goals score, when compared with the Desired Goals score (which is
set as 500 points), results in a smaller Gap of Goals (a negative, ‘-’, relationship is
formed). The perceived smaller goals gap will tend to enhance the implementation of
the five enablers (the Enablers score increases). For example, the lower number and
cost of accidents enhance the perception of, and commitment to, safety (Turner, 1991).
Therefore, the loop between the Enablers score, the Goals score, and the Gap of Goals
is considered a positive ( ) loop, i.e. the loop starts and closes by increasing the
Enablers score.
To further explain the interactions among the five CSC enablers, an example of a more
detailed causal loop diagram, showing the relationships among these five enablers and
Goals, is illustrated below.
+
+
A System Dynamics Approach to Construction Safety Culture
126
Safety Implementation(Pro)
Effective Safety Policy(Pol)
Resource Requirements(Prs)
Staff Participation(Ppl)
Management Commitment(Lds)
Cost of Accidents(Goals)
-
+++
+
++
-
+
+
Figure 6.5 A causal loop diagram of the five enablers and Goals
As shown in Figure 6.5, increased management commitment towards safety (Lds) will
tend to increase staff participation in safety activities (Ppl) (a ‘+’ link) (Teo et al.,
2005), which, in turn, will increase resource requirements (Prs) (a ‘+’ link)
(Pipitsupaphol and Watanabe, 2000). This increase in resource requirements (Prs) is
likely to enhance safety policy formulation (Pol) (a ‘+’ link), which sequentially
improves safety implementation (Pro) (a ‘+’ link). This assumption reflects the
recommendations made by Wright et al. (1999), which implied that resource
requirements are a fundamental element in formulating effective policies to improve
safety process implementation.
Further, improved safety implementation (Pro) will tend to reduce the cost of accidents
(Goals) (a ‘-’ link) (Pannirselvam and Ferguson, 2001). Undoubtedly, this reduced cost
of accidents (Goals) leads to more management commitment towards safety (Lds) (a ‘-’
link) (Turner, 1991). Thus, the feedback loop between Lds, Ppl, Prs, Pol, Pro, and
Goals is a positive loop.
A System Dynamics Approach to Construction Safety Culture
127
In this study, the causal loop diagrams of the CSC were converted into a so-called
‘stock-flow’ diagram, using a SD based software package ‘STELLA’ (Ithink, 2003), to
enable the simulations. The details are described in the next section.
6.4 CONSTRUCTION SAFETY CULTURE DYNAMIC MODEL
The formulated CSC dynamic model (as shown in Figure 6.6) captures the interactions
among the five enablers and Goals, where the CSC index represents the sum of the
Enablers score and the Goals score (with an overall score of 1,000 points, see Section
6.3.2). This dynamic model reflects the assumption that the CSC index can be
‘healthier’, provided that the organization focuses on improving the five enablers and,
accordingly, achieves higher safety goals. The dynamic models of the five enablers and
Goals are discussed in detail in the following sections.
6.4.1 Leadership Dynamic Model
The Leadership dynamic model, as shown in Figure 6.7, provides a simple
representation of the stock (leadership) and flow (rlds = leadership rate) diagram (refer
to the Acronyms list – page xxi). In this model the increase in the ‘rlds’ depends on: 1)
the value of the leadership (used_lds); 2) the leadership rate fraction (rldsf); 3) the gap
of goals (ggoals); 4) the gap of leadership (glds); and 5) the percentage of more effort
provided to improve the leadership score (plds) (in the initial base run of the model, the
organization considers all five enablers as having equal significance in improving the
CSC index, so the ‘plds’ is set as zero), as shown in Equations 6.1 and 6.2 (see full
details of the SD equations in Appendix 6).
A S
yste
m D
ynam
ics
App
roac
h to
Con
stru
ctio
n Sa
fety
Cul
ture
12
8
LE
AD
ER
SH
IP
rld
s
use
d l
ds
dld
s
GO
AL
S
rgo
als
CS
C I
ND
EX
gg
oal
sd
go
als
use
d l
ds
use
d p
ol
use
d p
pl
use
d p
rs
use
d p
ro
rld
sf
PE
OP
LE
rpp
l
use
d g
oal
s
Co l
ds
pp
l
DF
pp
l ld
sg
pp
l
use
d p
pl
gg
oal
s
dp
pl
DF
goal
s pro
Co
pro
go
als
PO
LIC
Y &
ST
RA
TE
GY
gp
ol
gp
ol
DF
pol
lds
rpol
DF
po
l p
rsC
o p
rs p
ol
Co
lds
po
l use
d p
ol
dpo
l
PA
RT
NE
RS
HIP
S &
RE
SO
UR
CE
S
rprs
use
d p
rs
dp
rs
gp
rs
DF
prs
lds
Co
ld
s p
rs
EN
AB
LE
RS
DF
prs
ppl
Co p
pl
prs
gp
rs
PR
OC
ES
SE
S
rproD
F p
ro p
pl
Co
ppl
pro
use
d p
rodpro
gp
ro
DF
pro
pol
Co
po
l pro
gp
ro
gld
s
pld
s
pp
ol
pp
pl
pp
rs
pp
ro
Not
e: C
o_ld
s_po
l, C
o_ld
s_pp
l, C
o_ld
s_pr
s, C
o_po
l_pr
o, C
o_pp
l_pr
o, C
o_pp
l_pr
s, C
o_pr
o_go
als,
and
Co_
prs_
pol =
Cor
rela
tions
bet
wee
n Ld
s an
d Po
l, Ld
s an
d Pp
l, Ld
s an
d Pr
s, P
ol a
nd
Pro,
Ppl
and
Pro
, Pp
l an
d Pr
s, P
ro a
nd G
oals
, an
d Pr
s an
d Po
l, re
spec
tivel
y. D
F_go
als_
pro,
DF_
pol_
lds,
DF_
pol_
prs,
DF_
ppl_
lds,
DF_
pro_
pol,
DF_
pro_
ppl,
DF_
prs_
lds,
and
D
F_pr
s_pp
l = D
ecis
ion
frac
tions
bet
wee
n G
oals
and
Pro
, Pol
and
Lds
, Pol
and
Prs
, Ppl
and
Lds
, Pro
and
Pol
, Pro
and
Ppl
, Prs
and
Lds
, and
Prs
and
Ppl
, res
pect
ivel
y. d
goal
s, d
lds,
dpo
l, dp
pl, d
pro,
dpr
s,=
Des
ired
Goa
ls, d
esir
ed L
ds, d
esir
ed P
ol, d
esir
ed P
pl, d
esir
ed P
ro, a
nd d
esir
ed P
rs, r
espe
ctiv
ely.
ggo
als,
gld
s, g
pol,
gppl
, gpr
o, a
nd g
prs
= G
aps
of G
oals
, Lds
, Pol
, Ppl
, Pr
o, a
nd P
rs, r
espe
ctiv
ely.
pld
s, p
pol,
pppl
, ppr
o, a
nd p
prs
= Pe
rcen
tage
eff
ort p
rovi
ded
to L
ds, P
ol, P
pl, P
ro, a
nd P
rs, r
espe
ctiv
ely.
rgoa
ls, r
lds,
rpol
rppl
, rpr
o, a
nd rp
rs =
Goa
ls, L
ds, P
ol,
Ppl,
Pro,
and
Prs
rate
s, re
spec
tivel
y. u
sed_
goal
s, u
sed_
lds,
use
d_po
l, us
ed_p
pl, u
sed_
pro,
and
use
d_pr
s =
Goa
ls, L
ds, P
ol, P
pl, P
ro, a
nd P
rs v
alue
s, re
spec
tivel
y. rl
dsf =
Lds
rate
frac
tion
Figu
re 6
.6 T
he C
SC d
ynam
ic m
odel
A System Dynamics Approach to Construction Safety Culture
129
Leadership
rlds
used lds
dlds
rldsf
ggoals
gldsplds
Leadership Dynamic Model
Note: See Acronyms list for the abbreviations. The symbols are described in Figure 2.6 and Table 2.1.
Figure 6.7 Leadership dynamic model
Eq. 6.1 Leadership(t) = Leadership(t - dt) + (rlds)*dt
Eq. 6.2 rlds = ((used_lds + ggoals)*rldsf) + (glds*plds)
The percentage of more effort provided to improve the Leadership score (plds) is the
effort (rather than what is normally provided) that the organization dedicates to boost
the value of Leadership to achieve its maximum score, i.e. 100 points, in a shorter
period of time. Further details are described in Section 6.5.2.
The value of the leadership rate fraction (rldsf) is constant, and was recommended by
the respondents (participating in the questionnaire survey) to be 0.08. It is the average
score derived from the ratio of budgets organizations reported to be spending in
implementing safety activities to their annual total budgets.
�
Specifically, then, Figure 6.7 may be explained as follows: when ‘ggoals’ is large (in
other words the score of Goals is low compared to the 500 targeted score), leadership
must try hard to reduce this gap. This is achieved by, for example, leaders committing
more to safety, encouraging more two-way communication, assigning safety
accountability to staff, and aligning productivity and safety targets. As a result, the
‘rlds’ increases. Naturally, the increased ‘rlds’ increases the ‘leadership’ stock, which,
in turn, increases the ‘used_lds’ value. The maximum score of ‘used_lds’ is controlled
= 0 (in base run)
A System Dynamics Approach to Construction Safety Culture
130
by the maximum ‘desired value of leadership’ (dlds), which is equal to 100 points (see
Figure 5.4).
Given that Leadership is assumed to drive People, Partnerships and Resources, and
Policy and Strategy, the newly obtained ‘used_lds’ value is transferred to these three
connected dynamic models. This transferred value is, however, influenced by the
strength of the correlation between Leadership and the three enablers (People,
Partnerships and Resources, and Policy and Strategy) (see Figure 5.4). For example,
the ‘used_lds’ value that is transferred to the People dynamic model will have a value
equal to the value of the ‘used_lds’, multiplied by the correlation coefficient between
Leadership and People, which is equal to 0.64 (0.64*‘use_lds’ value). The ‘used_lds’
values transferred to the Partnerships and Resources, and Policy and Strategy dynamic
models, on the other hand, are equal to (0.16*‘use_lds’ value), and (0.59*‘use_lds’
value), respectively.
6.4.2 People Dynamic Model
The relationship between Leadership and People has been confirmed and cited by many
research studies. Little (2002), for example, stated that leaders play an important role in
changing workers’ behaviour through demonstrating strong commitment and
accountability towards safety. The psychological link between management presence on
site and workers safe behaviour gives rise to a good perception about safety, and hence
getting workers to accept more safety responsibilities (Siu et al., 2004).
The People dynamic model, shown in Figure 6.8, illustrates that the ‘used_lds’ value
(obtained from Section 6.4.1) affects the people rate (rppl), as represented in Equations
6.3 and 6.4. This, in turn, influences the ‘used_ppl’ value.
A System Dynamics Approach to Construction Safety Culture
131
People
rppl
Co lds ppl
DF ppl lds gppl
used ppl
dppl
used lds
pppl
People Dynamic Model
Note: See Acronyms list for the abbreviations. The symbols are described in Figure 2.6 and Table 2.1.
Figure 6.8 People dynamic model
Eq. 6.3 People(t) = People(t - dt) + (rppl)*dt
Eq. 6.4 rppl = (used_lds*DF_ppl_lds) + (gppl*pppl)
‘DF_ppl_lds’, shown in Equation 6.4, depends on the ‘gap of people’ (gppl) value, and
the ‘correlation coefficient between the leadership and people’ (Co_lds_ppl) value. As
shown in Figure 5.4, the ‘Co_lds_ppl’ value equals 0.64, leading to Equations 6.5 and
6.6 of ‘DF_ppl_lds’ as below.
Eq. 6.5 DF_ppl_lds = gppl*Co_lds_ppl/100
Eq. 6.6 DF_ppl_lds = gppl*(0.64/100)
The ‘used_ppl’ score is controlled by the ‘dppl’ value (the desired score of the People
enabler, 90 points). This ‘used_ppl’ score is then transferred to the Partnerships and
Resources and Processes dynamic models (see causal links between Ppl and Prs, and
between Ppl and Pro in Figure 5.4).
= 0 (in base run)
A System Dynamics Approach to Construction Safety Culture
132
6.4.3 Partnerships and Resources Dynamic Model
People has a direct effect on Partnerships and Resources, as supported by Pipitsupaphol
and Watanabe (2000), who investigated the root causes of accidents in the Thai
construction industry. They concluded that workers must be provided with adequate
safety resources to facilitate performing the job safely. The ‘used_ppl’ value, therefore,
flows into the ‘partnerships and resources rate’ (rprs), as shown in Figure 6.9 and
Equations 6.7 and 6.8 below.
Partnerships and Resources
rprs
used lds
used prs
dprsgprs
DF prs lds
Co lds prs
used ppl DF prs ppl
Co ppl prs
gprs
pprs
Partnerships and Resources Dynamic Model
Note: See Acronyms list for the abbreviations. The symbols are described in Figure 2.6 and Table 2.1.
Figure 6.9 Partnerships and Resources dynamic model
Eq. 6.7 Partnerships_&_Resources(t) = Partnerships_&_Resources(t - dt) + (rprs)*dt
Eq. 6.8 rprs = (used_lds*DF_prs_lds) + (used_ppl*
DF_prs_ppl) + (gprs*pprs)
‘DF_prs_lds’, shown in Equation 6.9, depends on the ‘gap of partnerships and
resources’ (gprs) value, and the ‘correlation coefficient between the leadership and
partnerships and resources’ (Co_lds_prs) value. ‘DF_prs_ppl’, on the other hand,
depends on the ‘gprs’ value, and the ‘correlation coefficient between the people and
partnerships and resources’ (Co_ppl_prs) value (see Equation 6.10).
= 0 (in base run)
A System Dynamics Approach to Construction Safety Culture
133
Eq. 6.9 DF_prs_lds = gprs*Co_lds_prs/100
Eq. 6.10 DF_prs_ppl = gprs*Co_ppl_prs/100
Leadership also has an effect on Partnerships and Resources, although it is a weak one.
Consequently, the ‘used _lds’ value also flows into the ‘rprs’. The increase in the ‘rprs’
enhances the ‘used_prs’ value, and this, in turn, increases the ‘policy and strategy rate’
(rpol) of the Policy and Strategy dynamic model, as described in the following section.
6.4.4 Policy and Strategy Dynamic Model
Leadership and Partnerships and Resources influence the establishment of safety policy
and strategies in the organization (see Figure 5.4), leading to the flows of the ‘used_lds’
and ‘used_prs’ values into the ‘policy and strategy rate’ (rpol), as shown in Figure 6.10
and Equations 6.11 to 6.14.
used ldsPolicy and Strategy
gpolDF pol lds
used prsgpol
rpol
DF pol prs Co prs pol
Co lds pol
used pol
dpol
ppol
Policy and Strategy Dynamic Model
Note: See Acronyms list for the abbreviations. The symbols are described in Figure 2.6 and Table 2.1.
Figure 6.10 Policy and Strategy dynamic model
A System Dynamics Approach to Construction Safety Culture
134
Eq. 6.11 Policy_&_Strategy(t) = Policy_&_Strategy(t - dt) + (rpol)*dt
Eq.6.12 rpol = (used_lds*DF_pol_lds) + (used_prs*
DF_pol_prs) + (gpol*ppol)
Eq. 6.13 DF_pol_lds = gpol*Co_lds_pol/100
Eq. 6.14 DF_pol_prs = gpol*Co_prs_pol/100
The increased ‘rpol’ increases the ‘used_pol’ value. This ‘used_pol’ value is then
transferred to the Processes dynamic model, as described in the next section.
6.4.5 Processes Dynamic Model
Both People and Policy and Strategy play a key role in the successful safety
implementation; this is consistent, to some extent, with the process management studies
of Eskildsen and Dahlgaard (2000), and Pannirselvam and Ferguson (2001), where
process management was found to be directly related to strategic planning and human
resource management.
The Processes dynamic model, as depicted in Figure 6.11, demonstrates that the
increased ‘used_ppl’ and ‘used_pol’ values tend to increase the ‘processes rate’ (rpro)
(see Equations 6.15 to 6.18).
= 0 (in base run)
A System Dynamics Approach to Construction Safety Culture
135
Processes
rpro
used ppl
DF pro pplCo ppl pro
used pro
dpro
gpro
used polDF pro pol
Co pol pro
gpro
ppro
Processes Dynamic Model
Note: See Acronyms list for the abbreviations. The symbols are described in Figure 2.6 and Table 2.1.
Figure 6.11 Processes dynamic model
Eq. 6.15 Processes(t) = Processes(t - dt) + (rpro)*dt
Eq. 6.16 rpro = (used_ppl*DF_pro_ppl) + (used_pol*DF_pro_pol)
+ (gpro*ppro)
Eq. 6.17 DF_pro_ppl = gpro*Co_ppl_pro/100
Eq. 6.18 DF_pro_pol = gpro*Co_pol_pro/100
The increasing of ‘rpro’ will improve the ‘used_pro’ value, which, ultimately, will
enhance the ‘goals rate’ (rgoals), as described in the next section.
6.4.6 Goals Dynamic Model
In the Goals dynamic model, the Processes enabler appears to be very strongly
correlated to Goals (see Figure 5.4). The increase or decrease of the ‘used_pro’ value
has an effect on the ‘goals rate’ (rgoals), as shown in Figure 6.12 and Equations 6.19
and 6.20 below.
= 0 (in base run)
A System Dynamics Approach to Construction Safety Culture
136
Goals
rgoals
ggoals
dgoals
used proused goals
DF goals proCo pro goals
CSC Index
Goals Dynamic Model
Note: See Acronyms list for the abbreviations. The symbols are described in Figure 2.6 and Table 2.1.
Figure 6.12 Goals dynamic model
Eq. 6.19 Goals(t) = Goals(t - dt) + (rgoals)*dt
Eq. 6.20 rgoals = used_pro*DF_goals_pro
The increase of the ‘used_goals’ value depends on the ‘rgoals’ value, which appears to
increase when the ‘used_pro’ value increases. The increased ‘used_goals’ value reduces
the ‘gap of goals’ (ggoals), which then has an effect on the ‘rlds’ of the Leadership
dynamic model (see Figure 6.7).
The simulations of the CSC dynamic model iterate as cycles, from the Leadership to the
Goals dynamic models. In each cycle, the Enablers score and the CSC index are
calculated, as illustrated in Figure 6.13 and Equations 6.21 and 6.22 below. The cycles
continue until the CSC index reaches a score of 800 or more, indicating that the
organization has achieved the ‘continually improving’ (fifth) maturity level’.
A System Dynamics Approach to Construction Safety Culture
137
CSC Index
used lds
used polused ppl used prs
used pro
used goalsEnablers
CSC Index Dynamic Model
Note: See Acronyms list for the abbreviations. The symbols are described in Figure 2.6 and Table 2.1.
Figure 6.13 The CSC index dynamic model
Eq. 6.21 Enablers = used_lds + used_ppl + used_prs + used_pol + used_pro
Eq. 6.22 CSC index = Enablers + used_goals
The next section (Section 6.5) describes the simulation results of the CSC dynamic
model.
6.5 DYNAMIC SIMULATION RESULTS
6.5.1 Base Run Results
The CSC dynamic model was simulated using ‘STELLA’ software (Ithink, 2003). In
the ‘base run’ simulation, the initial values of the five enablers were set as zero to
manipulate the situation of organizations with no prior safety implementation. The
initial value of Goals, however, was not equal to zero. The linear regression was
performed with the SPSS program, using data from the questionnaire survey between
the Enablers score (as a sum data score of the five enablers) and the Goals score (see
A System Dynamics Approach to Construction Safety Culture
138
full details of linear regression results in Appendix 7). The results reveal that when the
Enablers value (which has a maximum score of 500 points) equals the hypothetical
value of zero, the Goals value (which also has a maximum score of 500 points) is equal
to 68 points (see Equation 6.23 below). This may be explained that some of the four
Goals’ attributes (including, ‘number of accidents’, ‘industrial image’, ‘workforce
morale’, and ‘cost of accidents’, see Figure 5.3) do not totally and exclusively depend
on the implementation of the five enablers. For example, organizations may experience
low ‘number of accidents’ due to low workload or sheer luck.
Eq. 6.23 Goals = 68 + (0.79*Enablers)
The initial values of Enablers (which was set at zero) and Goals (68 points) were
substituted in the SD equations (see Appendix 5). The dynamic model was simulated,
and the results are displayed graphically in Figures 6.14 to 6.16, and numerically in
Tables 6.1 and 6.2. The time units used in the simulation can be varied as SD model
allows the users to define their own units. For simplicity, however, the time unit used in
this study will be referred to as ‘years’, henceforth. Nevertheless, it is worth noting that
this ‘years’ unit does not represent the calendar year.
Page 11.00 5.00 9.00 13.00 17.00
Years
1:
1:
1:
0
250
500
1: Enablers Score
1
1
1
1
Figure 6.14 Graphical results of the Enablers score over time
A System Dynamics Approach to Construction Safety Culture
139
Page 2
1.00 5.00 9.00 13.00 17.00
Years
1:
1:
1:
68
284
500
1: used goals score
1 1
1
1
Figure 6.15 Graphical results of the Goals score over time
Page 3
1.00 5.00 9.00 13.00 17.00
Years
1:
1:
1:
68
534
1000
1: CSC Index Score
1
1
1
1
Figure 6.16 Graphical results of the CSC index over time
A System Dynamics Approach to Construction Safety Culture
140
Table 6.1 Simulation results of the five enablers and Goals
Year Score
Lds Pol Ppl Prs Pro Goals
Initial 0.00 0.00 0.00 0.00 0.00 68.00
1 2.64 0.47 0.56 0.21 0.13 68.00
2 5.50 2.36 2.66 1.68 1.82 68.13
3 8.55 6.14 6.30 5.52 7.31 69.12
4 11.73 12.28 11.40 12.50 18.66 72.08
5 14.84 20.91 17.70 22.68 36.73 77.84
6 23.20 31.92 25.42 35.34 60.02 114.61
7 30.80 44.85 35.61 49.67 85.27 156.79
8 42.10 57.14 46.98 63.51 107.57 244.71
9 52.23 66.79 58.16 74.71 123.35 329.25
10 60.37 73.17 68.07 82.23 132.44 381.64
11 70.79 76.76 75.77 86.46 136.86 461.90
12 78.61 78.59 81.40 88.53 138.78 491.36
13 85.54 79.42 85.07 89.43 139.54 498.07
14 92.69 79.78 87.32 89.79 139.83 499.57
15 100.00 79.92 88.62 89.92 139.94 499.91
16 100.00 79.97 89.31 89.97 139.98 499.98
17 100.00 79.99 89.66 89.99 139.99 500.00
18 100.00 80.00 89.83 90.00 140.00 500.00
19 100.00 80.00 89.91 90.00 140.00 500.00
20 100.00 80.00 89.96 90.00 140.00 500.00
21 100.00 80.00 89.98 90.00 140.00 500.00
22 100.00 80.00 89.99 90.00 140.00 500.00
23 100.00 80.00 89.99 90.00 140.00 500.00
24 100.00 80.00 90.00 90.00 140.00 500.00
Note: Maximum scores of Lds, Pol, Ppl, Prs, and Pro are 100, 80, 90, 90, and 140 points, respectively.
A System Dynamics Approach to Construction Safety Culture
141
Table 6.2 Simulation results of the enablers, Goals, and CSC index
Year Enablers Used_Goals CSC Index Level*
Score %Increasing Score %Increasing Score %Increasing
Initial 0.00 - 68.00 - 68.00 - 1st
1 4.00 0.80 68.00 0.00 71.98 0.40 1st
2 14.01 2.00 68.13 0.03 82.14 1.02 1st
3 33.82 3.96 69.12 0.20 102.94 2.08 1st
4 66.57 6.55 72.08 0.59 138.65 3.57 1st
5 112.86 9.26 77.84 1.15 190.71 5.21 1st
6 175.90 12.61 114.61 7.35 290.52 9.98 2nd
7 246.21 14.06 156.79 8.44 403.00 11.25 3rd
8 317.30 14.22 244.71 17.58 562.01 15.90 3rd
9 375.24 11.59 329.25 16.91 704.49 14.25 4th
10 416.27 8.21 381.64 10.48 797.91 9.34 4th
11 446.64 6.07 461.90 16.05 908.55 11.06 5th
12 465.90 3.85 491.36 5.89 957.26 4.87 5th
13 479.00 2.62 498.07 1.34 977.07 1.98 5th
14 489.40 2.08 499.57 0.30 988.97 1.19 5th
15 498.40 1.80 499.91 0.07 998.31 0.93 5th
16 499.23 0.17 499.98 0.01 999.21 0.09 5th
17 499.63 0.08 500.00 0.00 999.63 0.04 5th
18 499.82 0.04 500.00 0.00 999.82 0.02 5th
19 499.91 0.02 500.00 0.00 999.91 0.01 5th
20 499.96 0.01 500.00 0.00 999.96 0.01 5th
21 499.98 0.00 500.00 0.00 999.98 0.00 5th
22 499.99 0.00 500.00 0.00 999.99 0.00 5th
23 499.99 0.00 500.00 0.00 999.99 0.00 5th
24 500.00 0.00 500.00 0.00 1,000.00 0.00 5th
Note: Bold numbers refer to the time unit, where the organization reaches the fifth level of CSC maturity. (*) Level = CSC maturity level
As shown in Figures 6.14 through to 6.16, at the starting point, the Enablers value was
zero and the Goals value was 68, leading to the CSC index of 68 points. At this stage,
the gap of the Goals (ggoals) value was relatively large (500 – 68 = 432 points). This,
then, boosted the value of Leadership (see Equation 6.2), which, in turn, increased the
A System Dynamics Approach to Construction Safety Culture
142
values of the remaining four enablers, i.e. People, Policy and Strategy, Partnerships
and Resources, and Processes.
As the five enablers’ values increased (identifying an improvement in safety culture’s
implementations), the Goals value, and the CSC index increased. The simulation
continued until the CSC index reached the maximum score of 1,000 points (the
Enablers and Goals values reach their maximum 500 points). Table 6.2 showed that it
took 11 years for the organization, with a non-existent safety policy and safety
implementation process, to progress from the first to the fifth levels of CSC maturity
(the CSC index reached 800 points or more at the end of year 11).
The graphs shown in Figures 6.14 through to 6.16 showed similar S-shaped patterns,
with a slow increase at the beginning of the simulation. It took six years for the
organization to progress from the first to the second levels of culture maturity. This
result demonstrated that for an organization with a non-existent safety culture policy
and implementation process, it was hard to improve the CSC in the early stage of the
safety implementation. This is shown by a slow increase in the rate for Enablers, and
the even slower increased rate for Goals (see Table 6.2). After the organization reached
the second maturity level, however, the Enablers and Goals values increased rapidly, as
depicted by the sharp rises in the curves shown in Figures 6.14 and 6.15, respectively.
This, in turn, enhanced the CSC index, which could be seen as a steep incline of the
graph shown in Figure 6.16. The organization progressed from the second to the fifth
maturity levels over five years (at the end of year 11), showing a significant safety
improvement in the organization.
After year 11, it was difficult for the organization to increase the Enablers value, as
most of the safety implementations were accomplished. Moreover, the extra effort
needed to further improve safety in the organization might be switched to other
important areas. This, in turn, slows the increase rates of the Goals value and the CSC
index (see Figures 6.15 and 6.16, respectively). As shown in Table 6.2, the organization
achieved its CSC index of 1,000 points (representing the perfect safety implementation)
at the end of year 24. It appears to be very challenging to reach a perfect safety
A System Dynamics Approach to Construction Safety Culture
143
implementation; however, an organization can plan its safety implementation to
progress through to the fifth CSC maturity level by using the time frame shown in Table
6.2.
6.5.2 Base Run Results Examination
By observing the increasing rate of the five enablers’ values at the early stage of the
simulation (see Table 6.1 and Figure 6.17), it is clear that Leadership was the weakest
enabler in boosting the CSC index, as it produced the least scores compared with the
other four enablers. To explain, at the end of year six (when the organization reached
the second maturity level), the Leadership’s score was 23.20 out of 100 points,
representing 23.2% of the scores produced. On the other hand, People, Policy and
Strategy, Partnerships and Resources, and Processes produced 28.24, 39.90, 39.27, and
42.87% of their maximum scores, respectively. Therefore, to expedite the Enablers
value, and achieve higher CSC index in the early stages, an organization should focus
more on improving its Leadership.
Page 1
1.00 5.00 9.00 13.00 17.00
Years
1:
1:
1:
2:
2:
2:
3:
3:
3:
4:
4:
4:
5:
5:
5:
0
50
100
0
45
90
0
40
80
0
70
140
1: used lds score 2: used ppl score 3: used prs score 4: used pol score 5: used pro score
1
1
1
1
2
2
2
2
3
3
3
3
4
4
44
5
5
5 5
Figure 6.17 Graphical results of the five enablers over time
A System Dynamics Approach to Construction Safety Culture
144
To confirm whether the organization should concentrate on improving the
implementation of Leadership, a number of model runs were needed. First, the
organization was said to allocate 10% of more effort to focus on Leadership
improvement, i.e. apart from normally implementing this enabler, the organization puts
more time and effort (by 10%) into further enhancing this particular enabler’s
implementation. This means that the organization maintained its improvement of the
five enablers, but more attention was given to Leadership.
Consequently, the ‘plds’ value was set to 0.1, while the ‘pppl’, ‘pprs’, ‘ppol’, and ‘ppro’
were still set as zero. The CSC dynamic model was then simulated, and the results were
recorded. Next, the ‘plds’ was set back to zero, then the ‘pppl’ was set as 0.1 (meaning
that the organization now changed its focus, from improving Leadership’s
implementation, to the People enabler). The model was re-simulated, and the results
were recorded, then the ‘pppl’ was set back to zero.
The simulations were performed for all five enablers; the results (shown in Table 6.3)
demonstrate that, by focusing more on Leadership, the organization reached the second
maturity level in a much shorter time (three years), and achieved the fifth level of
maturity four years earlier (it took seven years, instead of 11 years, to achieve the fifth
maturity level). Therefore, for the organization starting at level one of CSC maturity,
attention should be paid, in the main, to improving the key attributes of Leadership to
successfully progress through to higher maturity levels. The leaders should thus take
safety seriously through being a role model (Dunlap, 2004; Teo et al., 2005), assigning
and communicating safety responsibilities clearly to all staff (Lardner et al., 2001), and
ensuring that the workload was reasonably balanced among workers to avoid unsafe
behaviours (Glendon and Litherland, 2001).
A System Dynamics Approach to Construction Safety Culture
145
Table 6.3 Experimentation with extra efforts given to improve the five enablers
Year Lds Pol Ppl Prs Pro
Index* Level* Index Level Index Level Index Level Index Level Initial 68.0 1st 68.0 1st 68.0 1st 68.0 1st 68.0 1st 1 87.0 1st 82.3 1st 85.5 1st 81.7 1st 86.9 1st
2 131.7 1st 109.0 1st 121.5 1st 104.5 1st 113.3 1st
3 210.9 2nd 150.0 1st 177.8 1st 141.1 1st 147.9 1st
4 357.0 2nd 203.1 2nd 272.4 2nd 193.5 1st 189.9 1st
5 540.4 3rd 315.4 2nd 394.6 2nd 290.7 2nd 282.8 2nd
6 735.8 4th 414.3 3rd 549.4 3rd 401.7 3rd 377.4 2nd
7 897.1 5th 561.6 3rd 700.7 4th 558.6 3rd 498.5 3rd
8 970.5 5th 693.8 4th 794.9 4th 699.8 4th 602.1 4th
9 991.1 5th 781.9 4th 903.3 5th 792.6 4th 740.2 4th
10 997.0 5th 876.1 5th 949.4 5th 902.9 5th 797.0 4th
11 998.9 5th 937.2 5th 967.1 5th 951.8 5th 896.9 5th
Note: Bold number refers to the CSC index reaching the fifth level of CSC maturity. (*) Level = CSC maturity level, (*) Index = CSC index
6.5.3 Model Verification and Validation
The CSC dynamic model was verified using the ‘logical’ test to assure its parameters,
the unit consistency, and the correct sequence of calculation (see Section 2.2.7.4). The
six constructs (the five enablers and the single set of Goals) and their attributes were
tested and confirmed by two statistical analyses: the EFA and the SEM. The ‘years’
time unit was consistently used throughout the simulation, and the sequence of the
calculation was correct, following the directional influences shown in Figure 5.4.
To validate the CSC dynamic model, a behavioural sensitivity analysis was conducted
(see Section 2.2.7.4) to test the robustness of the model, by ensuring that the
uncertainties and the estimating errors did not significantly affect the overall behaviour
of the model. It tested the limits of the model and its ability to adjust itself in response
to the changes. According to Tang and Ogunlana (2003a), a model is considered robust
A System Dynamics Approach to Construction Safety Culture
146
if its behaviour does not change drastically, when a parameter or behavioural
relationship is altered.
The sensitivity analysis of the CSC dynamic model was carried out by changing the
initial value of Leadership (by 25, 50, and 75%), meaning that the initial value was
changed from zero to 25, 50, and 75 points (out of a maximum of 100 points),
respectively. The simulation results, displayed graphically in Figures 6.18 and 6.19,
demonstrate that the change in the initial values of the Leadership only numerically
affects the model behaviour, not the pattern of the model, thus validating the model.
Page 2
1.00 5.00 9.00 13.00 17.00
Years
1:
1:
1:
0
50
100
used lds score: 1 - 2 - 3 - 4 -
1
1
1
1
2
2
2 2
3
3 3 34 4 4 4
Note: The numbers 1, 2, 3, and 4 displayed in the figure represent the initial values of Lds of zero, 25, 50, and 75 points, respectively.
Figure 6.18 Sensitivity results of the ‘used_lds’ value when its initial value is changed
A System Dynamics Approach to Construction Safety Culture
147
Page 8
1.00 5.00 9.00 13.00 17.00
Years
1:
1:
1:
0
500
1000
CSC Index Score: 1 - 2 - 3 - 4 -
1
1
1
1
2
2
2 2
3
33 3
4
4 4 4
Note: The numbers 1, 2, 3, and 4 displayed in the figure represent the initial values of Lds of zero, 25, 50, and 75 points, respectively.
Figure 6.19 Sensitivity results of the CSC index when the initial values of Lds are
changed
Sensitivity analyses (changing the initial values of People, Partnerships and Resources,
Policy and Strategy, and Processes) were also conducted; the results show that the
patterns of the model behaviour were not sensitive to the changes in all parameters (see
Appendix 8 for graphical results of the sensitivity analyses).
Changing the percentage of more effort provided to improve Leadership (plds) from
zero to 10, 20, and 30%, respectively, through another sensitivity test achieved the
simulation results shown in Figures 6.20 and 6.21, thus proving the non-sensitivity of
the model behaviour.
A System Dynamics Approach to Construction Safety Culture
148
Page 2
1.00 5.00 9.00 13.00 17.00
Years
1:
1:
1:
0
50
100
used lds score: 1 - 2 - 3 - 4 -
1
1
1
1
2
2
2 2
3
3 3 3
4
4 4 4
Note: The numbers 1, 2, 3, and 4 displayed in the figure represent ‘plds’ of zero, 0.1, 0.2, and 0.3, respectively.
Figure 6.20 Sensitivity results of the ‘used_lds’ value when the ‘plds’ value is changed
Page 8
1.00 5.00 9.00 13.00 17.00
Years
1:
1:
1:
0
500
1000
CSC Index Score: 1 - 2 - 3 - 4 -
1
1
1
1
2
2
2 2
3
3
3 3
4
44 4
Note: The numbers 1, 2, 3, and 4 displayed in the figure represent ‘plds’ of zero, 0.1, 0.2, and 0.3, respectively.
Figure 6.21 Sensitivity results of the CSC index when the ‘plds’ value is changed
A System Dynamics Approach to Construction Safety Culture
149
The results of the sensitivity analyses achieved by changing the values of the ‘pppl’,
‘pprs’, ‘ppol’, and ‘ppro’ from zero to 10, 20, and 30%, respectively, are illustrated in
Appendix 9. The results demonstrate the non-sensitivity of the pattern of the model.
The following chapter describes a number of model applications of the CSC dynamic
model. Several policy experiments were performed to underline the areas requiring
safety improvement and an enhancement of the CSC index. The cyclical style of safety
management that reflects management withdrawing attention from safety is also
modelled in the next chapter.
A System Dynamics Approach to Construction Safety Culture
150
[THIS PAGE IS INTENTIONALLY LEFT BLANK]
A System Dynamics Approach to Construction Safety Culture
151
77 SSYYSSTTEEMM DDYYNNAAMMIICCSS MMOODDEELLLLIINNGG OOFF
CCOONNSSTTRRUUCCTTIIOONN SSAAFFEETTYY CCUULLTTUURREE -- MMOODDEELL AAPPPPLLIICCAATTIIOONNSS
7.1 GENERAL OVERVIEW
This chapter presents the verified and validated CSC dynamic model experiment, along
with a set of safety policies examining their impact under different scenarios to
highlight areas for safety improvement, and to enhance the CSC index. The cyclical
style of safety management, modelled to reflect real-life situations, where management
tends to withdraw its attention away from safety following the realisation of excellent
performance record, is also discussed.
Section 7.2 (below) describes a number of policy analyses, as well as their simulation
results.
7.2 POLICY ANALYSIS
The five CSC enablers include those necessary elements that enable the organization to
improve its safety performance. Each enabler comprises a number of attributes, such as
management commitment, workers’ involvement, resource availability, and so on. Any
positive interventions in these attributes are expected to enable the organization to meet
its safety goals. These interventions can be in the form of improving safety training,
providing more safety resources, integrating safety in business goals, etc.
This study grouped all such interventions under enabler-specific improvement efforts,
and referring to these as a set of policy experiments. The formulation of an effective set
of policies through the dynamic model simulation for an improved and sustained
A System Dynamics Approach to Construction Safety Culture
152
organizational performance was the main objective of SD modelling (Tang and
Ogunlana, 2003b). The simulation was, therefore, used as a means for experimentation
with the model, understanding its behaviour, and identifying a framework for policy
intervention (Saeed and Brooke, 1996).
A set of policy experiments, along with comparative simulations, to demonstrate
improved performance, are described below. Two organizations (‘A’ and ‘B’) were
randomly chosen, among the 101 construction-contracting organizations responding to
the questionnaire, to represent two different maturity levels: organization ‘A’ is
currently in the second (managing) level of maturity, and organization ‘B’ is in the third
(involving) maturity level. The SD analysis was undertaken for each of the two
organizations in which the simulation results were called ‘the base run’. A number of
policies were then made for each organization in an attempt to enhance their safety
efforts.
Section 7.2.1 describes the base run for each of these organizations (‘A’ and ‘B’),
including the time period each organization (with different initial Enablers and Goals
values) needs to use to progress through to the fifth CSC maturity level.
7.2.1 Base Run
7.2.1.1 Base Run for Organization ‘A’
The initial values of the five enablers and Goals of organization ‘A’ (referring to the
questionnaire database) were as follows:
� ‘used_lds’ = 20.0 (out of 100 points)
� ‘used_ppl’ = 43.2 (out of 90 points)
A System Dynamics Approach to Construction Safety Culture
153
� ‘used_prs’ = 18.0 (out of 90 points)
� ‘used_pol’ = 19.2 (out of 80 points)
� ‘used_pro’ = 37.3 (out of 140 points)
� ‘used_goals’ = 129.0 (out of 500 points)
These scores were summed and gave rise to the base value for the CSC index of 266.7
points (the CSC index is the sum of the Enabler and Goals scores, see Section 6.3.2).
Organization ‘A’ was, therefore, currently at the second maturity level (see the score-
range in each maturity level in Section 3.5.2). The initial values of the five enablers,
Goals, and CSC index were then substituted in the SD equations (see SD equations of
organization ‘A’ in Appendix 10). In the ‘base run’, the organization considered all five
enablers as having equal significance in improving the CSC index, i.e. the ‘plds’, ‘pppl’,
‘ppol’, ‘pprs’ and ‘ppro’ were all set as zero. The dynamic model was run, leading to
the simulation results displayed numerically in Tables 7.1 and 7.2, and graphically in
Figures 7.1 to 7.4, respectively.
Table 7.1 Simulation results of the five enablers of organization ‘A’
Year Lds Pol Ppl Prs Pro
Score Gap* Score Gap Score Gap Score Gap Score Gap
Initial 20.00 80.00 19.20 60.80 43.20 46.80 18.00 72.00 37.30 102.7
1 26.73 73.27 32.33 47.67 49.62 40.38 44.20 45.80 69.32 70.68
2 36.89 63.11 47.48 32.52 56.87 33.13 63.08 26.92 97.43 42.57
3 46.23 53.77 60.53 19.47 64.56 25.44 75.59 14.41 117.83 22.17
4 55.50 44.5 69.59 10.41 71.80 18.2 82.99 7.01 129.83 10.17
5 64.95 35.05 74.87 5.13 77.70 12.3 86.87 3.13 135.76 4.24
6 73.20 26.80 77.68 2.32 82.26 7.74 88.71 1.29 138.34 1.66
7 79.89 20.11 79.01 0.99 85.38 4.62 89.50 0.50 139.38 0.62
8 86.62 13.38 79.60 0.40 87.38 2.62 89.81 0.19 139.77 0.23
9 93.79 6.21 79.85 0.15 88.58 1.42 89.93 0.07 139.92 0.08
Note: Bold numbers refer to the time unit where the organization reaches its fifth level of CSC maturity. (*) Gap = the difference between the maximum and the achieved scores
A System Dynamics Approach to Construction Safety Culture
154
Table 7.2 Simulation results of the Enablers, Goals, and CSC index of organization ‘A’
Year Score CSC Maturity Level
Enablers Goals CSC Index
Initial 137.70 129.00 266.70 2nd
1 222.21 155.75 377.96 2nd
2 301.76 226.78 528.54 3rd
3 364.73 300.88 665.61 4th
4 409.70 372.86 782.56 4th
5 440.16 444.63 884.79 5th
6 460.18 487.31 947.49 5th
7 473.15 497.16 970.31 5th
8 483.18 499.37 982.55 5th
9 492.07 499.86 991.93 5th
Note: Bold numbers refer to the time unit where the organization reaches its fifth level of CSC maturity.
Page 1
1.00 3.25 5.50 7.75 10.00
Years
1:
1:
1:
2:
2:
2:
3:
3:
3:
4:
4:
4:
5:
5:
5:
20
60
100
19
50
80
43
67
90
18
54
90
37
89
140
1: used lds score 2: used pol score 3: used ppl score 4: used prs score 5: used pro score
1
1
1
1
2
2
2
2
3
3
3
3
4
4
4 4
5
5
5 5
Figure 7.1 Graphical results of the five enablers of organization ‘A’ over time
A System Dynamics Approach to Construction Safety Culture
155
Page 2
1.00 3.25 5.50 7.75 10.00
Years
1:
1:
1:
138
319
500
1: Enablers Score
1
1
1
1
Figure 7.2 Graphical results of the Enablers score of organization ‘A’ over time
Page 3
1.00 3.25 5.50 7.75 10.00
Years
1:
1:
1:
129
315
500
1: used goals score
1
1
1
1
Figure 7.3 Graphical results of the Goals score of organization ‘A’ over time
A System Dynamics Approach to Construction Safety Culture
156
Page 4
1.00 3.25 5.50 7.75 10.00
Years
1:
1:
1:
267
633
1000
1: CSC Index Score
1
1
1
1
Figure 7.4 Graphical results of the CSC index of organization ‘A’ over time
The results show that it took five years for organization ‘A’ to progress from the second
to the fifth maturity levels. To demonstrate, it took two years for the organization to
reach the third maturity level. The organization then advanced through to the fourth
level in one year, and reached the fifth level of maturity at the end of year five.
When organization ‘A’ reached the fifth CSC maturity level, the scores of Partnerships
and Resources, Processes, and Policy and Strategy were close to their maximum scores
(the gaps between their maximum and achieved values were small, see Table 7.1), while
the gaps of the Leadership and People values were relatively large. Thus, to plan for
safety improvement and achieve the fifth maturity level in a shorter time frame, the
organization should pay more attention to improving the Leadership and People
enablers. This is further explained in Section 7.2.2.
The next section describes the base run simulation of organization ‘B’.
A System Dynamics Approach to Construction Safety Culture
157
7.2.1.2 Base Run for Organization ‘B’
The initial values of organization ‘B’ were: ‘used_lds’ = 85.0 points, ‘used_ppl’ = 43.2
points, ‘used_prs’ = 40.5 points, ‘used_pol’ = 35.2 points, ‘used_pro’ = 56.0 points and
‘used_goals’ = 200.0 points, leading to an initial CSC index of 459.9 points.
Organization ‘B’ was, therefore, at the third level of CSC maturity (see Section 3.5.2 for
the score-ranges of the five maturity levels). The initial value of Leadership was
relatively high (85 out of 100 points), demonstrating a strong management commitment
to safety.
The initial values were substituted in the SD equations, and the dynamic model was
simulated (see SD equations of organization ‘B’ in Appendix 11). The simulation
results are presented numerically in Tables 7.3 and 7.4, and graphically in Figures 7.5 to
7.8, respectively.
Table 7.3 Simulation results of the five enablers of organization ‘B’
Year Lds Pol Ppl Prs Pro
Score Gap* Score Gap Score Gap Score Gap Score Gap
Initial 85.00 15.00 35.20 44.80 43.20 46.80 40.50 49.50 56.00 84.00
1 98.72 1.28 59.56 20.44 64.95 25.05 63.96 26.04 91.96 48.04
2 100.00 0.00 72.05 7.95 77.50 12.50 78.85 11.15 118.47 21.53
3 100.00 0.00 77.08 2.92 83.77 6.23 85.71 4.29 131.50 8.50
4 100.00 0.00 78.95 1.05 86.90 3.10 88.44 1.56 136.83 3.17
5 100.00 0.00 79.63 0.37 88.46 1.54 89.45 0.55 138.85 1.15
6 100.00 0.00 79.87 0.13 89.23 0.77 89.81 0.19 139.59 0.41
7 100.00 0.00 79.95 0.05 89.62 0.38 89.93 0.07 139.85 0.15
8 100.00 0.00 79.98 0.02 89.81 0.19 89.98 0.02 139.95 0.05
9 100.00 0.00 79.99 0.01 89.91 0.09 89.99 0.01 139.98 0.02
Note: Bold numbers refer to the time unit where the organization reaches its fifth level of CSC maturity. (*) Gap = the difference between the maximum and the achieved scores
A System Dynamics Approach to Construction Safety Culture
158
Table 7.4 Simulation results of the Enablers, Goals, and CSC index of organization ‘B’
Year Score CSC Maturity Level
Enablers Goals CSC Index
Initial 259.90 200.00 459.90 3rd
1 379.14 249.42 628.56 4th
2 446.86 347.65 794.51 4th
3 478.06 449.15 927.21 5th
4 491.12 487.88 979.00 5th
5 496.38 497.25 993.63 5th
6 498.49 499.39 997.88 5th
7 499.35 499.86 999.22 5th
8 499.72 499.97 999.69 5th
9 499.87 499.99 999.87 5th
Note: Bold numbers refer to the time unit where the organization reaches its fifth level of CSC maturity.
Page 1
1.00 3.25 5.50 7.75 10.00
Years
1:
1:
1:
2:
2:
2:
3:
3:
3:
4:
4:
4:
5:
5:
5:
85
93
100
35
58
80
43
67
90
41
65
90
56
98
140
1: used lds score 2: used pol score 3: used ppl score 4: used prs score 5: used pro score
1
1 1 1
2
2
2 2
3
3
3 3
4
44 4
5
55 5
Figure 7.5 Graphical results of the five enablers of organization ‘B’ over time
A System Dynamics Approach to Construction Safety Culture
159
Page 2
1.00 3.25 5.50 7.75 10.00
Years
1:
1:
1:
260
380
500
1: Enablers Score
1
1
1 1
Figure 7.6 Graphical results of the Enablers score of organization ‘B’ over time
Page 3
1.00 3.25 5.50 7.75 10.00
Years
1:
1:
1:
200
350
500
used goals score: 1 -
1
1
1 1
Figure 7.7 Graphical results of the Goals score of organization ‘B’ over time
A System Dynamics Approach to Construction Safety Culture
160
Page 4
1.00 3.25 5.50 7.75 10.00
Years
1:
1:
1:
460
730
1000
CSC Index Score: 1 -
1
1
1 1
Figure 7.8 Graphical results of the CSC index of organization ‘B’ over time
It took three years for organization ‘B’ to progress from the third (involving) to the fifth
(continually improving) maturity levels (with a CSC index of 927.2 points at the end of
year three, see Table 7.4). The scores of the five enablers at the end of year three
indicated that organization ‘B’ should place more attention on the Processes and People
enablers, as they produced the largest score-gaps compared with the other three enablers
(see Table 7.3).
The following section describes this in more detail with a comparison between the base
run results of organizations ‘A’ and ‘B’. A number of policy analyses were performed
for each organization to enhance its safety performance, achieve a higher CSC index,
and reach the fifth maturity level in a shorter period of time.
7.2.2 Policy Experiments between Organizations ‘A’ and ‘B’
As stated earlier, SD modelling can assist in testing alternative strategies to improve
organizational safety culture, without having to implement them. Consequently,
A System Dynamics Approach to Construction Safety Culture
161
organizations ‘A’ and ‘B’ may need to experiment with different safety policy scenarios
to enhance their CSC indices, and select the best policy that matches their situation.
The next section presents the policy experimentations undertaken by organization ‘A’ to
reach the fifth maturity level within three years; the same as that achieved by
organization ‘B’.
7.2.2.1 Policy Experiments of Organization ‘A’
Organization ‘A’ took five years to mature (to reach the fifth level of CSC maturity, see
Table 7.2). To enable organization ‘A’ to plan to reach the CSC maturity earlier (such
as within three years), a number of policies experimentations need to be conducted.
The gaps of the Leadership and People values were relatively large, when compared
with those of the other three enablers (see Table 7.1). With this in mind, planning for
safety improvement should be performed in the Leadership and People areas, if
organization ‘A’ expects to achieve a higher CSC index value, and reach the fifth
maturity level in a shorter time period. Simulations, with different policy scenarios,
focusing on improving those two enablers (Leadership and People) may be conducted
to achieve the most effective policy. Examples of the policy experiments are described
below.
In reaction to the large gap in the Leadership score, the organization must allocate more
effort to improving the leadership’s four attributes (‘commitment’, ‘communication’,
‘accountability’ and ‘safety and productivity alignment’). The initial values of the five
enablers, Goals, and CSC index reflected the baseline scores listed in Section 7.2.1.1,
with the ‘plds’ then set to, say, as 0.1 (representing the 10% more effort). The dynamic
model was run, and the simulation results predict that the organization reached the CSC
maturity one year earlier (see Tables 7.5 and 7.6) (achieved the fifth level within four
years; one year faster than the base run results, see Table 7.2).
A System Dynamics Approach to Construction Safety Culture
162
Table 7.5 Simulation results of the five enablers of organization ‘A’ with ‘plds’ = 0.1
Year Lds Pol Ppl Prs Pro
Score Gap* Score Gap Score Gap Score Gap Score Gap
Initial 20.00 80.00 19.20 60.80 43.20 46.80 18.00 72.00 37.30 102.70
1 34.40 65.60 33.22 46.78 50.41 39.59 44.51 45.49 69.45 70.55
2 51.33 48.67 50.25 29.75 59.72 30.28 64.16 25.84 98.33 41.67
3 67.81 32.19 64.23 15.77 69.35 20.65 77.08 12.92 119.21 20.79
4 82.58 17.42 72.87 7.13 77.44 12.56 84.31 5.69 131.00 9.00
5 95.53 4.47 77.18 2.82 83.15 6.85 87.75 2.25 136.45 3.55
6 100.00 0.00 78.98 1.02 86.55 3.45 89.17 0.83 138.67 1.33
7 100.00 0.00 79.64 0.36 88.28 1.72 89.71 0.29 139.52 0.48
8 100.00 0.00 79.87 0.13 89.14 0.86 89.90 0.10 139.83 0.17
9 100.00 0.00 79.95 0.05 89.57 0.43 89.96 0.04 139.94 0.06
Note: Bold numbers refer to the time unit where the organization reaches its fifth level of CSC maturity. (*) Gap = the difference between the maximum and the achieved scores
Table 7.6 Simulation results of the Enablers, Goals, and CSC index of organization ‘A’
with ‘plds’ = 0.1
Year Score CSC Maturity Level
Enablers Goals CSC Index
Initial 137.70 129.00 266.70 2nd
1 232.00 155.76 387.76 2nd
2 323.78 227.08 550.87 3rd
3 397.70 319.96 717.65 4th
4 448.19 407.42 855.61 5th
5 480.06 477.79 957.86 5th
6 493.38 494.94 988.32 5th
7 497.14 498.87 996.02 5th
8 498.74 499.75 998.49 5th
9 499.43 499.95 999.38 5th
Note: Bold numbers refer to the time unit where the organization reaches its fifth level of CSC maturity.
Leadership still had the least score (in other words, the gap of the Leadership score was
the largest), when compared with the other four enablers (see Table 7.5). This outcome
A System Dynamics Approach to Construction Safety Culture
163
may indicate that more effort was still needed to further improve the Leadership value.
Thus, the organization may, for example, provide 20% more effort (instead of the 10%
given in the last simulation) to improving this particular enabler. The ‘plds’ was then set
at 0.2. The model was simulated; the results are shown in Tables 7.7 and 7.8.
Table 7.7 Simulation results of the five enablers of organization ‘A’ with ‘plds’ = 0.2
Year Lds Pol Ppl Prs Pro
Score Gap* Score Gap Score Gap Score Gap Score Gap
Initial 20.00 80.00 19.20 60.80 43.20 46.80 18.00 72.00 37.30 102.70
1 41.50 58.50 34.08 45.92 51.16 38.84 44.80 45.20 69.58 70.42
2 63.29 36.71 52.58 27.42 62.10 27.90 65.09 24.91 99.13 40.87
3 83.52 16.48 66.88 13.12 72.79 17.21 78.23 11.77 120.30 19.70
4 99.71 0.29 74.76 5.24 80.71 9.29 85.16 4.84 131.79 8.21
5 100.00 0.00 78.11 1.89 85.37 4.63 88.19 1.81 136.85 3.15
6 100.00 0.00 79.33 0.67 87.70 2.30 89.35 0.65 138.84 1.16
7 100.00 0.00 79.76 0.24 88.85 1.15 89.77 0.23 139.58 0.42
8 100.00 0.00 79.92 0.08 89.43 0.57 89.92 0.08 139.85 0.15
9 100.00 0.00 79.97 0.03 89.72 0.28 89.97 0.03 139.95 0.05
Note: Bold numbers refer to the time unit where the organization reaches its fifth level of CSC maturity. (*) Gap = the difference between the maximum and the achieved scores
A System Dynamics Approach to Construction Safety Culture
164
Table 7.8 Simulation results of the Enablers, Goals, and CSC index of organization ‘A’
with ‘plds’ = 0.2
Year Score CSC Maturity Level
Enablers Goals CSC Index
Initial 137.70 129.00 266.70 2nd
1 241.12 155.76 396.88 2nd
2 342.19 227.36 569.55 3rd
3 421.73 333.83 755.55 4th
4 472.13 431.83 903.96 5th
5 488.53 483.78 972.31 5th
6 495.21 496.32 991.54 5th
7 497.97 499.18 997.15 5th
8 499.11 499.82 998.93 5th
9 499.60 499.96 999.56 5th
Note: Bold numbers refer to the time unit where the organization reaches its fifth level of CSC maturity.
By setting the ‘plds’ = 0.2, the organization reached the fifth CSC maturity level in four
years (see Tables 7.7 and 7.8) (which was the same time frame of when the ‘plds’ =
0.1). However, the scores of the five enablers, as well as the CSC index, at the end of
year four, appeared to be higher than those obtained when the ‘plds’ = 0.1 (see Tables
7.5 and 7.6).
Consequently, the organization needed further experiments with different ‘extra’ efforts
for improving the scores of the five enablers to achieve a higher Goals score and reach
the CSC maturity as planned.
Three examples of policy scenarios that organization ‘A’ may apply to enhance the CSC
index, and achieve the fifth CSC maturity level within three years are shown below.
� Providing 20% more of effort to improve the Leadership, People, and Processes
enablers (‘plds’, ‘pppl’, and ‘ppro’ = 0.2). The CSC index at the end of year three
A System Dynamics Approach to Construction Safety Culture
165
was 802.21 points (the organization reached the fifth CSC maturity level in three
years), as shown in Table 7.9.
� Providing 30, 10, and 10% more of effort to increase the Leadership, People, and
Processes scores, respectively (‘plds’ = 0.3, ‘pppl’ = 0.1, and ‘ppro’ = 0.1). The
organization reached the fifth maturity level, with the CSC index of 804.37 points,
at the end of year three (see Table 7.10).
� Providing 20% more of effort to improving the Leadership and Processes enablers
(‘plds’ and ‘ppro’ = 0.2), and 10% more of effort to increase the People,
Partnership and Resources, and Policy and Strategy scores (‘pppl’, ‘pprs’, and
‘ppol’ = 0.1). The CSC index at the end of year three was 805.59 points (see Table
7.11).
Table 7.9 Simulation results of organization ‘A’ with ‘plds’, ‘pppl’, and ‘ppro’ = 0.2
Year Score Level*
Lds Pol Ppl Prs Pro Enablers Goals CSC Index
Initial 20.00 19.20 43.20 18.00 37.30 137.70 129.00 266.70 2nd
1 41.45 34.13 58.68 46.20 84.80 265.26 158.80 424.06 3rd
2 63.67 53.03 72.14 67.97 115.96 372.78 247.84 620.63 4th
3 83.14 67.34 81.33 80.62 131.49 443.92 358.29 802.21 5th
4 99.51 75.02 86.35 86.45 137.39 484.71 466.35 951.07 5th
5 100.00 78.21 88.58 88.74 139.25 494.78 492.42 987.20 5th
6 100.00 79.36 89.45 89.56 139.79 498.16 498.32 996.48 5th
7 100.00 79.77 89.78 89.85 139.94 499.35 499.63 998.98 5th
8 100.00 79.92 89.92 89.95 139.98 499.77 499.92 999.69 5th
9 100.00 79.97 89.97 89.98 140.00 499.92 499.98 999.90 5th
Note: Bold numbers refer to the time unit where the organization reaches the fifth level of CSC maturity. (*) Level = CSC maturity level
A System Dynamics Approach to Construction Safety Culture
166
Table 7.10 Simulation results of organization ‘A’ with ‘plds’ = 0.3, and ‘pppl’ and
‘ppro’ = 0.1
Year Score Level*
Lds Pol Ppl Prs Pro Enablers Goals CSC Index
Initial 20.00 19.20 43.20 18.00 37.30 137.70 129.00 266.70 2nd
1 48.03 34.92 55.74 45.79 77.66 262.13 157.33 419.46 3rd
2 73.71 54.86 69.36 67.42 109.13 374.48 242.62 617.10 4th
3 94.12 69.04 79.58 80.34 127.60 450.68 353.70 804.37 5th
4 100.00 75.94 85.34 86.35 135.67 483.30 463.91 947.21 5th
5 100.00 78.54 87.95 88.69 138.58 493.76 491.74 985.50 5th
6 100.00 79.48 89.09 89.54 139.55 497.66 498.16 995.82 5th
7 100.00 79.82 89.60 89.84 139.86 499.11 499.59 998.71 5th
8 100.00 79.93 89.82 89.94 139.96 499.66 499.91 999.57 5th
9 100.00 79.98 89.92 89.98 139.99 499.87 499.98 999.85 5th
Note: Bold numbers refer to the time unit where the organization reaches the fifth level of CSC maturity. (*) Level = CSC maturity level
Table 7.11 Simulation results of organization ‘A’ with ‘plds’ and ‘ppro’ = 0.2, and
‘pppl’, ‘pprs’, and ‘ppol’ = 0.1
Year Score Level*
Lds Pol Ppl Prs Pro Enablers Goals CSC Index
Initial 20.00 19.20 43.20 18.00 37.30 137.70 129.00 266.70 2nd
1 41.45 39.21 55.07 50.32 85.19 271.25 158.84 430.09 3rd
2 63.66 59.01 67.69 71.53 116.62 378.52 248.19 626.71 4th
3 83.10 71.40 77.80 82.74 131.88 446.91 358.68 805.59 5th
4 99.44 77.05 84.19 87.50 137.53 485.71 466.57 952.27 5th
5 100.00 79.07 87.43 89.21 139.29 495.01 492.48 987.49 5th
6 100.00 79.71 88.87 89.75 139.80 498.33 498.14 996.47 5th
7 100.00 79.91 89.50 89.92 139.95 499.63 499.28 998.91 5th
8 100.00 79.97 89.78 89.98 139.99 499.92 499.71 999.63 5th
9 100.00 79.99 89.90 89.99 140.00 499.98 499.88 999.87 5th
Note: Bold numbers refer to the time unit where the organization reaches the fifth level of CSC maturity. (*) Level = CSC maturity level
A System Dynamics Approach to Construction Safety Culture
167
In summary, organization ‘A’ is currently in the early stage of its maturity level, with
Leadership found to be crucial if the organization aspires to progress through to higher
maturity levels. The focus of the organization, therefore, should be on a more effective
implementation process of the Leadership’s attributes, as described below:
� Leaders must take safety more seriously (Lingard and Blismas, 2006).
� Leaders must act as a role model in behaving safely (Dunlap, 2004).
� Leaders should continue to encourage workers to give opinions and/or suggestions
on safety matters (Little, 2002).
� Leaders are expected to educate workers, and ensure that they hold safety
responsibilities, for both themselves, and their workmates (Dias and Coble, 1996).
� Leaders should respond quickly to correct safety problems when they are brought to
their attention (Teo et al., 2005).
7.2.2.2 Policy Experiments of Organization ‘B’
The base run results of organization ‘B’ illustrated that it took three years for the
organization to progress from the second to the fifth levels of CSC maturity. For
organization ‘B’ to achieve maturity earlier (less than three years), a number of
sensitivity analyses, with, say, 10% extra effort being given to improve each enabler
(the ‘plds’, ‘pppl’, ‘pprs’, ‘ppol’, and ‘ppro’) need to be undertaken. The analyses help
to identify which enabler has the potential to increase the CSC index so that the
organization reaches the CSC maturity level earlier.
The sensitivity analysis results, illustrated in Table 7.12, demonstrate that, by giving the
10% more effort to enhance the score of Processes (‘ppro’ = 0.1), organization ‘B’
achieved its maturity one year earlier, i.e. within two years. The results of the other four
enablers, however, showed no advancement in the organization achieving the fifth CSC
maturity level earlier.
A System Dynamics Approach to Construction Safety Culture
168
Table 7.12 The CSC index of organization ‘B’ when more of effort is given to enhance
each enabler
Year CSC Index
Base run Plds = 0.1 Ppol = 0.1 Pppl = 0.1 Pprs = 0.1 Ppro = 0.1
Initial 459.90 459.90 459.90 459.90 459.90 459.90
1 628.56 629.71 631.44 632.15 631.63 635.25
2 794.51 794.69 797.20 798.49 797.13 802.76
3 927.21 927.31 928.79 930.02 928.67 942.50
4 979.00 979.04 979.67 980.58 979.67 983.50
5 993.63 993.64 993.89 994.47 993.91 994.93
Note: Bold numbers refer to the time unit where the organization reaches the fifth level of CSC maturity
In conclusion, then, to improve safety performance and achieve the fifth CSC maturity
level, in a shorter time period, organization ‘B’ should focus on enhancing the
improvement of the Processes enabler, as it facilitates a faster CSC maturity
achievement. The organization may need to:
� Provide adequate safety training, especially for new staff, to ensure that the job is
performed safely (Tam et al., 2004);
� Have a routine risk and hazard assessment (Berg, 2006);
� Keep the site housekeeping at a high level (Zohar, 1980);
� Adopt a no-blame approach, and learn from experience (ICAO, 1992); and
� Have a good safety benchmarking system to compare the organization’s safety with
that of other construction organizations (Taylor, 2003).
7.3 THE CYCLICAL STYLE OF SAFETY MANAGEMENT
The policy experiments, described in Section 7.2, demonstrate the value in identifying
areas for safety improvement to progress through to higher CSC maturity levels, and to
A System Dynamics Approach to Construction Safety Culture
169
achieve the maximum score of the CSC index (1,000 points). In real-life situations,
however, the CSC index may never reach its maximum score. One key reason is top
management’s view of the fifth maturity level as a target, not as means of continual
improvement. Once the fifth maturity level is reached, top management tends to slow
the momentum behind all safety activities. This phenomenon is known as ‘attention
withdrawal’, and is illustrated in the accident cycle (shown in Figures 7.9 and 7.10),
where top management gradually and slowly withdraws its attention to safety when
safety performance reflects the highest level of maturity (NPS Risk Management
Division, 2006).
M anagem ent tak es pro m pt safe ty ac tion s
M anagem ent slow ly w ithdraw s its attention to safety
(M anagem en t satisfaction w ith reach ing the fifth C S C m aturity level)
Safety Perform ance
Acc
iden
t dec
reas
es
Acc
iden
t inc
reas
es
U pper lim it(M anagem en t sa tisfaction w ith safety perform ance)
T im e
L ow er lim it
Figure 7.9 The accident cycle (Adapted from NPS Risk Management Division, 2006)
A System Dynamics Approach to Construction Safety Culture
170
Stage 1
Occurrence of a detrimental event
Stage 2
Identification of shortfalls and recrimination
Stage 3
Spotlight firmly on safety
Stage 4
Strongly supported campaign for major safety improvements and implementation
Stage 5
A growing history of safety success comfort
Stage 6
Attention slowly being focused elsewhere
Stage 7
Safety concerns raised but judged as not sufficiently important (or welcomed) in the overall scheme of things
Figure 7.10 The normal accident cycle (Adapted from Jones, 2007)
When the accident rate is high (reaches the lower limit, see Figure 7.9), it becomes top
management’s priority to reduce the number of accidents (as shown in Stages 1 to 3 of
Figure 7.10). Top management identifies the shortfalls and priorities, and then acts
promptly to enhance safety improvements. Indeed, top management may consider, for
example, providing more safety resources to carry out the job safely, empowering safety
responsibilities to staff, and encouraging more two-way safety communication (as
shown in Stage 4 of Figure 7.10).
As top management pays more attention to reducing the number of accidents, the
accident rate decreases (representing Stage 5 of Figure 7.10). This continues until the
accident rate reaches the hypothetical upper limit of management satisfaction with
safety performance (see Figure 7.9), then top management starts to unintentionally
withdraw its attention to safety (as shown in Stage 6 of Figure 7.10). For example, the
financial resources provided to support the acquisition of safety resources may be cut
A System Dynamics Approach to Construction Safety Culture
171
down (see the negative relationship between safety resources and management
commitment shown in the feedback loop ‘B’ of Figure 6.3). Further, top management
may also shift the focus to increasing production (as shown in Stage 7 of Figure 7.10),
in which, then, puts pressure on the staff. As pressure increases, the distress increases,
leading to the intensifying of the accident rate (see the positive relationship between the
distress and accident rate shown in the feedback loop ‘A’ of Figure 6.3).
Consequently, the accident rate increases as management withdraws its safety attention
until the hypothetical lower limit is reached (see Figure 7.9). Top management then
reacts promptly and swiftly by taking actions to reduce the accident rate again, and thus
the cycle continues.
SD modelling is used to better understand the changes in the Enablers, Goals, and CSC
index scores resulting from the cyclical style of safety management. The details of these
changes are discussed below.
7.3.1 The Dynamic Model of the Cyclical Style of Safety Management
The cyclical style of safety management was modelled with SD modelling. The
assumption made in the modelling process was that top management withdraws its
attention to safety when the CSC index reaches 95% of its maximum score
(representing the upper limit of management satisfaction with safety performance, see
Figure 7.11). The 95% level was selected as it represents a very high confidence in the
organization’s safety management ability, and any accidents that might occur could be
largely traced to random events represented by the 5% error level. At this point, top
management starts gradually shifting its safety attention to other areas for improvement,
believing that an adequate safety management system is in place, and the effective
implementation of this system will continue, regardless of the level of management
support/attention.
A System Dynamics Approach to Construction Safety Culture
172
Management takes prompt safety actions
(i.e. safety performance reflects the fourth CSC maturity level)
Management withdraws its attention to safety
(i.e. high CSC index value)
(Management satisfaction with reaching the fifth CSC maturity level)
Upper limit(Management satisfaction with much improved safety performance)
CSC Index
Time
950
800 Lower limit
The
fift
h C
SC m
atur
ity le
vel
Posi
tive
slop
e Negative slope
“Comfort Zone”
Figure 7.11 The CSC index cycle as management withdraws attention to safety
Eighty percent of the maximum score of the CSC index, on the other hand, was chosen
as the lower limit (see Figure 7.11). At this point, the organization is falling into the
lower maturity level, i.e. from the fifth to the fourth maturity levels. Top management
realises the problem, and starts taking actions to improve the CSC index.
The dynamic model of the cyclical style of safety management is shown in Figure 7.12.
The upper and lower limits (the CSC index of 950 and 800 points, respectively) were
used in the SD equations, as described in the dynamic models of the five enablers and
Goals below.
A System Dynamics Approach to Construction Safety Culture
173
LEADERSHIP
rlds
used lds
dlds
GOALS
rgoals
CSC INDEX
ggoals
dgoals
slope
used pol
used ppl used prs
used pro
rldsf
PEOPLE
rppl
used goals
Co lds ppl
DF ppl ldsgppl
used ppl
ggoals
dppl
used ldsDF goals pro Co pro goals
POLICY & STRATEGY
gpol
gpol
DF pol lds
rpol
DF pol prs
Co prs pol
Co lds pol
used pol
dpol
PARTNERSHIPS & RESOURCES
rprs
used pol
used prsdprs
gprs
DF prs lds
Co lds prs
rlds2
ENABLERS
DF prs ppl
Co ppl prs
gprs
PROCESSES
rpro
DF pro ppl
Co ppl pro
used pro
dpro
gpro
DF pro pol
Co pol pro
rlds2
rlds2 rlds2
gpro
rlds2
rppl2
rlds2
rprs2 rpol2
Co lds pol
rpro2
glds
Co lds ppl
Co ppl pro
Co pol pro
slope
slope
desired CSC INDEX
rgoals2 CSC flow
Co lds ppl
CSC INDEX
CSC INDEX
rlds3
rlds3
rlds3rlds3
rlds3
rlds3dCSC INDEX
dCSC INDEX
Co prs pol
Note: Co_lds_pol, Co_lds_ppl, Co_lds_prs, Co_pol_pro, Co_ppl_pro, Co_ppl_prs, Co_pro_goals, and Co_prs_pol = Correlations between Lds and Pol, Lds and Ppl, Lds and Prs, Pol and Pro, Ppl and Pro, Ppl and Prs, Pro and Goals, and Prs and Pol, respectively. DF_goals_pro, DF_pol_lds, DF_pol_prs, DF_ppl_lds, DF_pro_pol, DF_pro_ppl, DF_prs_lds, and DF_prs_ppl = Decision fractions between Goals and Pro, Pol and Lds, Pol and Prs, Ppl and Lds, Pro and Pol, Pro and Ppl, Prs and Lds, and Prs and Ppl, respectively. dgoals, dlds, dpol, dppl, dpro, dprs,= Desired Goals, desired Lds, desired Pol, desired Ppl, desired Pro, and desired Prs, respectively. ggoals, glds, gpol, gppl, gpro, and gprs = Gaps of Goals, Lds, Pol, Ppl, Pro, and Prs, respectively. plds, ppol, pppl, ppro, and pprs = Percentage effort provided to Lds, Pol, Ppl, Pro, and Prs, respectively. rgoals, rlds, rpol rppl, rpro, and rprs = Goals, Lds, Pol, Ppl, Pro, and Prs rates, respectively. used_goals, used_lds, used_pol, used_ppl, used_pro, and used_prs = Goals, Lds, Pol, Ppl, Pro, and Prs values, respectively. rldsf = Lds rate fraction
Figure 7.12 The dynamic model of the cyclical style of safety management
A System Dynamics Approach to Construction Safety Culture
174
7.3.1.1 Leadership Dynamic Model
The Leadership dynamic model is shown in Figure 7.13. The upper and lower limits of
the CSC index (800 and 950 points, respectively) were adopted in the Lds equations
(see full details of SD equations of the cyclical style of safety management model in
Appendix 12). Equations 7.1 to 7.3 demonstrate that there will be no inflows of the Lds
score (‘rlds’ = 0), if the CSC index exceeds the upper limit of management satisfaction
with safety performance (950 points, see Figure 7.11); or the slope of the CSC index is
negative (slope < 0, see Figure 7.11), which represents the onset of withdrawing
attention to safety.
As the attention to safety gets withdrawn, the Lds score decreases (there are outflows of
the Lds score, ‘rlds2’ or ‘rlds3’ >0), leading to a reduced CSC index. The ‘attention
withdrawal’ continues until the CSC index reaches the lower limit (800 points, see
Figure 7.11), then top management reacts through taking actions to increase the CSC
index again (the Lds score increases, ‘rlds’ > 0, and ‘rlds2’ or ‘rlds3’ = 0).
LEADERSHIP
rlds
used lds
dlds
rldsf
ggoals
rlds2
gldsslope
slope
CSC INDEX
CSC INDEX
rlds3
dCSC INDEX
Leadership Dynamic Model
Note: See Acronyms list for the abbreviations. The symbols are described in Figure 2.6 and Table 2.1.
Figure 7.13 Leadership dynamic model
A System Dynamics Approach to Construction Safety Culture
175
Inflows:
Eq 7.1 rlds = IF (CSC_INDEX > dCSC_INDEX) OR ((800 <
CSC_INDEX < dCSC_INDEX) AND (slope < 0)) THEN
(0) ELSE ((used_lds + ggoals)*rldsf
Outflows:
Eq. 7.2 rlds2 = IF (CSC_INDEX < 800) OR ((800 < CSC_INDEX <
dCSC_INDEX) AND (slope > 0)) THEN (0) ELSE ((glds
+ ggoals)*rldsf
Eq. 7.3 rlds3 = IF (CSC_INDEX > dCSC_INDEX) AND (rlds = 0) AND
(rlds2 = 0) THEN ((glds + ggoals)*rldsf) ELSE (0)
The decreased Lds score negatively affects the score of the People enabler, as these two
enablers have a clear positive relationship (see Figure 5.4). The details of this effect are
described in the following section.
7.3.1.2 People Dynamic Model
The People dynamic model is illustrated in Figure 7.14. As top management withdraws
attention to safety (‘rlds2’ or ‘rlds3’ > 0), there tends to be a decrease in people’s
perception of, and participation in, safety (see the feedback loop ‘B’ of Figure 6.3),
which, in turn, reduces the Ppl score (there is an outflow of the Ppl score, ‘rppl2’ > 0)
(see Equations 7.4 and 7.5). Normally, the amount of score reduction depends on the
correlation strength between the Lds and Ppl enablers, as demonstrated in Equation 7.5.
A System Dynamics Approach to Construction Safety Culture
176
used lds
PEOPLE
rppl
Co lds ppl
DF ppl ldsgppl
used ppl
dppl
rlds2
rppl2
rlds3
People Dynamic Model
Note: See Acronyms list for the abbreviations. The symbols are described in Figure 2.6 and Table 2.1.
Figure 7.14 People dynamic model
Inflows:
Eq. 7.4 rppl = (used_lds*DF_ppl_lds)
Outflows:
Eq. 7.5 rppl2 = IF (rlds2 > 0) OR (rlds3 > 0) THEN (rppl + (rppl*
Co_lds_ppl)) ELSE (0)
As both scores of the Lds and Ppl enablers continue to decrease, they tend to reduce the
Partnerships and Resources score (see Figure 6.5 for the positive relationships between
these three enablers). The details for this model are described below.
7.3.1.3 Partnerships and Resources Dynamic Model
The reduction in the Lds score, affected by the cyclical style of safety management,
decreases the Partnerships and Resources score directly and indirectly, through the Ppl
enabler (see Figure 7.15, and Equations 7.6 and 7.7). For instance, the financial
resources provided to support the acquisition of safety resources may be cut down as
A System Dynamics Approach to Construction Safety Culture
177
management shifts its focus to increasing production (management withdraws attention
from safety to other areas needing improvement).
used lds
PARTNERSHIPS & RESOURCES
rprs
used ppl
used prs
dprs
gprs
DF prs lds
Co lds prs
DF prs ppl
Co ppl prs
Co lds ppl
gprs
rlds2
rprs2
rlds3
Partnerships and Resources Dynamic Model
Note: See Acronyms list for the abbreviations. The symbols are described in Figure 2.6 and Table 2.1.
Figure 7.15 Partnerships and Resources dynamic model
Inflows:
Eq. 7.6 rprs = (used_lds*DF_prs_lds) + (used_ppl*DF_prs_ppl)
Outflows:
Eq. 7.7 rprs2 = IF (rlds2 > 0) OR (rlds3 > 0) THEN (rprs + (rprs*
Co_lds_prs) + (rprs* Co_lds_ppl*Co_ppl_prs)) ELSE (0)
The amount of the Prs score decrease depends on the correlation strength between this
particular enabler, and the Lds and Ppl enablers, as illustrated in Equation 7.7. As the
Lds and Prs enablers have a direct effect on the Policy and Strategy enabler, the
decreased Lds and Prs scores, undoubtedly, reduce the Pol score. The details of the
Policy and Strategy dynamic model are described below.
A System Dynamics Approach to Construction Safety Culture
178
7.3.1.4 Policy and Strategy Dynamic Model
The dynamic model of Policy and Strategy (see Figure 7.16) demonstrates that the
increased Lds score increases the Pol score (the Lds enabler has a positive relationship
with the Pol enabler, see Figure 6.5) (see Equation 7.8). As top management’s attention
to safety being withdrawn (in other words, the Lds score is reduced), the Pol score
decreases. The amount of the Pol score decrease depends on the correlation strength
between this enabler, and the Lds and Prs enablers (as Lds has direct and indirect,
through Prs, effects on Pol) (see Equation 7.9).
used ldsPOLICY & STRATEGY
gpol
gpol
DF pol lds
used prs
rpol
DF pol prs
Co prs pol
Co lds pol
used pol
dpol
rlds2
rpol2
Co lds prs
rlds3
Policy and Strategy Dynamic Model
Note: See Acronyms list for the abbreviations. The symbols are described in Figure 2.6 and Table 2.1.
Figure 7.16 Policy and Strategy dynamic model
Inflows:
Eq. 7.8 rpol = (used_lds*DF_pol_lds) + (used_prs*DF_pol_prs)
Outflows:
Eq. 7.9 rpol2 = IF (rlds2 > 0) OR (rlds3 > 0) THEN (rpol + (rpol*
Co_lds_pol) + (rpol* Co_lds_prs*Co_prs_pol)) ELSE (0)
A System Dynamics Approach to Construction Safety Culture
179
The ‘attention withdrawal’ (the decrease in the Lds score) not only has a direct effect on
the Ppl, Prs, and Pol scores, but also has an indirect influence in the reduction of the
Process score through the Ppl and Pol enablers (see Figure 5.4). This decrease is
described in the Processes dynamic model below.
7.3.1.5 Processes Dynamic Model
The Processes dynamic model (see Figure 7.17) shows that the decreased Lds, Ppl, and
Pol scores, as management withdraws its attention to safety, lower the Pro score (see
Equations 7.10 and 7.11). The score decrease depends on the correlation strength
between the Pro and the Lds, Ppl, and Pol enablers.
used ppl
used polPROCESSES
rpro
DF pro ppl
Co ppl pro
used prodprogpro
DF pro pol
Co pol progpro rlds2
rpro2
Co lds ppl
Co lds pol
rlds3
Processes Dynamic Model
Note: See Acronyms list for the abbreviations. The symbols are described in Figure 2.6 and Table 2.1.
Figure 7.17 Processes dynamic model
Inflows:
Eq. 7.10 rpro = (used_ppl*DF_pro_ppl) + (used_pol*DF_pro_pol)
A System Dynamics Approach to Construction Safety Culture
180
Outflows:
Eq. 7.11 rpro2 = IF (rlds2 > 0) OR (rlds3 > 0) THEN (rpro + (rpro*
Co_lds_ppl*Co_ppl_pro) + (rpro*Co_lds_pol*
Co_pol_pro))ELSE (0)
The decrease of the Pro score, undoubtedly, negatively affects the Goals score (as these
two constructs have a strong positive relationship, see Figure 5.4). The details of the
Goals dynamic model are presented below.
7.3.1.6 Goals Dynamic Model
In the Goals dynamic model (see Figure 7.18), the increased Pro score enhances the
Goals score (as they have a strong positive relationship, see Figure 5.4) (see Equation
7.12); the reduced Pro score leads to the decreased Goals score (as demonstrated in
Equation 7.13).
GOALS
rgoals
ggoals
dgoals
used pro
used goalsDF goals pro Co pro goals
rlds2Co lds ppl
Co ppl pro
rgoals2
Co lds pol
Co pol pro
rlds3
CSC INDEX
Goals Dynamic Model
Note: See Acronyms list for the abbreviations. The symbols are described in Figure 2.6 and Table 2.1.
Figure 7.18 Goals dynamic model
A System Dynamics Approach to Construction Safety Culture
181
Inflows:
Eq. 7.12 rgoals = used_pro*DF_goals_pro
Outflows:
Eq. 7.13 rgoals2 = IF (rlds2 > 0) OR (rlds3 > 0) THEN (rgoals + (rgoals*
Co_lds_pol*Co_pol_pro*Co_pro_goals) + (rgoals*
Co_lds_ppl*Co_ppl_pro* Co_pro_goals)) ELSE (0)
The reduction of the Goals score (Equation 7.13), as affected by the ‘attention
withdrawal’, is influenced by Pro, which is also being effected by Lds through Ppl and
Pol. The smaller Goals score, certainly, leads to a larger ‘gap of goals’ (ggoals), which,
in turn, will affect the Lds enabler (as ‘rlds’ depends on ‘ggoals’, see Equation 7.1), and
the simulation repeats as cycles.
The ‘attention withdrawal’ causes a reduction in the CSC index. This continues until
the index score reaches the lower limit (800 points, see Figure 7.11), then management
takes prompt actions to improve safety implementation. Management achieves this, for
example, by getting workers involved in safety activities, encouraging feedback on
safety, providing adequate safety resources, and reassigning safety responsibilities to
staff. As top management returns its focus on safety improvement, the CSC index
begins to increase. This continues until the index score exceeds the upper limit
corresponding to management’s satisfaction with the safety record (assumed herein as
950 points, see Figure 7.11), then management starts to withdraw its attention away
from safety again, and the cycle continues.
The simulation results, as an effect of the ‘attention withdrawal’ phenomenon, are
described in Section 7.3.2 below.
A System Dynamics Approach to Construction Safety Culture
182
7.3.2 The Simulation Results
The dynamic model of the cyclical style of safety management is simulated to
investigate the changes in the Enablers, Goals, and CSC index scores as reflected by
management attention withdrawal. The lower limit used in the simulation was, as stated
earlier, set at the CSC index = 800 points. According to the base run results (previously
shown in Tables 6.1 and 6.2), this index score was almost achieved (the CSC index =
797.90 points) at the end of year 10. Therefore, the scores of the five enablers and Goals
at this point of time were used as the initial values for the simulation. The specifics are
detailed below:
� ‘used_lds’ = 60.37 (out of 100 points)
� ‘used_ppl’ = 68.07 (out of 90 points)
� ‘used_prs’ = 82.23 (out of 90 points)
� ‘used_pol’ = 73.17 (out of 80 points)
� ‘used_pro’ = 132.44 (out of 140 points)
� ‘used_goals’ = 381.64 (out of 500 points)
The initial values of the five enablers and Goals were adopted in the SD equations, and
the model was simulated. The simulation results illustrate that at the beginning of the
simulation, the CSC index, which reflects the fourth CSC maturity level, increased as
the Enablers and Goals’ scores increased (see Table 7.13, and Figures 7.19 to 7.21),
demonstrating safety improvement. At the end of year three, however, the CSC index
reached its specified upper limit (950 points, see Figure 7.11), and management started
to unintentionally withdraw its attention from safety to other areas requiring
improvement. Thus, there was a slight drop in the five enablers’ score, which led to a
decrease in the Goals and the CSC index scores.
A System Dynamics Approach to Construction Safety Culture
183
Table 7.13 Simulation results of the cyclical style of safety management
Year Score Lds Pol Ppl Prs Pro Enablers Goals CSC Index Initial 60.37 73.17 68.07 82.23 132.44 416.28 381.64 797.92 1 72.05 76.76 75.79 86.46 136.87 447.93 426.25 874.18
2 83.05 78.62 81.56 88.54 138.78 470.54 460.13 930.67
3 89.63 79.21 84.07 89.17 139.33 481.42 472.84 954.25*
4 86.19 78.69 81.61 88.55 138.85 473.89 462.28 936.16
5 81.47 77.84 78.31 87.48 138.05 463.15 447.68 910.83
6 74.98 76.52 74.03 85.75 136.74 448.02 427.59 875.61
7 66.05 74.57 68.78 83.06 134.64 427.10 400.15 827.25
8 60.41 73.31 65.88 81.26 133.18 414.04 382.97 797.01**
9 63.96 75.31 70.31 83.95 135.54 429.08 401.28 830.36
10 75.98 77.86 77.65 87.33 138.20 457.02 446.37 903.39
11 85.94 79.11 82.85 88.93 139.31 476.14 471.15 947.29
12 89.27 79.28 83.68 89.12 139.45 480.79 475.09 955.88*
13 86.01 78.80 81.07 88.44 139.05 473.38 465.39 938.77
14 81.57 78.03 77.55 87.32 138.40 462.87 451.97 914.84
15 75.49 76.82 72.98 85.49 137.32 448.11 433.47 881.58
16 67.16 75.03 67.33 82.68 135.60 427.80 408.10 835.90
17 58.95 74.29 65.12 81.37 134.88 414.61 385.58 800.18**
18 71.62 77.28 73.85 85.96 137.86 446.58 437.68 884.26
19 81.81 78.83 80.36 88.30 139.16 468.46 466.43 934.89
20 90.90 79.53 84.64 89.34 139.68 484.10 482.00 966.10*
21 88.44 79.22 82.37 88.83 139.46 478.31 474.97 953.29
22 85.10 78.70 79.25 87.96 139.08 470.10 465.23 935.33
23 80.57 77.88 75.08 86.53 138.45 458.50 451.74 910.24
24 74.38 76.60 69.67 84.27 137.42 442.35 433.14 875.49
25 65.93 74.72 63.04 80.88 135.81 420.38 407.63 828.01
26 60.60 73.51 59.36 78.69 134.72 406.88 391.58 798.46**
27 64.00 75.43 65.01 81.93 136.49 422.86 407.82 830.67
28 75.58 77.90 74.30 86.26 138.54 452.58 450.09 902.67
29 85.25 79.12 80.88 88.45 139.43 473.13 473.18 946.31
30 94.18 79.66 85.05 89.40 139.79 488.08 485.63 973.71*
31 92.33 79.42 82.87 88.94 139.64 483.19 480.02 963.21
32 89.80 79.02 79.79 88.14 139.38 476.13 472.22 948.36
. . . . . . . . .
. . . . . . . . . 112 100.00 80.00 90.00 90.00 140.00 500.00 500.00 1,000.00 Note: * The CSC index reaches its upper limit. ** The CSC index reaches its lower limit.
A System Dynamics Approach to Construction Safety Culture
184
Page 11.00 7.00 13.00 19.00 25.00
Years
1:
1:
1:
350
425
500
1: Enablers Score
11
11
Figure 7.19 Graphical results of the Enablers score as the effect of
the attention withdrawal
Page 41.00 7.00 13.00 19.00 25.00
Years
1:
1:
1:
350
425
500
1: used goals score
1
1
1
1
Figure 7.20 Graphical results of the Goals score as the effect of
the attention withdrawal
A System Dynamics Approach to Construction Safety Culture
185
Page 21.00 7.00 13.00 19.00 25.00
Years
1:
1:
1:
700
850
1000
1: CSC Index Score
1
1
1
1
Figure 7.21 Graphical results of the CSC index as the effect of the attention withdrawal
The CSC index continued to decrease gradually until it reached the lower limit of 800
points at the end of year eight. Subsequently, top management took prompt actions to
improve safety implementation, in response to the fear that the organization would fall
to a lower maturity level (from the fifth to the fourth maturity levels). As a result, there
was a relatively large increment in the Lds score from 60.41 points, at the end of year
eight, to 79.58 points, at the end of year 10 (an increase of more than 19% of its
maximum score, see Table 7.13).
Leadership action to improve safety implementation, obviously, enhanced the
implementation of the Ppl, Pol, Prs, and Pro enablers (as seen by the increase of these
four enablers’ scores, see Table 7.13), leading to a higher Goals score and, ultimately,
the CSC index. The actions taken to improve the CSC index continues until the index
exceeds the assumed upper limit (the CSC index of 950 points), then the ‘attention
withdrawal’ takes place again (top management shifts attention from safety to other
areas for improvement), and the cycle continues.
Simulation results (see Figure 7.21 and Table 7.13) show that the CSC index score
oscillates between the fourth and the fifth CSC maturity levels. However, it slowly aims
A System Dynamics Approach to Construction Safety Culture
186
towards the maximum score of 1,000 point (the maximum score of the CSC index),
over a very long term.
7.3.3 Conclusion of the Cyclical Style of Safety Management
SD modelling was used to model the cyclical style of safety management to imitate the
situation where management withdraws its attention to safety. The ‘attention
withdrawal’ occurs when the CSC index exceeds the assumed upper limit of 950 points.
Top management, then, started to gradually withdraw its attention to safety (for
example, its focus from safety implementation to product enhancement), believing that
safety management system is in place, and the effective safety implementation will
continue, regardless of management support. These leadership actions negatively
affected the workers’ perception of, and participation in, safety, which, in turn, led to
less job satisfaction and lower workforce morale (as indicated by the lower scores of
Ppl, Prs, Pol, and Pro, as well as Goals).
The decreased Enablers and Goals scores undoubtedly reduce the CSC index. This
continues until the index score reaches the assumed lower limit of 800 points, indicating
that the organization has fallen into the lower (the fourth) maturity level. At that stage,
top management takes prompt safety actions to increase the CSC index by: 1)
encouraging staff to put forward their opinions about how to improve safety, 2)
providing safety training, especially to new staff; and 3) promoting safety campaigns to
enhance safety awareness.
Such positive safety implementation improves the Enablers score, which, in turn,
enhances the Goals and CSC index scores. This approach continues until the CSC index
surpasses the upper limit of management satisfaction with its safety performance, then
the ‘attention withdrawal’ starts again, and the cycle continues.
The results from this study show that the CSC index score oscillates between the fourth
and the fifth CSC maturity levels, as an effect of the cyclical nature of safety
A System Dynamics Approach to Construction Safety Culture
187
management. However, the organization, once again, turns towards the maximum index
score of 1,000 point, over a very long period.
A System Dynamics Approach to Construction Safety Culture
188
[THIS PAGE IS INTENTIONALLY LEFT BLANK]
A System Dynamics Approach to Construction Safety Culture
189
88 SSTTUUDDYY FFIINNDDIINNGGSS AANNDD RREECCOOMMMMEENNDDAATTIIOONNSS
FFOORR FFUUTTUURREE RREESSEEAARRCCHH
8.1 GENERAL OVERVIEW
This chapter brings together the major findings of this study. A review and synthesis of
the existing body of knowledge is presented below. The implications of this research
and the findings for the Thai construction industry, and the recommendations for future
research are discussed at the end of the chapter.
8.2 MAJOR FINDINGS
The main objective of this study was to investigate the interactions and causal
relationships among the key factors (enablers and Goals) of the construction safety
culture (CSC). An understanding of their individual, or combined, effects on an
organization’s ability to achieve safety performance improvements was also sought.
With this in mind, the following secondary objectives were identified:
� Review the literature regarding the nature of the CSC and its key components, and
identify the tools (well-established performance measurement systems) available for
measuring safety culture.
� Develop a CSC model based on a widely used performance measurement system,
the EFQM Excellence model, and investigate the interactions and causal
relationships among its key factors (CSC enablers and Goals).
� Obtain Thai industry input, via a questionnaire survey, the data acquired to be used
for the statistical analyses and SD modelling.
A System Dynamics Approach to Construction Safety Culture
190
� Perform the EFA and SEM analyses to confirm the construct validity of the
proposed CSC model.
� Develop a CSC dynamic model utilizing SD software ‘STELLA’ to examine the
interactions and causal relationships among the five enablers and Goals, over a
period of time.
� Verify and validate the developed dynamic model.
� Assess CSC maturity levels using the scores from the developed CSC index.
� Identify areas for safety improvement to achieve a higher CSC index, and progress
through to higher maturity levels.
These objectives were successfully achieved, as presented in Chapters 1 to 7. In Chapter
1, a literature review of the characteristics of the construction industry, safety culture
definitions, and the measuring of safety culture was conducted. Two major
shortcomings seemed apparent, viz�
� No safety culture model included either the causal relationships among the key
factors of the CSC, or any feedback mechanism between these factors.
� No tool existed for appropriately assessing the CSC maturity levels or the evolution
to higher maturity levels, over time.
Consequently, these shortcomings became the research objectives, while a number of
research aims were identified to fulfil those research gaps.
Chapter 2 presented the research methodology adopted for this study, including the
research design and the research activities, along with the expected outputs. The
research activities and expected outcomes led to the specific steps required to fulfil the
research aims.
A System Dynamics Approach to Construction Safety Culture
191
As discussed in Chapter 2, a questionnaire survey was adopted as the most appropriate
method for data collection. The techniques for data screening and the preliminary
analyses were explained. In addition, the EFA, SEM, and SD modelling, as well as their
uses in this study were also introduced.
Three well-known performance measurement systems (the MBNQA framework, the
BSC framework, and the EFQM Excellence model), were critically examined and
compared in Chapter 3. Based on this comparison, the EFQM Excellence model was
identified as the basis for the CSC model development. The proposed CSC model
consisted of six constructs, i.e. five enablers, namely Leadership, Policy and Strategy,
People, Partnerships and Resources, and Processes, and the single set of Goals. Their
associated attributes were identified from the literature review, and were used in
developing the questionnaire survey. The survey was drawn up to define and then
operationalise the six constructs of the CSC model.
In Chapter 4, the questionnaire, which consisted of 34 statements covering 34 attributes
of the CSC, was drawn up. It was rated on a five-point Likert scale. The questionnaire
was sent to over 100 Thai construction-contracting organizations, with 53.6% being
returned. Three surveys were unusable due to data incompleteness, and so they were
dropped from the data file. One hundred and fifteen surveys were used in the analyses.
The preliminary analyses and data screening were performed to increase confidence in
the data. The results showed that less than 5% of the missing values were found in each
attribute (item), and that all the attributes displayed a normal distribution (their
skewness and kurtosis values were in acceptable ranges). Confidence in the data was
therefore increased. Only a single outlier was found (in data number ‘76’). As a result,
this case was removed from the data file, leading to a retained total of 114 data sets for
the analyses. The retained data were used to test the internal consistency of the 34
attributes within the six constructs of the CSC. The results demonstrated high reliability
values (Cronbach’s alpha higher than 0.7), and hence increased confidence in the
attributes to the measurement of their respective constructs.
A System Dynamics Approach to Construction Safety Culture
192
The screened data were analysed further, using principal axis factoring with varimax
rotation (see Chapter 5). A total of 27 attributes of the CSC enablers were identified,
categorised into a number of factors that represented the interrelations among the set of
attributes. Three attributes were removed from the data file, leaving 24 attributes
grouped into five enabler-constructs (labelled Leadership, Policy and Strategy, People,
Partnerships and Resources, and Processes). Among those groups, nine attributes were
relocated from one proposed enabler to another.
After this relocation, the SEM was undertaken to provide further evidence for the
construct validity of the CSC model. The 24 attributes, grouped to explain the five
enablers, and the seven attributes grouped to explain Goals, were examined, using the
confirmatory factor analysis (CFA) technique, to further specify the posited relations of
the attributes to their underlying constructs. A number of GOF indices, such as �2/DF,
RMSEA, CFI, and NFI, were used to assess the model fit. The proposed CSC model
was modified by eliminating the links with low correlations, and removing the attributes
with high multicollinearity. In doing so, seven attributes were removed, leading to a
best-fit measurement model that comprised 20 attributes grouped to represent the six
constructs of the CSC model.
The fitted measurement model was subsequently tested to examine the direction of the
assumed relationships between the six constructs, which were reflected by the arrows
connecting them. It was first assumed that bi-directional relationships existed between
the three enablers (Policy and Strategy, People, and Partnerships and Resources), due
to their potential to affect each other. The model was then analysed with different
directional influences among these three enablers. As a result, one attribute with high
multicollinearity was removed from the Leadership construct. The fitted structural
model confirmed the direction of relationships among the five enablers and Goals of the
CSC.
Leadership appears to directly influence the implementation of the People, Policy and
Strategy, and Partnerships and Resources enablers. However, Leadership has indirect
effect on Partnerships and Resources; through the implementation of the People
A System Dynamics Approach to Construction Safety Culture
193
enabler, the relationship was found to be stronger than the direct one. This result fits
well with the assumption that Thai managers consider teamwork more important, in
improving safety implementation, than the provision of safety resources (Aksorn and
Hadikusumo, 2006).
Partnerships and Resources, on the other hand, was found to indirectly affect Processes
through Policy and Strategy, which, likewise, appears to be indirectly influenced by
People. Both People and Policy and Strategy have significant relationships with
Processes, which appears to have a strong effect on the achievement of Goals. The
fitted structural model was labelled the final CSC model. The directions of the
relationships among its five enablers and Goals, as well as their correlation coefficients,
were then used in developing the CSC dynamic model.
In Chapter 6, the CSC dynamic model was formulated to capture the interactions and
causal relationships among the six constructs (five enablers and Goals) of the CSC
model, over a period of time. Model verification and validation were undertaken to
increase confidence in the developed model. The CSC index, developed through the
dynamic model, represented the sum of the five enablers and Goals’ values at a point in
time, and was used together with the five levels of CSC maturity to indicate the current
CSC maturity level.
Base run results revealed that an organization with ad-hoc safety implementation should
primarily focus on enhancing the Leadership and People enablers to successfully
progress through to higher CSC maturity levels in the future. This finding perfectly
matches with the respondents’ perspective that People and Leadership are among the
most influential enablers in significantly improving the CSC.
As presented in Chapter 7, policy analyses were performed with two organizations
(randomly chosen from the data file), currently in the second (managing) and third
(involving) levels of CSC maturity, respectively. A number of safety policies were
A System Dynamics Approach to Construction Safety Culture
194
tested to identify the most effective policy each organization could apply to enhance its
CSC, and progress through to the fifth (continually improving) CSC maturity level.
A cyclical style of safety management was also modelled to imitate real-life situations
where top management gradually withdrew its attention from safety when the CSC
index exceeds the upper limit of management satisfaction with safety performance. This
‘attention withdrawal’ negatively affects the Enablers and Goals scores, and ultimately
the CSC index. The CSC index decreases as top management withdraws its attention
from safety. This decrease continues until the index score reaches the lower limit,
meaning that the organization falls into the lower CSC maturity level (the fourth CSC
maturity level). as a consequence of this fall, top management takes prompt actions to
improve safety implementation, and thus to increase the Enablers and Goals scores, as
well as the CSC index. Once again, the CSC index rises until it exceeds the upper limit,
then management starts to withdraw its attention from safety again, and the cycle
continues.
While being affected by the cyclical nature of safety management, the organization,
however, slowly progresses towards the maximum CSC index score of 1,000 point, over
a very long period of time.
8.3 CONTRIBUTIONS TO THE EXISTING BODY OF KNOWLEDGE
Despite a large number of research studies focusing on measuring safety culture,
virtually no research has been undertaken: 1) to investigate the interactions and causal
relationships among the key factors of the CSC; and 2) to assess the CSC maturity level
and determine areas for improvement to progress through to higher maturity levels, over
a period of time.
The results from this study have contributed to the existing body of knowledge in the
following ways:
A System Dynamics Approach to Construction Safety Culture
195
� Previously, no study had comprehensively modelled the CSC, and most studies had
only tested the interactions among selected enablers, in isolation. Those studies were
unable to identify the causal mechanisms that researchers/managers need to
understand and so enhance safety culture. The CSC model, developed in this study,
explores the causal relationships among the five enablers and Goals, thus extending
knowledge and understanding of the key factors and their respective, as well as
collective, influences on CSC implementation and output.
� The research facilitated the examination of the relationships among the enablers in a
user-friendly graphical format. To the best of the author’s knowledge, this had not
previously been undertaken in the area of the CSC.
� The developed CSC model provides an integrated framework for understanding how
decisions and behaviours of leadership are linked to safety processes, and how these
translate into desired safety goals (better safety performance).
Such contributions provide a strong foundation for understanding the CSC and its key
factors, as well as the relationships among those key factors, thus adding value to future
research.
8.4 IMPLICATIONS FOR THE THAI CONSTRUCTION INDUSTRY
The development of a CSC dynamic model provides a number of benefits to the Thai
construction industry, as discussed below.
� Although extracting from international literature, most of the attributes (more than
82%), associated with each enabler and Goals, in the final CSC model give a good
representation of safety practices in the Thai construction industry. Such attributes
include training, safety standards, human resources, financial resources,
stakeholders’ cooperation, management commitment, workers involvement, workers
relationships, safety standards, safety resources, number of accidents, cost of
accidents, safety awareness, safety initiatives, safety integration in business goals,
A System Dynamics Approach to Construction Safety Culture
196
accountability, safety responsibilities, communication, and feedback (Boonrod et al.,
1998; Pipitsupaphol and Watanabe, 2000; Embassy of Denmark, Bangkok, 2006;
Aksorn and Hadikusumo, 2007; Wangniwetkul, 2007). The utilization of the study’s
results in realistically improving the CSC in Thailand can be said to be reinforced.
� By the literature review, Leadership was the most influential factor for a successful
safety program implementation in the Thai construction industry (Aksorn and
Hadikusumo, 2007). In an organization with no prior safety implementation,
leaders must be a role model in behaving safely. They must also be opened-mind,
and encourage two-way communication, and thus share the ideas of how to improve
safety.
� People and Policy and Strategy also play a key role in successful safety
implementation in the Thai construction industry. Workers involvement in safety
related activities, clear assigned safety responsibilities, and realistic safety rules help
enhance safety performance in an organization (Boonrod et al., 1998; Embassy of
Denmark, Bangkok, 2006).
� Further, the use of personal protective equipment (PPE) helps reduce site accidents
(Wangniwetkul, 2007). While adequately provided, the workers, sometimes, do not
use their PPE because it is inconvenient to do so (Pipitsupaphol and Watanabe,
2000). Consequently, the leaders are the key factor to solving this problem. They
must be role models, wearing the PPE everytime they enter the site. This action will
raise the workers’ perception and awareness of safety, and they will, then, follow the
leaders’ example (this is consistent with the direct relationships among the
Leadership, People, and Partnerships and Resources enablers found in this study).
� SD modelling was a useful tool in understanding the interactions and causal
relationships among the key factors, and has been similarly used in a number of
research projects focused on the Thai construction industry (Ogunlana et al., 1996;
Chritamara and Ogunlana, 2002; Ogunlana et al., 2002). The developed CSC
dynamic model provides insights into, and guidance and understanding of, the
interactions and casual relationships among the five enablers and Goals of the CSC,
over a period of time. In the Thai construction industry, these interactions and causal
relationships are important, as a change in one enabler may largely affect the change
in another enabler(s). This conclusion is supported by Aksorn and Hadikusumo
A System Dynamics Approach to Construction Safety Culture
197
(2004), who stated that a lack of management support and management pressure
was found to be associated with workers’ unsafe acts, which led to high accident
rates. An organization could improve its safety performance if top management
committed more to improving safety, for instance, promoting realistic and workable
safety rules, providing adequate safety resources, and encouraging safety feedback
(Aksorn and Hadikusumo, 2007).
� The developed CSC index will assist organizations in assessing their current CSC
maturity levels, and identifying areas for safety improvement to enable progress
through to higher maturity levels. Organizations with different maturity levels will
need different safety policy and safety implementation processes, which cannot be
imitated. The use of SD modelling, with the developed CSC index, will help
organizations to plan the most effective safety implementation process to achieve
their safety goals within a planned time frame.
� Making investment decisions to improve one or more enabler is one of the most
essential roles played by management. However, these decisions have different
levels of impact on safety goals. Thus, managers who evaluate different policies
must do so based on informed decisions. The developed CSC dynamic model has
the capability to facilitate to test alternative strategies, through a number of model
simulations, to improve safety culture, by do not actually have to implement them.
Nevertheless, this approach helps to save costs that may occur from not
implementing the best safety strategy.
8.5 LIMITATIONS AND RECOMMENDATIONS FOR FUTURE
RESEARCH
The limitations of this study are presented below:
� The data used were based on input provided only by medium-to-large construction
contracting companies operating in Thailand.
A System Dynamics Approach to Construction Safety Culture
198
� The targeted respondents were in senior positions, such as directors and project
managers, to facilitate the capture of a macro-level perspective for the CSC.
Workers were not included in the sample.
� The numbers of attributes used to operationalise each of the six constructs of the
CSC (Leadership, Policy and Strategy, People, Partnerships and Resources, and
Processes) were extracted from the international literature review, and were not
specifically limited to the Thai practices.
� The subcontractors’ role in improving the CSC was not of explicit interest (it is only
included in the ‘stakeholders’ cooperation’ attribute in the Partnerships and
Resources enabler).
� While the ‘workload’ and ‘work pressure’ attributes were key factors influencing
the construction site safety in Thailand (Boonrod et al., 1998; Pipitsupaphol and
Watanabe, 2000), they were not included in the final CSC model as a result of the
EFA and SEM.
� The final CSC model was developed based on the questionnaire surveys targeting
Thai construction organizations, thus, it might not be a best normative model to
prescribe the way of developing CSC in other countries.
� The study offers maximum flexibility to the users in utilizing their best judgement to
determine the relevant contributions each item makes to the operationalisation of a
certain enabler. In other words, the study allows the users to perform the analysis
using different levels of extra efforts per enabler (as a whole) without examining the
contributions made by each item. However, this could be a problem if the user does
not have enough experience or knowledge in planning for safety improvements.
The recommendations for areas of future research are listed below:
� The study was conducted using data from Thai construction organizations, in which
Thailand is considered a developing country. Thus, a comparative study may be
performed between developed and developing countries to investigate the
differences in CSC’s perspectives.
A System Dynamics Approach to Construction Safety Culture
199
� A comparison study between top management and workers’ perceptions of safety
could capture the macro-level, as well as micro-level perspectives, of CSC.
� Different score-ranges of the five levels of CSC maturity might be altered
(whenever appropriate) to investigate the results over a period of time.
� Case studies could be conducted, over a period of time, to examine the deviations
between the simulation and the real life results to help further refine the model.
8.6 CLOSURE
This study made fundamental contributions to the area of construction safety culture
(CSC). The developed CSC dynamic model provided an insight into the interactions and
influences each enabler has in improving the CSC. The developed CSC index will guide
organizations benchmarking their CSC maturity level, and planning safety
improvements.
A System Dynamics Approach to Construction Safety Culture
200
�
AAppppeennddiixx 11
QQuueessttiioonnnnaaiirree SSuurrvveeyy
A System Dynamics Approach to Construction Safety Culture
201
Centre for Infrastructure Engineering and Management Griffith University
Questionnaire Survey
Safety Culture in Construction
Organizations
Thanwadee Chinda
PMB 50 Gold Coast Mail Centre
Centre for Infrastructure Engineering and Management
Griffith University
Gold Coast Campus
Queensland 9726 Australia
A System Dynamics Approach to Construction Safety Culture
202
Centre for Infrastructure Engineering and Management
Griffith University, Gold Coast Campus
PMB 50 Gold Coast Mail Centre 9726, Queensland, Australia
February 28, 2008
Dear Sir/ Madam;
The following questionnaire has been developed to assess the organizational safety
culture in the context of construction. Please set aside 20 minutes for filling out this
questionnaire. The information will be used for academic purposes only, as integral part
of a PhD study. Individual responses will be kept confidential. Only a consolidated
summary of the result may be published.
The questionnaire comprises a series of statements relating to different aspects of
construction safety culture. Please indicate your level of agreement or disagreement
using a five-point Likert scale (for example 1 = strongly disagree; 5 = strongly agree).
Once you have completed the questionnaire, please return it in the envelope provided as
soon as possible. Should you have any questions, please do not hesitate to contact me at
+61-422062007 (Australia); 01-8802408 (Thailand) or e-mail me at:
[email protected] or write to me: Thanwadee Chinda 153/3 Moo 12 Khaowang-
Numpu Rd. Jadeehuk Muang Ratchaburi 70000 Thailand. Thank you in advance for
your cooperation.
Yours sincerely, Endorsement
Thanwadee Chinda Prof. Sherif Mohamed
(PhD Candidate) (Principal Supervisor)
A System Dynamics Approach to Construction Safety Culture
203
������������ ��������������� ����� ������������� ���
����������� ������ � ������!"� ������"� ��#��$�������� %�������"���$�
6 &�'���� � 2549
��$���(���$�����;
�))�)*��+&��$,����-�!.,� ���/+�/� ��%��������0� �����������%��'�� (safety culture) /���� �1
�� ���� �� (����� �&2�������� 20 ���$���/+�/� �� � �))�)*���$,!�����$3����,������*� 1
�-�3%/+����������� ���. 4���(���,� ��1 �%5��(����.�!� ���-����������)� �"��!������� ����$
����� /����!� “A system dynamics approach to safety culture” !������,������*� � 6)3���(���$������$1
���7����(�8���7�����$3���� ��%��������(���,�
�))�)*��+&��$,�)(��%5� 3 �(��
�(���$ 1 �%5��-�*�����3%� $�� �)�(�������� ��$�(���-������(
�(���$ 2 %�� )3%����!�������,���� 34 !������ $�� �)��0� �����������%��'��/���� ��� �1
��� �� (����� �&2��� ������)������6��������3�(��6�����!��(��/��"(��!����� (�������) 1
��(� �)3�(��6������� �$�&� ����) 3 ��(� �)3�(�+����6��������3�(��6����� ����) 5 ��(� �)��6������� �$�&�)
�(���$ 3 �%5� ����&%7�������6�!��(�� ��� ��/��!�������
����(��")�))�)*���$,���6����� �&2��(� ��)���/�(/�9�������"����"�%:�$��)������1������1
7������-������*3%��)�))�)*��3��*��"�� �� *���(���$!������%�� ��/� �&2�"��"(�$
+61-422062007 (Australia)
+66-81-8802408 (Thailand) ���
E-mail ���$ [email protected]
!!)�&2�(������/� ��/�������(����
�������������1 1 1 1 1 1 1 ����)��1
1
����$1�����1 1 1 1 1 1 1 ;<=>?1@AB<C>1D=AEFBG1
H�� �. 4�I1 1 1 1 1 1 1 H�������$%�. 4�I1
�
A System Dynamics Approach to Construction Safety Culture
204
Part I: Personal Information (����������� �����������������������������������������)
Please fill in the blanks or tick the appropriate box. ( �&2�")�-�*����/�+(��(��1��� �+(�
�$�(��"�� ����� )
�
1. The name of your organization (optional) ___________________________________
+���� �!��(��1H��� ")���3�(")3��I11
2. Your job title: ______________________________________
"-����(�!��(��11
3. How long have you been working in this organization? __________ year/s
�(���-����/���� ��$,�����1 1111 1 1 1 1 11111111%J1
�
4. How long have you been working in the construction industry? __________ year/s
�(���-����/�����&"��� ����� ���� �� (����������11 1111111111 1 11111111111%J1
5. Does your organization have a formal safety policy (such as safety manuals)?
��� �!��(���$���)��� $�� �)����%��'���(���%5���� �����3�(1 1
� Yes � No
���������$1 1 1111 1 1 11113�(�$1
6. Does your current role have direct safety responsibilities?
)�)��!��(��/���� �%K��&)���$�(��� $��!�� �)������)7��+)��������%��'�����3�(1 1
� Yes � No
11111�$1 1 1111 1 1 11113�(�$1
If yes, is it in:
*���$1�%5�������)7��+)����1
� Planning and/or reviewing safety policy� � � Safety auditing
����7�1���L���1�)������)������%��'��11 1111"����)��������%��'��1
� Supervising safety in site works � � � � Other (please specify) �
�%5��$%�. 4���������%��'��/�1site ��� 1 1 1111��M1H�%����)&I1
1
�
A System Dynamics Approach to Construction Safety Culture
205
7. How often do you engage in safety related activities (such as safety training, safety
auditing, safety meeting, etc.)?
�(���$�(��� $��!������$�(���(��/� �� �����������%��'��1 H�+(�1 ��NO N����� ��%P�)�"�������1
����%��'��Q1 ��"����)��������%��'��Q1 ���!���(��%��+&���������%��'��I1 �� ���
��$��/�1 1 1
� Once a month � Once every 6 months �
�& ����1 111 11111 1 1111�& 1R1����111111 1 1111 1
� Once a year � Other (please specify) _______________
�& %J1 1 1 1 1 1111��M1H�%����)&I1
8. When was the last time you have engaged in safety related activities (such as safety
training, safety auditing, safety meeting, etc.)?
�(���$�(��� $��!������$�(���(��/� �� �����������%��'��1 H�+(� ��NO N����� ��%P�)�"�������1
����%��'��Q1 ��"����)��������%��'��Q1 ���!���(��%��+&���������%��'��I1 ���,��&�����1
���3�1
� Last month � Last 6 months
�����$����1 1 1 1 1111R1�����$����1
� Last year � Other (please specify) _______________
%J�$����1 1 1 1 1111��M1H�%����)&I1
9. In your opinion, how would you rank your organization’ s safety performance
compared to the national’ s average record?
/����������6�!��(��1 �(����/��� 2S�����%����� �'����������%��'��/���� �!��(�����1
�%�$�)��$�) �)� 2S��8�$�/�%������(��3�1 1
� Above � Same � Below
�$ �(�� 2S��8�$�1 11111 1111��(� �)� 2S��8�$�1 1 1111"-� �(�� 2S��8�$�1
A System Dynamics Approach to Construction Safety Culture
206
Part II: Safety Culture (����������������� ������ !�� "��)
This part contains 34 statements relating to construction safety culture. Please complete
this part by circling the score that best reflects the level of your agreement or
disagreement with each statement.
(�(���$,%�� )����!�������,����1 TU1 !������ $�� �)��0� �����������%��'��/���� ��� ����1
�� (�����1 �&2��� ������)������6��������3�(��6�����!��(��/��"(��!�����)
The meaning of each score is shown below. (��������!��"(������)������%5�����$,)
1 2 3 4 5 N/A
Strongly Disagree
3�(��6������� �$�&�1
Disagree
3�(��6�����1
Neither Agree nor Disagree
3�(�+����6��������3�(��6�����1
Agree
��6�����1
Strongly Agree
��6������� �$�&�1
!3�(")1
No. Statement (����� ) Score
(�� �������)
Leadership (��� �!#�$�����)�
1. In our organization, management takes safety seriously NV��)�����/���� �!����/�������-���W� $�� �)����%��'���(���� 1
1 2 3 4 5 N/A
2.
Management encourages workers to give opinions and/or suggestions on safety matters NV��)�����+� +��/����� ���H�����,������I ���������6����L���!�1��������������%��'���
1 2 3 4 5 N/A
3. In our organization, management makes sure that workers hold their responsibilities for their own safety NV��)�����/���� �!�����,-����/����(/��(��������)7��+)/������$����1����%��'��!�"�1
1 2 3 4 5 N/A
4.
Management acts quickly to correct safety problems when brought to his/her attention NV��)������ �3!%KW����������%��'������(������$%KW��� ��!.,�1
1 2 3 4 5 N/A
Policy and Strategy (�%��������&��'��(��)
5. It is our policy to recognize workers with good safe behaviours ����$���)��/� ��+��+�������$�$����%���X"���������%��'���$��$��1
1 2 3 4 5 N/A
6. It is our policy to give safety the same priority as production ����$���)��/� ��/�������-���W��������%��'����(� �)���� ��7��"1
1 2 3 4 5 N/A
7. Our organization has a safety policy that gets reviewed and upgraded regularly ��� �!�����$���)����������%��'���$3����) ���)���%��)%�&�/���$!.,�1���1
1 2 3 4 5 N/A
8. In our organization, safety initiatives are proactively planned in order to continually improve our safety standards /���� �!�������������������������%��'��3����) ������7��%5��(���$1���%����+��/� ����0����"�#����������%��'��1
1 2 3 4 5 N/A
9. In our organization, safety is an integral part in formulating our business decisions and goals /���� �!��������%��'��*� ����%5��(����.�!��7�����%Y���������1 &� ��1
1 2 3 4 5 N/A
A System Dynamics Approach to Construction Safety Culture
207
No. Statement (����� ) Score (�� �������)
People ()������)
10. Our project staff (including workers) believe that our organization is genuinely concerned about workplace safety ��� ������� ��H�����,������I�+��(���� �!�����-��.�*.�����%��'��1/��$�-����
1 2 3 4 5 N/A
11. Our project staff (including workers) fully understand their safety responsibilities ��� ������� ��H�����,������I!�����$�����!��/�/������$������)7��+)��������%��'��!�"�1
1 2 3 4 5 N/A
12. In our organization, workers can seek advice on safety matters from their immediate boss, such as project manager, safety manager, supervisor etc. /���� �!��������������*!�-�����-���������%��'��3���� �������!�"�1 �+(�1�� 7��)��������� ��Q17��)�������������%��'��Q1�$%�. 4�1���1��M1
1 2 3 4 5 N/A
13. In our organization, project staff (including workers) are involved, formally and/or informally, in safety related issues /���� �!����1 ��� ������� ��H�����,������I�$�(���(��1 ��,��(���%5�1��� �����L���3�(�%5���� ��1/����!�� $�� �)����%��'��1
1 2 3 4 5 N/A
14. In our organization, workmates often give suggestions to each other on how to work safely /���� �!���������(������� /���-�����-�/� ���-�����(��%��'��� (1 ����� ���
1 2 3 4 5 N/A
15. In our organization, workload is reasonably balanced among workers so that they can get the job done safely /���� �!����1 %����2����$�������&��� �)�-�������������(��������13���-�����(��%��'��1
1 2 3 4 5 N/A
16. Our organization ensures that workers are not under pressure to avoid unsafe behaviours ��� �!�����,-����/����(/��(������3�(3���-����'��/"����� ����1���1��$ ��$���X"� ����$��$��"(����%��'��1
1 2 3 4 5 N/A
Partnerships and Resources (*&������+�$�� �������� ��������)����)
17.
Project participants, such as subcontractors, cooperate with us in following our safety standards 7���$�(���(��/����� ���+(�7����)��������(�1 /�������(����/� ��%P�)�"�"��1��"�#����������%��'��!���� �!����1
1 2 3 4 5 N/A
18. In our organization, financial resources are adequately provided to support the implementation of our safety policy ��� �!����/�� �����)��&����� �������(����$���1 ����-�3%/+�/� ��1�-����� ��"�����)����������%��'��!���� �1
1 2 3 4 5 N/A
19. Our organization has sufficient necessary safety resources available so that workers can carry out their jobs safely ��� �!�����$&% �2����������������������%��'���$�-���W�(��1���$��1�������/+����1���/��������-����3���(��%��'��
1 2 3 4 5 N/A
20. Our organization endeavours to have adequate human resources to get the job done safely ��� �!�����$������ ���&4���(����$����$���-�/�����"(��M�-���6�3���(��1%��'��1
1 2 3 4 5 N/A
A System Dynamics Approach to Construction Safety Culture
208
No. Statement (����� ) Score ,�� �������-
Processes (���������)
21. In our organization, we provide adequate training for those performing new tasks safely ����$���/��(�$3�(����-��� (�1 ��� �!������/�� �����)��&� ��NO N�1���� ��%P�)�"�����������%��'���(�����$��1 ���/�� ���-�����%5�3%1�(��%��'��1
1 2 3 4 5 N/A
22. In our organization, risk and hazard assessment is a part of our routine safety planned activities /���� �!����1 ��%������7�����������$�������"����$��� ��!.,��� 1 ���-����1�%5��(����.�!� �� �����������%��'��%���-����1
1 2 3 4 5 N/A
23. Feedback on safety implementation is encouraged within the organization in order to improve safety performance ��� �!�������)��&�/���$ ��+$,���!��$���L���!�) ��(�!� ���-�����1�����������%��'��1 ���7�/� ���-�3%%��)%�&������0��%����� �'������1����%��'���
1 2 3 4 5 N/A
24. Our organization adopts a no-blame approach so that workers always report near misses and accidents they experience or witness ��� �!�����-���)) ��3�("��"$��1 3�( �(����41 ���3�(����4��/+�1 �����,�1�����!�����.��"6�/�������&)�"���"&�$"���%���)����)��6����3�(%��%K�1
1 2 3 4 5 N/A
25. In our organization, site housekeeping is maintained at a high level /���� �!����1 �������� 4�1 ZC[B1 ���/���$�����������%��'����(/�1����)���1
1 2 3 4 5 N/A
26. Our organization keeps accidents records to investigate their causes ��� �!����� 6)����������&)�"���"&����-�3%/+�/� ����)������"&!�1&)�"���"&��,�M
1 2 3 4 5 N/A
27. Our organization has a good safety benchmarking system to compare with other construction organizations ��� �!�����$��)) ���%�$�)��$�)%������7�H\B]̂AFE<_C]̀I��������1%��'���$�$1 ����-�3%/+�/� ���%�$�)��$�)%������7� �)��� ��� ����1 �� (�������M�
1 2 3 4 5 N/A
Goals ($���� ����.��������� *���)
28. Workers are generally satisfied with the way we currently manage safety in our organization ������$�����.��/�/� ��)����������������%��'��!���� �/�1%K��&)��1
1 2 3 4 5 N/A
29. The way we currently manage safety in our organization promotes safe work behaviour ��)����������������%��'��!���� �/�%K��&)���$�(��+(��/� ��1���)��&�/��� ���X"� ���/� ���-�����(��%��'��1
1 2 3 4 5 N/A
30. The way we currently manage safety in our organization helps us reduce the number of severe accidents and safety related incidents ��)����������������%��'��!���� �/�%K��&)��1 �$�(��+(��/� ����1�-����&)�"���"&�$� ��!.,�/���� �1
1 2 3 4 5 N/A
31. The way we currently manage safety in our organization helps us meet our clients’ expectations ��)����������������%��'��!���� �/�%K��&)��1 �$�(��+(��/� ��1 )���&����"�� ����������������!��� ���1
1 2 3 4 5 N/A
32. Public perceive our organization with a good safety image �� ��2+������ �!�����(��$'��������������%��'���$�$1
1 2 3 4 5 N/A
33. The way we currently manage safety in our organization has a positive influence on workers’ morale ��)����������������%��'��!���� �/�%K��&)��1 �$7�����)� "(1�$� ��������!������11
1 2 3 4 5 N/A
A System Dynamics Approach to Construction Safety Culture
209
No. Statement (����� ) Score (�� �������)
Goals (Cont.) ($���� ����.��������� *����(��)
34. The way we currently manage safety in our organization leads to reduction in the total costs associated with accidents ��)����������������%��'��!���� �/�%K��&)��1 �-�3%��( ������!�1�(�/+��(����,�����$� $��!�� �)&)�"���"&1 �+(�1 �(���$�����Q1 �(��� 4����)��1�����M1
1 2 3 4 5 N/A
Your opinions (��� �. �*/�������)
1. Considering the five enablers listed above (Leadership, Policy and Strategy, People, Partnerships and Resources and Processes), which enabler do you think the Thai construction industry considers being the most influential in significantly improving safety culture? (Please tick only one box) ��������2�1a1���!���� !���)�1H9.�%�� )����1�����%5�7���-�Q1���)������&� ���"��Q1��� ���Q1�&���(��17���$�(���(��1���1������ �Q1��� ��)�� ��I1�(������(����!�3���$�&(��*��� %��������������������0��� �-��.��(��$�� ����� �$�&�/� ��%��)%�&������0����0� �����������1%��'��1H �&2���� ���$��1b1+(�I11
� Leadership � Policy and Strategy 1111�����%5�7���-�1 1 1 1 1111���)������&� ���"��1� People � Partnerships and Resources
1 1111��11111 1 1 111111111111111�&���(��17���$�(���(��1���1������ �1 � Processes ��)�� ��1 2. Considering the five enablers listed above (Leadership, Policy and Strategy, People,
Partnerships and Resources and Processes), which enabler do you think your organization considers being the most influential in significantly improving safety culture? (Please tick only one box) ��������2�1a1���!���� !���)�1H9.�%�� )����1�����%5�7���-�Q1���)������&� ���"��Q1��� ���Q1�&���(��17���$�(���(��1���1������ �Q1��� ��)�� ��I1�(������(����!�3���$�������������-��.��(��$1 �� ����� �$�&�1 /� ��%��)%�&������0����0� �����������%��'��1 H �&2���� ���$��1 b1+(�I1
� Leadership � Policy and Strategy 1111�����%5�7���-�1 1 1 1 11111���)������&� ���"��1� People � Partnerships and Resources
1 1111��11111 1 1 1 1 11111�&���(��17���$�(���(��1���1������ �1 � Processes ��)�� ��1
A System Dynamics Approach to Construction Safety Culture
210
3. How will you rank the five enablers listed above (Leadership, Policy and Strategy, People, Partnerships and Resources and Processes) according to its importance in enabling a more positive safety culture? �(������$���-���)1 a1 ���!���� !���)�1 H9.�%�� )����1 �����%5�7���-�Q1 ���)������&� ���"��Q1��� ���Q1�&���(��17���$�(���(��1��������� �Q1��� ��)�� ��I1�(��3�1���/+�� 2S������-���W/�1 ���-�/��� ����0� �����������%��'������)� �� !.,�1
___ Leadership ___ Policy and Strategy 11111111�����%5�7���-�1 1 1 11111111���)������&� ���"��1___ People ___ Partnerships and Resources
1 1111111��11111 1 1 1 11111111�&���(��17���$�(���(��1���1������ �1 ___ Processes 11 1111111 ��)�� ��1
4. Do you have other suggestions to improve the safety culture on construction sites? �(���$!�������/� ��%��)%�&������0����0� �����������%��'��1 /�1 ZC[B1 ���!� ��1 �� ���� �� (��������3�(1
� Yes � No 11111�$1 1 1111 1 1 11113�(�$1
If yes, please write them down in the space provided: *���$1 �&2��!$��!�������!��(�������(���$,1
_____________________________________________________________________________________ __________________________________________________________________________________________________________________________________________________________________________ __________________________________________________________________________________________________________________________________________________________________________
Thank you very much for your time and effort ����&1����2���������������2*���� ��� 3��
A System Dynamics Approach to Construction Safety Culture
211
AAppppeennddiixx 22
RRaaww DDaattaa
A S
yste
m D
ynam
ics
App
roac
h to
Con
stru
ctio
n Sa
fety
Cul
ture
21
2
case
cm
mt
Com
m.
accn
ld
bx
awrn
al
gn
stnd
in
it in
tg
prcp
re
sp
sppt
in
vm
rlsp
w
kld
prsr
1
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
2 ��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
3 ��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
4 ��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
5 ��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
6 ��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
7 ��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
8 ��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
9 ��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
10
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
11
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
12
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
13
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
14
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
15
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
16
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
17
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
18
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
19
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
20
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
21
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
22
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
23
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
24
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
25
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
26
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
27
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
28
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
29
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
30
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
Not
e: c
mm
t = th
e ‘c
omm
itmen
t’ i
tem
, com
m =
the
‘com
mun
icat
ion’
item
, acc
n =
the
‘acc
ount
abili
ty’
item
, ldb
x =
the
‘lea
ding
by
exam
ple’
item
, aw
rn =
the
‘saf
ety
awar
enes
s’ it
em,
algn
= th
e ‘s
afet
y an
d pr
oduc
tivity
alig
nmen
t’ it
em, s
tnd
= th
e ‘s
afet
y st
anda
rds’
item
, ini
t = th
e ‘s
afet
y in
itiat
ives
’ ite
m, i
ntg
= th
e ‘s
afet
y in
tegr
atio
n in
bus
ines
s go
als’
item
, prc
p =
the
‘sha
red
perc
eptio
ns’
item
, res
p =
the
‘saf
ety
resp
onsi
bilit
ies’
item
, spp
t = th
e ‘s
uppo
rtiv
e en
viro
nmen
t’ it
em, i
nvm
= th
e ‘w
orke
rs’
invo
lvem
ent’
item
, rls
p =
the
‘wor
kers
’ re
latio
nshi
ps’
item
, wkl
d =
the
‘wor
kloa
d’ it
em, p
rsr =
the
‘wor
k pr
essu
re’ i
tem
A System Dynamics Approach to Construction Safety Culture
213
case cmmt Comm. accn ldbx awrn algn stnd init intg prcp resp sppt invm rlsp wkld prsr 31 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
32 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
33 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
34 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
35 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
36 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
37 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
38 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
39 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
40 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
41 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
42 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
43 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
44 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
45 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
46 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
47 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
48 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
49 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
50 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
51 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
52 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
53 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
54 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
55 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
56 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
57 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
58 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
59 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
60 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
Note: cmmt = the ‘commitment’ item, comm = the ‘communication’ item, accn = the ‘accountability’ item, ldbx = the ‘leading by example’ item, awrn = the ‘safety awareness’ item, algn = the ‘safety and productivity alignment’ item, stnd = the ‘safety standards’ item, init = the ‘safety initiatives’ item, intg = the ‘safety integration in business goals’ item, prcp = the ‘shared perceptions’ item, resp = the ‘safety responsibilities’ item, sppt = the ‘supportive environment’ item, invm = the ‘workers’ involvement’ item, rlsp = the ‘workers’ relationships’ item, wkld = the ‘workload’ item, prsr = the ‘work pressure’ item
A System Dynamics Approach to Construction Safety Culture
214
case cmmt Comm. accn ldbx awrn algn stnd init intg prcp resp sppt invm rlsp wkld prsr 61 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
62 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
63 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
64 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
65 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
66 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
67 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
68 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
69 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
70 �� �� �� �� � �� �� �� �� �� �� �� �� �� �� ��
71 �� �� �� �� �� �� � � �� �� �� �� �� �� �� ��
72 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
73 � � �� �� �� �� �� �� �� �� �� �� �� �� �� ��
74 �� �� �� �� � �� �� �� �� �� �� �� �� �� �� ��
75 �� �� �� �� �� �� �� �� � �� �� �� � �� �� ��
76 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
77 �� �� �� �� �� �� �� �� �� �� �� �� �� �� � ��
78 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
79 �� �� �� �� �� �� � �� �� �� �� �� �� �� �� ��
80 �� �� �� �� �� � �� �� �� �� �� � �� �� �� ��
81 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
82 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
83 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
84 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
85 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
86 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
87 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
88 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
89 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
90 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
Note: cmmt = the ‘commitment’ item, comm = the ‘communication’ item, accn = the ‘accountability’ item, ldbx = the ‘leading by example’ item, awrn = the ‘safety awareness’ item, algn = the ‘safety and productivity alignment’ item, stnd = the ‘safety standards’ item, init = the ‘safety initiatives’ item, intg = the ‘safety integration in business goals’ item, prcp = the ‘shared perceptions’ item, resp = the ‘safety responsibilities’ item, sppt = the ‘supportive environment’ item, invm = the ‘workers’ involvement’ item, rlsp = the ‘workers’ relationships’ item, wkld = the ‘workload’ item, prsr = the ‘work pressure’ item
A System Dynamics Approach to Construction Safety Culture
215
case cmmt Comm. accn ldbx awrn algn stnd init intg prcp resp sppt invm rlsp wkld prsr 91 �� �� �� �� �� �� � �� �� �� �� �� � �� �� ��
92 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
93 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
94 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
95 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
96 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
97 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
98 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
99 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
100 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
101 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
102 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
103 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
104 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
105 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
106 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
107 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
108 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
109 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
110 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
111 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
112 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
113 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
114 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
115 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
Note: cmmt = the ‘commitment’ item, comm = the ‘communication’ item, accn = the ‘accountability’ item, ldbx = the ‘leading by example’ item, awrn = the ‘safety awareness’ item, algn = the ‘safety and productivity alignment’ item, stnd = the ‘safety standards’ item, init = the ‘safety initiatives’ item, intg = the ‘safety integration in business goals’ item, prcp = the ‘shared perceptions’ item, resp = the ‘safety responsibilities’ item, sppt = the ‘supportive environment’ item, invm = the ‘workers’ involvement’ item, rlsp = the ‘workers’ relationships’ item, wkld = the ‘workload’ item, prsr = the ‘work pressure’ item
A System Dynamics Approach to Construction Safety Culture
216
case coop finc Resc hmnr trng risk fdbk nobm hskp docu bnmk jstf swbh acci cstm imge mrle cost 1 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
2 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
3 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
4 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
5 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
6 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
7 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
8 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
9 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
10 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
11 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
12 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
13 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
14 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
15 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
16 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
17 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
18 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
19 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
20 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
21 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
22 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
23 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
24 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
25 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
26 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
27 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
28 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
29 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
30 �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� �� ��
Note: coop = the ‘stakeholders’ cooperation’ item, finc = the ‘financial resources’ item, resc = the ‘safety resources’ item, hmnr = the ‘human resources’ item, trng = the ‘training’ item, risk = the ‘risk assessment’ item, fdbk = the ‘feedback’ item, nobm = the ‘no-blame approach’ item, hskp = the ‘housekeeping’ item, docu = the ‘safety documentation’ item, bnmk = the ‘benchmarking system’ item, jstf = the ‘job satisfaction’ item, swbh = the ‘safe work behaviour’ item, acci = the ‘number of accidents’ item, cstm = the ‘customers’ expectations’ item, imge = the ‘industrial image’ item, mrle = the ‘workforce morale’ item, cost = the ‘cost of accidents’ item
A S
yste
m D
ynam
ics
App
roac
h to
Con
stru
ctio
n Sa
fety
Cul
ture
21
7
case
co
op
finc
Res
c hm
nr
trng
ri
sk
fdbk
no
bm
hskp
do
cu
bnm
k js
tf
swbh
ac
ci
cstm
im
ge
mrl
e co
st
31
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
32
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
33
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
34
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
35
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
36
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
37
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
38
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
39
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
40
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
41
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
42
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
43
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
44
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
45
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
46
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
47
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
48
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
49
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
50
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
51
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
52
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
53
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
54
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
55
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
56
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
57
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
58
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
59
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
60
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
Not
e: c
oop
= th
e ‘s
take
hold
ers’
coo
pera
tion’
item
, fin
c =
the
‘fin
anci
al r
esou
rces
’ ite
m, r
esc
= th
e ‘s
afet
y re
sour
ces’
item
, hm
nr =
the
‘hum
an r
esou
rces
’ ite
m, t
rng
= th
e ‘t
rain
ing’
item
, ri
sk =
the
‘ris
k as
sess
men
t’ it
em, f
dbk
= th
e ‘f
eedb
ack’
item
, nob
m =
the
‘no-
blam
e ap
proa
ch’ i
tem
, hsk
p =
the
‘hou
seke
epin
g’ it
em, d
ocu
= th
e ‘s
afet
y do
cum
enta
tion’
item
, bnm
k =
the
‘ben
chm
arki
ng s
yste
m’
item
, jst
f =
the
‘job
sat
isfa
ctio
n’ it
em, s
wbh
= th
e ‘s
afe
wor
k be
havi
our’
item
, acc
i = th
e ‘n
umbe
r of
acc
iden
ts’
item
, cst
m =
the
‘cus
tom
ers’
exp
ecta
tions
’ ite
m,
imge
= th
e ‘i
ndus
tria
l im
age’
item
, mrl
e =
the
‘wor
kfor
ce m
oral
e’ it
em, c
ost =
the
‘cos
t of a
ccid
ents
’ ite
m
A S
yste
m D
ynam
ics
App
roac
h to
Con
stru
ctio
n Sa
fety
Cul
ture
21
8
case
co
op
finc
Res
c hm
nr
trng
ri
sk
fdbk
no
bm
hskp
do
cu
bnm
k js
tf
swbh
ac
ci
cstm
im
ge
mrl
e co
st
61
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
62
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
63
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
64
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
65
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
66
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
67
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
68
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
69
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
70
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
71
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
72
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
73
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
74
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
75
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
76
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
77
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
78
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
79
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
80
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
81
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
82
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
83
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
84
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
85
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
86
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
87
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
88
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
89
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
90
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
Not
e: c
oop
= th
e ‘s
take
hold
ers’
coo
pera
tion’
item
, fin
c =
the
‘fin
anci
al r
esou
rces
’ ite
m, r
esc
= th
e ‘s
afet
y re
sour
ces’
item
, hm
nr =
the
‘hum
an r
esou
rces
’ ite
m, t
rng
= th
e ‘t
rain
ing’
item
, ri
sk =
the
‘ris
k as
sess
men
t’ it
em, f
dbk
= th
e ‘f
eedb
ack’
item
, nob
m =
the
‘no-
blam
e ap
proa
ch’ i
tem
, hsk
p =
the
‘hou
seke
epin
g’ it
em, d
ocu
= th
e ‘s
afet
y do
cum
enta
tion’
item
, bnm
k =
the
‘ben
chm
arki
ng s
yste
m’
item
, jst
f =
the
‘job
sat
isfa
ctio
n’ it
em, s
wbh
= th
e ‘s
afe
wor
k be
havi
our’
item
, acc
i = th
e ‘n
umbe
r of
acc
iden
ts’
item
, cst
m =
the
‘cus
tom
ers’
exp
ecta
tions
’ ite
m,
imge
= th
e ‘i
ndus
tria
l im
age’
item
, mrl
e =
the
‘wor
kfor
ce m
oral
e’ it
em, c
ost =
the
‘cos
t of a
ccid
ents
’ ite
m
A S
yste
m D
ynam
ics
App
roac
h to
Con
stru
ctio
n Sa
fety
Cul
ture
21
9
case
co
op
finc
Res
c hm
nr
trng
ri
sk
fdbk
no
bm
hskp
do
cu
bnm
k js
tf
swbh
ac
ci
cstm
im
ge
mrl
e co
st
91
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
92
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
93
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
94
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
95
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
96
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
97
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
98
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
99
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
100
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
101
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
102
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
103
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
104
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
105
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
106
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
107
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
108
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
109
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
110
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
111
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
112
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
113
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
114
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
115
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
Not
e: c
oop
= th
e ‘s
take
hold
ers’
coo
pera
tion’
item
, fin
c =
the
‘fin
anci
al r
esou
rces
’ ite
m, r
esc
= th
e ‘s
afet
y re
sour
ces’
item
, hm
nr =
the
‘hum
an r
esou
rces
’ ite
m, t
rng
= th
e ‘t
rain
ing’
item
, ri
sk =
the
‘ris
k as
sess
men
t’ it
em, f
dbk
= th
e ‘f
eedb
ack’
item
, nob
m =
the
‘no-
blam
e ap
proa
ch’ i
tem
, hsk
p =
the
‘hou
seke
epin
g’ it
em, d
ocu
= th
e ‘s
afet
y do
cum
enta
tion’
item
, bnm
k =
the
‘ben
chm
arki
ng s
yste
m’
item
, jst
f =
the
‘job
sat
isfa
ctio
n’ it
em, s
wbh
= th
e ‘s
afe
wor
k be
havi
our’
item
, acc
i = th
e ‘n
umbe
r of
acc
iden
ts’
item
, cst
m =
the
‘cus
tom
ers’
exp
ecta
tions
’ ite
m,
imge
= th
e ‘i
ndus
tria
l im
age’
item
, mrl
e =
the
‘wor
kfor
ce m
oral
e’ it
em, c
ost =
the
‘cos
t of a
ccid
ents
’ ite
m
A System Dynamics Approach to Construction Safety Culture
220
AAppppeennddiixx 33
SSttaannddaarrddiizzeedd SSccoorreess ((ZZ--SSccoorreess))
A S
yste
m D
ynam
ics
App
roac
h to
Con
stru
ctio
n Sa
fety
Cul
ture
22
1
zc
mm
t Z
com
m
zacc
n zl
dbx
zaw
rn
zalg
n zs
tnd
zini
t zi
ntg
zprc
p zr
esp
zspp
t zi
nvm
zr
lsp
zwkl
d zp
rsr
1 1.
0 1.
4 1.
2 0.
8 0.
4 0.
2 0.
2 0.
2 0.
2 1.
4 0.
4 0.
0 0.
3 -0
.8
0.2
0.3
2 1.
0 0.
4 1.
2 0.
8 0.
4 0.
2 1.
3 0.
2 0.
2 0.
1 -1
.0
0.0
0.3
-0.8
0.
2 0.
3 3
1.0
1.4
1.2
0.8
1.3
1.2
1.3
1.3
1.2
1.4
0.4
0.0
1.6
0.4
1.4
1.3
4 -0
.2
0.4
1.2
-0.3
0.
4 0.
2 1.
3 0.
2 0.
2 0.
1 0.
4 -1
.3
-0.9
-0
.8
0.2
-0.8
5
1.0
0.4
0.0
0.8
0.4
0.2
0.2
1.3
1.2
0.1
0.4
1.3
0.3
1.6
1.4
0.3
6 -0
.2
0.4
0.0
-0.3
-0
.6
-0.8
0.
2 -0
.9
0.2
0.1
0.4
-1.3
0.
3 0.
4 -1
.0
0.3
7 -0
.2
0.4
0.0
0.8
-0.6
-0
.8
0.2
0.2
-0.8
0.
1 -1
.0
0.0
0.3
0.4
0.2
0.3
8 1.
0 1.
4 1.
2 0.
8 1.
3 1.
2 1.
3 1.
3 1.
2 1.
4 0.
4 0.
0 0.
3 0.
4 0.
2 0.
3 9
1.0
1.4
1.2
0.8
1.3
1.2
1.3
1.3
1.2
1.4
1.8
1.3
1.6
1.6
1.4
1.3
10
1.0
-0.7
1.
2 -0
.3
-0.6
-0
.8
-3.1
-3
.2
-0.8
-1
.2
-2.4
1.
3 -0
.9
-0.8
-2
.1
-1.9
11
-0
.2
0.4
-1.2
-0
.3
0.4
1.2
-0.9
0.
2 1.
2 -1
.2
0.4
0.0
1.6
0.4
-1.0
-0
.8
12
-0.2
0.
4 0.
0 -0
.3
0.4
-0.8
-0
.9
0.2
0.2
-1.2
-1
.0
0.0
0.3
0.4
-1.0
-0
.8
13
-0.2
0.
4 0.
0 0.
8 1.
3 0.
2 0.
2 0.
2 0.
2 0.
1 0.
4 1.
3 0.
3 1.
6 0.
2 1.
3 14
-0
.2
0.4
0.0
-0.3
-0
.6
0.2
1.3
0.2
0.2
0.1
-1.0
-1
.3
-0.9
0.
4 0.
2 1.
3 15
1.
0 0.
4 0.
0 -0
.3
-0.6
0.
2 1.
3 0.
2 0.
2 -1
.2
-1.0
-1
.3
0.3
0.4
0.2
1.3
16
-0.2
0.
4 0.
0 -0
.3
0.4
-0.8
0.
2 0.
2 0.
2 0.
1 -1
.0
-1.3
-0
.9
0.4
0.2
1.3
17
-1.5
-0
.7
0.0
0.8
0.4
0.2
0.2
0.2
0.2
0.1
-1.0
0.
0 0.
3 0.
4 0.
2 -0
.8
18
1.0
0.4
1.2
0.8
-0.6
0.
2 0.
2 0.
2 1.
2 0.
1 0.
4 -1
.3
0.3
0.4
0.2
0.3
19
-0.2
-0
.7
0.0
-0.3
0.
4 0.
2 0.
2 0.
2 0.
2 1.
4 1.
8 0.
0 0.
3 -0
.8
0.2
0.3
20
-0.2
0.
4 1.
2 -0
.3
0.4
0.2
0.2
0.2
-0.8
0.
1 0.
4 0.
0 0.
3 0.
4 1.
4 1.
3 21
-0
.2
-0.7
0.
0 0.
8 -1
.5
-0.8
0.
2 0.
2 0.
2 -1
.2
-1.0
0.
0 0.
3 -0
.8
-1.0
0.
3 22
-0
.2
-0.7
-1
.2
-1.4
-1
.5
-1.8
0.
2 0.
2 -0
.8
-1.2
-1
.0
-1.3
-2
.1
-0.8
-1
.0
-0.8
23
-0
.2
-0.7
-1
.2
0.8
-2.4
-1
.8
-0.9
-2
.0
-1.8
-1
.2
0.4
0.0
-0.9
-2
.0
0.2
-0.8
24
1.
0 1.
4 1.
2 -3
.5
-2.4
-1
.8
-0.9
0.
2 -0
.8
0.1
-1.0
0.
0 0.
3 -0
.8
-1.0
-0
.8
25
1.0
-0.7
0.
0 -2
.4
-0.6
-0
.8
-0.9
-0
.9
0.2
-2.6
-1
.0
-1.3
-2
.1
0.4
-1.0
-0
.8
26
1.0
0.4
0.0
0.8
-0.6
0.
2 0.
2 0.
2 0.
2 1.
4 0.
4 0.
0 0.
3 0.
4 0.
2 -0
.8
27
-1.5
0.
4 0.
0 0.
8 0.
4 0.
2 -0
.9
-0.9
-0
.8
0.1
-1.0
0.
0 -0
.9
-0.8
0.
2 -0
.8
28
1.0
1.4
0.0
0.8
-0.6
1.
2 -0
.9
-0.9
-0
.8
-2.6
-1
.0
0.0
-2.1
-2
.0
0.2
0.3
29
-1.5
-0
.7
0.0
-0.3
0.
4 -0
.8
-0.9
-0
.9
-0.8
0.
1 0.
4 0.
0 -0
.9
0.4
0.2
0.3
30
1.0
1.4
0.0
0.8
1.3
1.2
1.3
0.2
1.2
1.4
1.8
1.3
1.6
0.4
1.4
0.3
Not
e: z
cmm
t = z
-sco
re o
f the
‘com
mitm
ent’
item
, zco
mm
= z
-sco
re o
f the
‘com
mun
icat
ion’
item
, zac
cn =
z-s
core
of t
he ‘a
ccou
ntab
ility
’ ite
m, z
ldbx
= z
-sco
re o
f the
‘lea
ding
by
exam
ple’
ite
m, z
awrn
= z
-sco
re o
f the
‘sa
fety
aw
aren
ess’
item
, zal
gn =
z-s
core
of t
he ‘
safe
ty a
nd p
rodu
ctiv
ity a
lignm
ent’
item
, zst
nd =
z-s
core
of t
he ‘
safe
ty s
tand
ards
’ ite
m, z
init
= z-
scor
e of
the
‘saf
ety
initi
ativ
es’
item
, zin
tg =
z-s
core
of
the
‘saf
ety
inte
grat
ion
in b
usin
ess
goal
s’ it
em, z
prcp
= z
-sco
re o
f th
e ‘s
hare
d pe
rcep
tions
’ ite
m, z
resp
= z
-sco
re o
f th
e ‘s
afet
y re
spon
sibi
litie
s’
item
, zsp
pt =
z-s
core
of
the
‘sup
port
ive
envi
ronm
ent’
item
, zin
vm =
z-s
core
of
the
‘wor
kers
’ in
volv
emen
t’ it
em, z
rlsp
= z
-sco
re o
f th
e ‘w
orke
rs’
rela
tions
hips
’ ite
m, z
wkl
d =
z-sc
ore
of
the
‘wor
kloa
d’ it
em, z
prsr
= z
-sco
re o
f the
‘wor
k pr
essu
re’ i
tem
A S
yste
m D
ynam
ics
App
roac
h to
Con
stru
ctio
n Sa
fety
Cul
ture
22
2
case
zc
mm
t zc
omm
za
ccn
zldb
x za
wrn
za
lgn
zstn
d zi
nit
zint
g zp
rcp
zres
p zs
ppt
zinv
m
zrls
p zw
kld
zprs
r 31
-0
.2
-2.8
-1
.2
-2.4
-2
.4
-0.8
-0
.9
-0.9
-0
.8
0.1
0.4
0.0
-0.9
-2
.0
-1.0
-0
.8
32
-0.2
0.
4 0.
0 0.
8 0.
4 0.
2 0.
2 0.
2 -0
.8
0.1
0.4
-1.3
0.
3 0.
4 0.
2 0.
3 33
-1
.5
-1.8
-1
.2
0.8
-1.5
0.
2 0.
2 0.
2 0.
2 -1
.2
0.4
-1.3
-0
.9
-2.0
-1
.0
0.3
34
-0.2
-0
.7
0.0
0.8
-0.6
-0
.8
-0.9
-0
.9
-0.8
0.
1 0.
4 0.
0 0.
3 0.
4 0.
2 0.
3 35
-0
.2
-0.7
0.
0 -0
.3
0.4
0.2
0.2
0.2
-0.8
0.
1 0.
4 0.
0 0.
3 -0
.8
-2.1
-0
.8
36
1.0
1.4
1.2
0.8
1.3
1.2
1.3
1.3
1.2
1.4
1.8
1.3
1.6
1.6
1.4
1.3
37
-2.7
-1
.8
-1.2
-2
.4
-0.6
-1
.8
-3.1
-2
.0
-0.8
0.
1 -1
.0
-1.3
-0
.9
-0.8
0.
2 -3
.0
38
-0.2
-0
.7
0.0
-0.3
-0
.6
0.2
1.3
0.2
0.2
1.4
0.4
0.0
0.3
-0.8
0.
2 0.
3 39
1.
0 0.
4 1.
2 0.
8 0.
4 1.
2 0.
2 0.
2 1.
2 0.
1 0.
4 0.
0 0.
3 -0
.8
0.2
0.3
40
-0.2
0.
4 0.
0 -0
.3
0.4
0.2
0.2
0.2
-0.8
0.
1 0.
4 0.
0 0.
3 0.
4 0.
2 0.
3 41
-0
.2
-0.7
0.
0 0.
8 0.
4 0.
2 0.
2 0.
2 -0
.8
0.1
0.4
0.0
-0.9
1.
6 0.
2 0.
3 42
-0
.2
0.4
0.0
-0.3
0.
4 0.
2 0.
2 0.
2 0.
2 0.
1 -1
.0
-1.3
0.
3 -0
.8
0.2
0.3
43
1.0
1.4
1.2
0.8
0.4
1.2
1.3
1.3
1.2
1.4
1.8
1.3
1.6
-0.8
0.
2 1.
3 44
-0
.2
0.4
1.2
0.8
1.3
0.2
0.2
1.3
1.2
1.4
1.8
0.0
0.3
1.6
1.4
1.3
45
-0.2
-0
.7
0.0
-0.3
0.
4 -0
.8
0.2
0.2
0.2
0.1
-1.0
0.
0 0.
3 -0
.8
-1.0
0.
3 46
1.
0 1.
4 1.
2 -0
.3
0.4
0.2
0.2
0.2
1.2
1.4
0.4
1.3
0.3
0.4
0.2
-0.8
47
1.
0 -0
.7
0.0
-1.4
0.
4 0.
2 0.
2 0.
2 1.
2 0.
1 -1
.0
0.0
0.3
-0.8
1.
4 -0
.8
48
-0.2
0.
4 0.
0 -0
.3
0.4
0.2
0.2
0.2
0.2
0.1
0.4
0.0
0.3
-0.8
-1
.0
0.3
49
1.0
1.4
1.2
0.8
1.3
1.2
1.3
1.3
1.2
1.4
1.8
1.3
1.6
1.6
1.4
1.3
50
-0.2
-0
.7
0.0
-0.3
-1
.5
0.2
0.2
0.2
-0.8
-1
.2
0.4
0.0
0.3
0.4
0.2
-0.8
51
-0
.2
0.4
0.0
-0.3
0.
4 0.
2 0.
2 0.
2 0.
2 0.
1 0.
4 0.
0 0.
3 0.
4 0.
2 0.
3 52
1.
0 1.
4 0.
0 0.
8 0.
4 0.
2 0.
2 1.
3 0.
2 1.
4 1.
8 0.
0 -2
.1
0.4
1.4
1.3
53
-0.2
0.
4 1.
2 0.
8 0.
4 1.
2 1.
3 0.
2 1.
2 1.
4 1.
8 0.
0 1.
6 1.
6 1.
4 1.
3 54
1.
0 0.
4 0.
0 0.
8 0.
4 0.
2 0.
2 0.
2 0.
2 0.
1 0.
4 0.
0 0.
3 0.
4 0.
2 0.
3 55
1.
0 1.
4 1.
2 0.
8 1.
3 1.
2 1.
3 1.
3 1.
2 1.
4 0.
4 0.
0 1.
6 0.
4 1.
4 1.
3 56
1.
0 0.
4 0.
0 0.
8 1.
3 1.
2 0.
2 0.
2 0.
2 0.
1 0.
4 0.
0 0.
3 0.
4 1.
4 0.
3 57
-0
.2
0.4
0.0
-0.3
0.
4 0.
2 0.
2 0.
2 0.
2 0.
1 0.
4 0.
0 0.
3 0.
4 0.
2 0.
3 58
-0
.2
0.4
0.0
0.8
0.4
1.2
0.2
0.2
0.2
0.1
0.4
0.0
0.3
0.4
0.2
0.3
59
1.0
0.4
0.0
0.8
0.4
0.2
0.2
1.3
0.2
-1.2
-1
.0
1.3
-0.9
0.
4 0.
2 0.
3 60
1.
0 -0
.7
-1.2
-1
.4
-0.6
1.
2 -2
.0
1.3
-2.8
-1
.2
-1.0
-1
.3
0.3
-0.8
-3
.3
-1.9
N
ote:
zcm
mt =
z-s
core
of t
he ‘c
omm
itmen
t’ it
em, z
com
m =
z-s
core
of t
he ‘c
omm
unic
atio
n’ it
em, z
accn
= z
-sco
re o
f the
‘acc
ount
abili
ty’ i
tem
, zld
bx =
z-s
core
of t
he ‘l
eadi
ng b
y ex
ampl
e’
item
, zaw
rn =
z-s
core
of t
he ‘
safe
ty a
war
enes
s’ it
em, z
algn
= z
-sco
re o
f the
‘sa
fety
and
pro
duct
ivity
alig
nmen
t’ it
em, z
stnd
= z
-sco
re o
f the
‘sa
fety
sta
ndar
ds’
item
, zin
it =
z-sc
ore
of th
e ‘s
afet
y in
itiat
ives
’ ite
m, z
intg
= z
-sco
re o
f th
e ‘s
afet
y in
tegr
atio
n in
bus
ines
s go
als’
item
, zpr
cp =
z-s
core
of
the
‘sha
red
perc
eptio
ns’
item
, zre
sp =
z-s
core
of
the
‘saf
ety
resp
onsi
bilit
ies’
ite
m, z
sppt
= z
-sco
re o
f th
e ‘s
uppo
rtiv
e en
viro
nmen
t’ it
em, z
invm
= z
-sco
re o
f th
e ‘w
orke
rs’
invo
lvem
ent’
item
, zrl
sp =
z-s
core
of
the
‘wor
kers
’ re
latio
nshi
ps’
item
, zw
kld
= z-
scor
e of
th
e ‘w
orkl
oad’
item
, zpr
sr =
z-s
core
of t
he ‘w
ork
pres
sure
’ ite
m
A System Dynamics Approach to Construction Safety Culture
223
case zcmmt zcomm zaccn zldbx zawrn zalgn zstnd zinit zintg zprcp zresp zsppt zinvm zrlsp zwkld zprsr 61 -1.5 -0.7 0.0 -0.3 0.4 -0.8 -0.9 -0.9 0.2 0.1 0.4 -1.3 -0.9 0.4 -1.0 -1.9 62 -0.2 -1.8 -1.2 -0.3 -0.6 0.2 -0.9 -0.9 -0.8 0.1 -1.0 -1.3 -0.9 -0.8 -1.0 0.3 63 -1.5 -1.8 -2.4 -1.4 -0.6 -0.8 0.2 -0.9 -1.8 -1.2 -1.0 0.0 -2.1 -0.8 0.2 -0.8 64 1.0 0.4 0.0 -0.3 0.4 0.2 -0.9 -0.9 -0.8 0.1 0.4 0.0 0.3 0.4 0.2 1.3 65 -0.2 0.4 0.0 -0.3 0.4 0.2 0.2 0.2 0.2 0.1 0.4 0.0 0.3 0.4 0.2 0.3 66 1.0 -2.8 1.2 0.8 1.3 -2.8 1.3 1.3 1.2 1.4 1.8 0.0 1.6 0.4 1.4 1.3 67 -0.2 -1.8 0.0 -0.3 0.4 0.2 0.2 0.2 0.2 0.1 0.4 0.0 -0.9 -0.8 0.2 0.3 68 -1.5 0.4 1.2 0.8 -0.6 0.2 1.3 0.2 0.2 0.1 0.4 1.3 1.6 -0.8 0.2 0.3 69 -0.2 0.4 0.0 -0.3 1.3 0.2 0.2 0.2 1.2 0.1 0.4 1.3 0.3 0.4 0.2 0.3 70 -0.2 0.4 0.0 -0.3 . 0.2 0.2 0.2 0.2 0.1 -1.0 0.0 -0.9 0.4 0.2 0.3 71 -1.5 0.4 0.0 0.8 -0.6 0.2 . . 0.2 -1.2 -1.0 0.0 -2.1 -0.8 0.2 -0.8 72 -1.5 -0.7 -1.2 -0.3 -0.6 -0.8 -0.9 0.2 0.2 0.1 0.4 1.3 0.3 -0.8 -1.0 -0.8 73 . . -2.4 0.8 -2.4 -0.8 -2.0 -3.2 -2.8 0.1 -1.0 -2.6 -2.1 0.4 -1.0 0.3 74 -0.2 0.4 0.0 0.8 . 0.2 0.2 0.2 1.2 0.1 0.4 1.3 0.3 0.4 -1.0 0.3 75 -0.2 0.4 0.0 -0.3 0.4 0.2 1.3 1.3 . 0.1 0.4 0.0 . 0.4 1.4 1.3 76 -3.9 -2.8 -3.6 -3.5 -2.4 -2.8 -2.0 -3.2 -2.8 -1.2 -2.4 -1.3 -0.9 -3.2 -1.0 -0.8 77 1.0 0.4 1.2 0.8 0.4 1.2 1.3 1.3 1.2 0.1 0.4 1.3 0.3 0.4 . 0.3 78 -0.2 0.4 0.0 -1.4 1.3 1.2 1.3 1.3 -0.8 -1.2 0.4 1.3 0.3 0.4 0.2 -0.8 79 -0.2 0.4 0.0 -0.3 0.4 0.2 . 0.2 0.2 0.1 0.4 0.0 0.3 -0.8 0.2 1.3 80 -0.2 0.4 0.0 -0.3 0.4 . -0.9 -0.9 -0.8 0.1 0.4 . 0.3 0.4 0.2 0.3 81 1.0 1.4 1.2 0.8 1.3 1.2 1.3 1.3 1.2 0.1 0.4 0.0 1.6 0.4 0.2 1.3 82 1.0 1.4 1.2 0.8 1.3 -2.8 1.3 1.3 1.2 0.1 1.8 1.3 1.6 1.6 0.2 1.3 83 -0.2 0.4 1.2 0.8 1.3 1.2 1.3 0.2 0.2 -1.2 -1.0 1.3 0.3 0.4 0.2 1.3 84 -1.5 -1.8 -2.4 -0.3 -2.4 -1.8 -0.9 -0.9 -1.8 -1.2 -1.0 0.0 0.3 -2.0 -1.0 -1.9 85 1.0 0.4 0.0 -0.3 -0.6 0.2 0.2 0.2 0.2 1.4 0.4 1.3 0.3 0.4 1.4 0.3 86 -0.2 -0.7 -1.2 -1.4 0.4 0.2 0.2 0.2 0.2 0.1 0.4 0.0 0.3 0.4 1.4 0.3 87 1.0 1.4 1.2 -0.3 0.4 1.2 -0.9 1.3 1.2 1.4 1.8 1.3 0.3 0.4 1.4 0.3 88 -0.2 1.4 1.2 -0.3 1.3 1.2 1.3 1.3 1.2 1.4 1.8 1.3 1.6 1.6 1.4 1.3 89 -0.2 0.4 0.0 0.8 0.4 -0.8 0.2 0.2 0.2 0.1 -1.0 0.0 1.6 1.6 1.4 1.3 90 -0.2 0.4 0.0 0.8 1.3 0.2 0.2 0.2 0.2 0.1 0.4 1.3 0.3 1.6 1.4 0.3
Note: zcmmt = z-score of the ‘commitment’ item, zcomm = z-score of the ‘communication’ item, zaccn = z-score of the ‘accountability’ item, zldbx = z-score of the ‘leading by example’ item, zawrn = z-score of the ‘safety awareness’ item, zalgn = z-score of the ‘safety and productivity alignment’ item, zstnd = z-score of the ‘safety standards’ item, zinit = z-score of the ‘safety initiatives’ item, zintg = z-score of the ‘safety integration in business goals’ item, zprcp = z-score of the ‘shared perceptions’ item, zresp = z-score of the ‘safety responsibilities’ item, zsppt = z-score of the ‘supportive environment’ item, zinvm = z-score of the ‘workers’ involvement’ item, zrlsp = z-score of the ‘workers’ relationships’ item, zwkld = z-score of the ‘workload’ item, zprsr = z-score of the ‘work pressure’ item
A System Dynamics Approach to Construction Safety Culture
224
case zcmmt zcomm zaccn zldbx zawrn zalgn zstnd zinit zintg zprcp zresp zsppt zinvm zrlsp zwkld zprsr 91 1.0 1.4 0.0 0.8 1.3 1.2 . 0.2 1.2 0.1 0.4 0.0 . -0.8 0.2 0.3 92 -0.2 -0.7 -1.2 -0.3 -1.5 0.2 -0.9 -0.9 0.2 0.1 0.4 1.3 0.3 0.4 -1.0 -1.9 93 -0.2 -0.7 -1.2 -0.3 -0.6 -0.8 -0.9 -0.9 -0.8 -1.2 -1.0 0.0 0.3 -0.8 0.2 0.3 94 1.0 0.4 1.2 0.8 -0.6 1.2 0.2 0.2 -0.8 0.1 0.4 1.3 -2.1 1.6 0.2 0.3 95 1.0 -0.7 0.0 0.8 -0.6 0.2 -0.9 -0.9 -0.8 0.1 -1.0 0.0 -0.9 0.4 -1.0 -0.8 96 1.0 -0.7 1.2 0.8 -0.6 -0.8 0.2 -0.9 0.2 1.4 -1.0 -1.3 -2.1 -0.8 -2.1 -1.9 97 -0.2 -0.7 0.0 -0.3 -0.6 0.2 -0.9 -0.9 -0.8 0.1 0.4 1.3 -0.9 0.4 0.2 -0.8 98 1.0 -0.7 -1.2 -1.4 -0.6 1.2 -2.0 1.3 -2.8 -1.2 -1.0 -1.3 0.3 -0.8 -3.3 -1.9 99 -0.2 -0.7 -1.2 -0.3 -0.6 -0.8 -0.9 -0.9 -0.8 -1.2 -1.0 0.0 0.3 -0.8 0.2 0.3
100 -0.2 -0.7 0.0 -0.3 0.4 -0.8 0.2 0.2 0.2 0.1 -1.0 0.0 0.3 -0.8 -1.0 0.3 101 -1.5 -0.7 0.0 -0.3 0.4 -0.8 -0.9 -0.9 0.2 0.1 0.4 -1.3 -0.9 0.4 -1.0 -1.9 102 -0.2 -0.7 0.0 0.8 0.4 0.2 0.2 0.2 -0.8 0.1 0.4 0.0 -0.9 1.6 0.2 0.3 103 1.0 -0.7 0.0 -1.4 0.4 0.2 0.2 0.2 1.2 0.1 -1.0 0.0 0.3 -0.8 1.4 -0.8 104 -0.2 0.4 0.0 -0.3 0.4 0.2 0.2 0.2 0.2 0.1 0.4 0.0 0.3 -0.8 -1.0 0.3 105 1.0 0.4 1.2 0.8 -0.6 0.2 0.2 0.2 1.2 0.1 0.4 -1.3 0.3 0.4 0.2 0.3 106 1.0 -0.7 0.0 0.8 -0.6 0.2 -0.9 -0.9 -0.8 0.1 -1.0 0.0 -0.9 0.4 -1.0 -0.8 107 1.0 1.4 0.0 0.8 1.3 1.2 1.3 0.2 1.2 1.4 1.8 1.3 1.6 0.4 1.4 0.3 108 -0.2 0.4 0.0 -0.3 0.4 -0.8 -0.9 0.2 0.2 -1.2 -1.0 0.0 0.3 0.4 -1.0 -0.8 109 1.0 0.4 1.2 0.8 -0.6 1.2 0.2 0.2 -0.8 0.1 0.4 1.3 -2.1 1.6 0.2 0.3 110 -0.2 0.4 0.0 0.8 1.3 0.2 0.2 0.2 0.2 0.1 0.4 1.3 0.3 1.6 0.2 1.3 111 1.0 1.4 1.2 0.8 1.3 1.2 1.3 1.3 1.2 1.4 0.4 0.0 0.3 0.4 0.2 0.3 112 -0.2 -0.7 -1.2 0.8 -2.4 -1.8 -0.9 -2.0 -1.8 -1.2 0.4 0.0 -0.9 -2.0 0.2 -0.8 113 -2.7 -1.8 -2.4 -2.4 -1.5 -2.8 -2.0 -2.0 -1.8 -2.6 -2.4 -3.8 -0.9 -0.8 -1.0 -1.9 114 -2.7 -1.8 -2.4 -2.4 -1.5 -1.8 -2.0 -2.0 -1.8 -2.6 -2.4 -2.6 -0.9 -2.0 -2.1 -1.9 115 -2.7 -1.8 -2.4 -2.4 -1.5 -1.8 -2.0 -2.0 -1.8 -2.6 -2.4 -2.6 -0.9 -2.0 -2.1 -1.9 Note: zcmmt = z-score of the ‘commitment’ item, zcomm = z-score of the ‘communication’ item, zaccn = z-score of the ‘accountability’ item, zldbx = z-score of the ‘leading by example’ item, zawrn = z-score of the ‘safety awareness’ item, zalgn = z-score of the ‘safety and productivity alignment’ item, zstnd = z-score of the ‘safety standards’ item, zinit = z-score of the ‘safety initiatives’ item, zintg = z-score of the ‘safety integration in business goals’ item, zprcp = z-score of the ‘shared perceptions’ item, zresp = z-score of the ‘safety responsibilities’ item, zsppt = z-score of the ‘supportive environment’ item, zinvm = z-score of the ‘workers’ involvement’ item, zrlsp = z-score of the ‘workers’ relationships’ item, zwkld = z-score of the ‘workload’ item, zprsr = z-score of the ‘work pressure’ item
A System Dynamics Approach to Construction Safety Culture
225
case zcoop zfinc zresc zhmnr ztrng zrisk zfdbk znobm zhskp zdocu zbnmk zjstf zswbh zacci zcstm zimge zmrle zcost 1 -0.7 -0.8 0.1 -1.1 -0.9 -0.5 -0.9 -0.2 -0.9 -0.9 -0.3 -0.8 -1.0 -0.1 0.1 0.2 0.2 0.0 2 0.4 0.3 0.1 0.1 0.2 0.5 0.2 0.7 0.3 0.1 0.7 0.4 1.4 1.1 0.1 0.2 0.2 1.1 3 0.4 1.4 1.2 1.2 1.3 1.5 1.3 -0.2 0.3 1.1 1.6 1.7 1.4 1.1 1.3 1.4 1.5 1.1 4 -0.7 -0.8 0.1 0.1 1.3 0.5 0.2 -0.2 0.3 0.1 0.7 -0.8 0.2 -0.1 0.1 0.2 -1.0 0.0 5 0.4 1.4 0.1 1.2 0.2 1.5 0.2 -0.2 0.3 1.1 0.7 0.4 0.2 1.1 0.1 0.2 0.2 1.1 6 -0.7 -0.8 0.1 -1.1 -0.9 0.5 0.2 -0.2 -0.9 1.1 -0.3 -0.8 0.2 1.1 0.1 -0.9 -1.0 0.0 7 0.4 0.3 0.1 -1.1 0.2 -0.5 -0.9 -0.2 -0.9 0.1 -0.3 0.4 0.2 -0.1 0.1 -0.9 0.2 0.0 8 0.4 0.3 0.1 1.2 0.2 -0.5 0.2 -0.2 0.3 0.1 0.7 0.4 0.2 -0.1 0.1 0.2 0.2 0.0 9 1.5 1.4 1.2 1.2 1.3 0.5 0.2 -0.2 1.6 1.1 0.7 0.4 0.2 -0.1 0.1 0.2 0.2 1.1
10 -1.8 -0.8 -2.0 -2.2 0.2 -2.4 -2.0 1.6 -0.9 -3.0 -2.2 -2.0 -2.3 -1.4 -1.0 -3.2 -2.2 -3.3 11 -0.7 -0.8 0.1 0.1 0.2 0.5 -0.9 -0.2 0.3 0.1 0.7 -0.8 -1.0 -0.1 0.1 0.2 -1.0 0.0 12 -0.7 0.3 -0.9 0.1 0.2 0.5 0.2 0.7 -0.9 0.1 -0.3 0.4 0.2 -1.4 -1.0 0.2 0.2 0.0 13 0.4 1.4 0.1 0.1 0.2 0.5 0.2 -0.2 0.3 0.1 0.7 0.4 0.2 1.1 0.1 0.2 1.5 1.1 14 0.4 0.3 0.1 -1.1 -0.9 -0.5 -0.9 -0.2 0.3 -0.9 -0.3 -0.8 0.2 1.1 1.3 0.2 0.2 0.0 15 -0.7 0.3 0.1 0.1 -0.9 -0.5 -0.9 0.7 0.3 -0.9 -0.3 -0.8 0.2 1.1 1.3 0.2 0.2 0.0 16 0.4 0.3 0.1 -1.1 0.2 -0.5 0.2 -0.2 1.6 0.1 -0.3 -0.8 0.2 1.1 1.3 0.2 0.2 0.0 17 0.4 0.3 0.1 0.1 0.2 -0.5 -0.9 -0.2 0.3 0.1 -0.3 -0.8 -1.0 -0.1 -2.2 1.4 -3.4 1.1 18 0.4 1.4 1.2 0.1 -0.9 0.5 1.3 -0.2 0.3 1.1 0.7 0.4 1.4 -0.1 0.1 0.2 0.2 0.0 19 0.4 0.3 0.1 0.1 0.2 1.5 0.2 -0.2 1.6 1.1 0.7 0.4 1.4 1.1 1.3 0.2 0.2 0.0 20 -0.7 1.4 0.1 1.2 0.2 0.5 0.2 0.7 1.6 1.1 1.6 0.4 0.2 1.1 1.3 1.4 0.2 0.0 21 0.4 -1.8 0.1 0.1 -0.9 0.5 -0.9 -1.1 0.3 0.1 -0.3 0.4 0.2 -0.1 -1.0 -0.9 0.2 0.0 22 -0.7 -1.8 -2.0 -1.1 -0.9 -0.5 -0.9 -0.2 0.3 0.1 -1.3 0.4 0.2 -0.1 -1.0 -0.9 -1.0 0.0 23 -1.8 -0.8 -0.9 0.1 -0.9 -1.4 -2.0 -0.2 -0.9 -0.9 -1.3 -0.8 -1.0 -0.1 -1.0 -2.1 0.2 0.0 24 -0.7 0.3 -0.9 -1.1 -0.9 0.5 -0.9 0.7 -0.9 0.1 -0.3 -0.8 0.2 -1.4 0.1 -0.9 0.2 -1.1 25 -2.8 -1.8 -0.9 0.1 -2.0 -0.5 0.2 -0.2 0.3 0.1 -0.3 -0.8 -1.0 -0.1 0.1 -2.1 -2.2 0.0 26 0.4 0.3 1.2 -1.1 1.3 0.5 0.2 0.7 0.3 0.1 0.7 -0.8 -1.0 1.1 0.1 0.2 0.2 0.0 27 -0.7 -1.8 -0.9 0.1 1.3 0.5 0.2 -0.2 -0.9 -0.9 0.7 0.4 -1.0 -1.4 -1.0 0.2 0.2 0.0 28 -1.8 -1.8 -2.0 -1.1 -2.0 -1.4 -0.9 0.7 0.3 -3.0 -2.2 -0.8 -1.0 -0.1 -1.0 -0.9 -1.0 0.0 29 0.4 -0.8 -0.9 0.1 -0.9 0.5 -0.9 0.7 0.3 0.1 0.7 0.4 0.2 -0.1 0.1 0.2 0.2 0.0 30 1.5 1.4 1.2 1.2 1.3 1.5 1.3 -1.1 0.3 1.1 0.7 1.7 1.4 1.1 1.3 1.4 1.5 1.1
Note: zcoop = z-score of the ‘stakeholders’ cooperation’ item, zfinc = z-score of the ‘financial resources’ item, zresc = z-score of the ‘safety resources’ item, zhmnr = z-score of the ‘human resources’ item, ztrng = z-score of the ‘training’ item, zrisk = z-score of the ‘risk assessment’ item, zfdbk = z-score of the ‘feedback’ item, znobm = z-score of the ‘no-blame approach’ item, zhskp = z-score of the ‘housekeeping’ item, zdocu = z-score of the ‘safety documentation’ item, zbnmk = z-score of the ‘benchmarking system’ item, zjstf = z-score of the ‘job satisfaction’ item, zswbh = z-score of the ‘safe work behaviour’ item, zacci = z-score of the ‘number of accidents’ item, zcstm = z-score of the ‘customers’ expectations’ item, zimge = z-score of the ‘industrial image’ item, zmrle = z-score of the ‘workforce morale’ item, zcost = z-score of the ‘cost of accidents’ item
A S
yste
m D
ynam
ics
App
roac
h to
Con
stru
ctio
n Sa
fety
Cul
ture
22
6
case
zc
oop
zfin
c zr
esc
zhm
nr
ztrn
g zr
isk
zfdb
k zn
obm
zh
skp
zdoc
u zb
nmk
zjst
f zs
wbh
za
cci
zcst
m
zim
ge
zmrl
e zc
ost
31
0.4
-0.8
-0
.9
0.1
0.2
0.5
0.2
-1.1
0.
3 0.
1 -1
.3
0.4
0.2
-0.1
-1
.0
0.2
-2.2
0.
0 32
0.
4 0.
3 0.
1 0.
1 0.
2 -0
.5
0.2
0.7
0.3
0.1
0.7
0.4
0.2
-0.1
0.
1 0.
2 0.
2 1.
1 33
-1
.8
-0.8
-0
.9
-1.1
-0
.9
-1.4
-0
.9
-0.2
-0
.9
-0.9
-1
.3
-0.8
-1
.0
-1.4
-1
.0
0.2
0.2
0.0
34
0.4
-0.8
-0
.9
0.1
0.2
-0.5
0.
2 -0
.2
0.3
0.1
0.7
0.4
0.2
-1.4
-1
.0
0.2
0.2
0.0
35
-2.8
-1
.8
-0.9
-2
.2
-0.9
-1
.4
-0.9
-2
.0
0.3
-0.9
-1
.3
-3.3
-1
.0
1.1
0.1
0.2
0.2
0.0
36
1.5
1.4
1.2
1.2
1.3
0.5
1.3
1.6
1.6
1.1
-2.2
0.
4 1.
4 1.
1 1.
3 1.
4 0.
2 -3
.3
37
0.4
-0.8
1.
2 -1
.1
0.2
-1.4
-2
.0
0.7
0.3
-0.9
-1
.3
0.4
-2.3
-0
.1
1.3
0.2
-1.0
-1
.1
38
0.4
1.4
1.2
1.2
1.3
0.5
-0.9
0.
7 0.
3 0.
1 -0
.3
0.4
1.4
1.1
0.1
0.2
0.2
0.0
39
0.4
0.3
0.1
0.1
0.2
0.5
1.3
0.7
0.3
0.1
0.7
0.4
0.2
1.1
0.1
1.4
0.2
1.1
40
0.4
0.3
0.1
0.1
0.2
0.5
0.2
0.7
0.3
0.1
0.7
0.4
0.2
-0.1
0.
1 0.
2 0.
2 0.
0 41
0.
4 -0
.8
0.1
-1.1
-0
.9
-0.5
0.
2 0.
7 -0
.9
0.1
0.7
0.4
0.2
-0.1
-1
.0
0.2
-1.0
0.
0 42
0.
4 0.
3 0.
1 0.
1 0.
2 0.
5 0.
2 0.
7 0.
3 0.
1 0.
7 0.
4 0.
2 -0
.1
0.1
0.2
0.2
0.0
43
1.5
1.4
0.1
1.2
1.3
1.5
1.3
-0.2
0.
3 0.
1 -0
.3
0.4
1.4
1.1
1.3
0.2
1.5
0.0
44
1.5
0.3
0.1
1.2
0.2
1.5
1.3
0.7
1.6
1.1
1.6
0.4
1.4
1.1
1.3
1.4
1.5
1.1
45
-0.7
-0
.8
0.1
-1.1
0.
2 -0
.5
0.2
-0.2
0.
3 0.
1 0.
7 0.
4 0.
2 -0
.1
0.1
0.2
-1.0
-1
.1
46
0.4
1.4
1.2
0.1
1.3
1.5
1.3
0.7
0.3
1.1
0.7
0.4
0.2
-1.4
0.
1 1.
4 0.
2 -1
.1
47
0.4
0.3
1.2
1.2
1.3
-0.5
-0
.9
0.7
0.3
1.1
0.7
0.4
0.2
-0.1
1.
3 1.
4 1.
5 1.
1 48
0.
4 0.
3 0.
1 0.
1 0.
2 0.
5 0.
2 0.
7 -2
.2
0.1
0.7
0.4
0.2
-0.1
0.
1 0.
2 0.
2 0.
0 49
1.
5 0.
3 1.
2 1.
2 1.
3 1.
5 1.
3 0.
7 0.
3 1.
1 0.
7 0.
4 1.
4 1.
1 1.
3 1.
4 1.
5 1.
1 50
0.
4 0.
3 0.
1 0.
1 0.
2 -0
.5
0.2
-1.1
0.
3 0.
1 -0
.3
0.4
0.2
-0.1
0.
1 0.
2 0.
2 0.
0 51
0.
4 0.
3 0.
1 0.
1 0.
2 0.
5 0.
2 0.
7 0.
3 0.
1 0.
7 0.
4 0.
2 -0
.1
0.1
0.2
0.2
0.0
52
1.5
1.4
1.2
1.2
1.3
1.5
1.3
1.6
0.3
1.1
0.7
0.4
0.2
-0.1
0.
1 0.
2 0.
2 1.
1 53
0.
4 1.
4 1.
2 1.
2 1.
3 0.
5 1.
3 0.
7 1.
6 1.
1 0.
7 0.
4 1.
4 1.
1 1.
3 0.
2 1.
5 1.
1 54
0.
4 0.
3 0.
1 0.
1 0.
2 0.
5 0.
2 0.
7 0.
3 0.
1 0.
7 0.
4 0.
2 -0
.1
0.1
0.2
0.2
0.0
55
0.4
1.4
1.2
1.2
1.3
1.5
1.3
-0.2
0.
3 1.
1 1.
6 1.
7 1.
4 1.
1 1.
3 1.
4 1.
5 1.
1 56
0.
4 1.
4 1.
2 1.
2 0.
2 0.
5 1.
3 1.
6 0.
3 1.
1 0.
7 0.
4 0.
2 -0
.1
0.1
0.2
0.2
0.0
57
0.4
0.3
0.1
0.1
0.2
0.5
0.2
0.7
0.3
0.1
0.7
0.4
0.2
-0.1
0.
1 0.
2 0.
2 0.
0 58
0.
4 0.
3 0.
1 0.
1 0.
2 0.
5 0.
2 0.
7 0.
3 0.
1 0.
7 0.
4 0.
2 -0
.1
0.1
0.2
0.2
0.0
59
-0.7
0.
3 0.
1 0.
1 -0
.9
-0.5
0.
2 0.
7 0.
3 0.
1 0.
7 0.
4 0.
2 -0
.1
-1.0
0.
2 -1
.0
-1.1
60
-1
.8
-1.8
-0
.9
-2.2
1.
3 -1
.4
-0.9
-2
.0
-2.2
-2
.0
-0.3
-3
.3
-2.3
-1
.4
-2.2
-2
.1
-1.0
-1
.1
Not
e: z
coop
= z
-sco
re o
f th
e ‘s
take
hold
ers’
coo
pera
tion’
ite
m, z
finc
= z
-sco
re o
f th
e ‘f
inan
cial
res
ourc
es’
item
, zre
sc =
z-s
core
of
the
‘saf
ety
reso
urce
s’ i
tem
, zhm
nr =
z-s
core
of
the
‘hum
an r
esou
rces
’ ite
m, z
trng
= z
-sco
re o
f th
e ‘t
rain
ing’
item
, zri
sk =
z-s
core
of
the
‘ris
k as
sess
men
t’ it
em, z
fdbk
= z
-sco
re o
f th
e ‘f
eedb
ack’
item
, zno
bm =
z-s
core
of
the
‘no-
blam
e ap
proa
ch’
item
, zhs
kp =
z-s
core
of
the
‘hou
seke
epin
g’ it
em, z
docu
= z
-sco
re o
f th
e ‘s
afet
y do
cum
enta
tion’
item
, zbn
mk
= z-
scor
e of
the
‘ben
chm
arki
ng s
yste
m’
item
, zjs
tf =
z-sc
ore
of
the
‘job
satis
fact
ion’
item
, zsw
bh =
z-s
core
of
the
‘saf
e w
ork
beha
viou
r’ it
em, z
acci
= z
-sco
re o
f th
e ‘n
umbe
r of
acc
iden
ts’
item
, zcs
tm =
z-s
core
of
the
‘cus
tom
ers’
exp
ecta
tions
’ ite
m,
zim
ge =
z-s
core
of t
he ‘i
ndus
tria
l im
age’
item
, zm
rle
= z-
scor
e of
the
‘wor
kfor
ce m
oral
e’ it
em, z
cost
= z
-sco
re o
f the
‘cos
t of a
ccid
ents
’ ite
m
A S
yste
m D
ynam
ics
App
roac
h to
Con
stru
ctio
n Sa
fety
Cul
ture
22
7
case
zc
oop
zfin
c zr
esc
zhm
nr
ztrn
g zr
isk
zfdb
k zn
obm
zh
skp
zdoc
u zb
nmk
zjst
f zs
wbh
za
cci
zcst
m
zim
ge
zmrl
e zc
ost
61
-0.7
0.
3 0.
1 -1
.1
-2.0
-1
.4
0.2
-0.2
-2
.2
0.1
-1.3
-0
.8
0.2
-0.1
-1
.0
-0.9
-1
.0
0.0
62
-1.8
-0
.8
-0.9
-1
.1
0.2
0.5
-2.0
-0
.2
-0.9
-0
.9
-1.3
-0
.8
-1.0
-0
.1
0.1
0.2
0.2
-1.1
63
-0
.7
-0.8
0.
1 0.
1 -0
.9
0.5
-0.9
0.
7 -0
.9
-2.0
-0
.3
-0.8
-1
.0
-0.1
0.
1 0.
2 0.
2 0.
0 64
1.
5 0.
3 0.
1 0.
1 0.
2 -0
.5
0.2
-1.1
0.
3 0.
1 -0
.3
-0.8
0.
2 -0
.1
-1.0
0.
2 0.
2 0.
0 65
0.
4 0.
3 0.
1 0.
1 0.
2 0.
5 0.
2 0.
7 0.
3 0.
1 0.
7 0.
4 0.
2 -0
.1
0.1
0.2
0.2
0.0
66
1.5
1.4
1.2
1.2
1.3
1.5
1.3
1.6
1.6
1.1
-0.3
1.
7 1.
4 1.
1 1.
3 -0
.9
0.2
1.1
67
0.4
-0.8
0.
1 -1
.1
0.2
0.5
0.2
-0.2
0.
3 0.
1 0.
7 0.
4 0.
2 -0
.1
0.1
0.2
0.2
0.0
68
0.4
1.4
1.2
1.2
0.2
-0.5
0.
2 1.
6 1.
6 0.
1 -0
.3
0.4
0.2
-0.1
0.
1 0.
2 0.
2 1.
1 69
0.
4 0.
3 1.
2 0.
1 0.
2 0.
5 0.
2 0.
7 0.
3 0.
1 0.
7 0.
4 0.
2 -0
.1
0.1
0.2
0.2
0.0
70
0.4
0.3
0.1
0.1
0.2
0.5
0.2
-0.2
-0
.9
0.1
0.7
-0.8
0.
2 -0
.1
0.1
0.2
0.2
0.0
71
-0.7
-0
.8
-2.0
0.
1 -0
.9
-1.4
0.
2 -0
.2
-0.9
-2
.0
-0.3
-2
.0
-2.3
-1
.4
-2.2
0.
2 0.
2 -2
.2
72
0.4
0.3
0.1
0.1
. .
-0.9
-0
.2
. -0
.9
-0.3
.
-1.0
-1
.4
0.1
. -1
.0
0.0
73
1.5
. .
0.1
-0.9
0.
5 0.
2 1.
6 -0
.9
-3.0
-2
.2
-0.8
0.
2 1.
1 0.
1 0.
2 0.
2 1.
1 74
0.
4 0.
3 0.
1 0.
1 0.
2 0.
5 0.
2 -0
.2
1.6
0.1
0.7
0.4
0.2
1.1
1.3
1.4
1.5
1.1
75
. .
1.2
1.2
1.3
. 1.
3 .
1.6
1.1
1.6
. 0.
2 1.
1 0.
1 .
. 0.
0 76
-2
.8
-2.9
-3
.1
-3.3
-3
.1
-2.4
-3
.1
-2.0
.
-3.0
-2
.2
. -3
.5
-3.9
-3
.4
-3.2
-3
.4
-3.3
77
0.
4 0.
3 1.
2 1.
2 1.
3 0.
5 1.
3 0.
7 1.
6 1.
1 0.
7 0.
4 0.
2 1.
1 0.
1 .
1.5
1.1
78
. -0
.8
0.1
0.1
-0.9
-0
.5
0.2
1.6
0.3
1.1
0.7
-0.8
0.
2 -1
.4
. .
-1.0
0.
0 79
0.
4 -0
.8
-0.9
0.
1 -0
.9
0.5
0.2
0.7
-0.9
0.
1 0.
7 0.
4 0.
2 -0
.1
1.3
0.2
0.2
1.1
80
-0.7
0.
3 0.
1 0.
1 .
0.5
0.2
. .
. 0.
7 0.
4 0.
2 -0
.1
0.1
0.2
0.2
1.1
81
1.5
1.4
1.2
1.2
1.3
0.5
1.3
-0.2
0.
3 1.
1 -0
.3
0.4
1.4
1.1
1.3
1.4
1.5
1.1
82
1.5
1.4
1.2
1.2
1.3
1.5
1.3
1.6
1.6
1.1
0.7
1.7
1.4
1.1
1.3
1.4
1.5
1.1
83
-0.7
-0
.8
1.2
1.2
-0.9
0.
5 1.
3 0.
7 1.
6 1.
1 -0
.3
-0.8
0.
2 1.
1 -1
.0
-0.9
0.
2 1.
1 84
0.
4 -0
.8
1.2
-1.1
-0
.9
-0.5
0.
2 -2
.0
0.3
1.1
-1.3
-0
.8
0.2
-0.1
-1
.0
0.2
-1.0
1.
1 85
0.
4 0.
3 0.
1 0.
1 1.
3 0.
5 0.
2 -2
.0
0.3
1.1
0.7
0.4
0.2
1.1
1.3
0.2
-1.0
1.
1 86
-0
.7
0.3
0.1
0.1
-0.9
-0
.5
0.2
0.7
0.3
0.1
0.7
0.4
0.2
-0.1
0.
1 0.
2 0.
2 0.
0 87
1.
5 1.
4 1.
2 1.
2 1.
3 1.
5 1.
3 1.
6 0.
3 1.
1 1.
6 1.
7 0.
2 -0
.1
0.1
0.2
0.2
1.1
88
1.5
1.4
1.2
1.2
1.3
1.5
1.3
-2.0
1.
6 1.
1 1.
6 1.
7 1.
4 1.
1 1.
3 1.
4 1.
5 1.
1 89
0.
4 0.
3 1.
2 1.
2 1.
3 0.
5 1.
3 1.
6 1.
6 1.
1 0.
7 -0
.8
1.4
1.1
0.1
0.2
0.2
1.1
90
0.4
0.3
0.1
1.2
1.3
0.5
0.2
0.7
0.3
0.1
0.7
0.4
0.2
1.1
0.1
0.2
0.2
0.0
Not
e: z
coop
= z
-sco
re o
f th
e ‘s
take
hold
ers’
coo
pera
tion’
ite
m, z
finc
= z
-sco
re o
f th
e ‘f
inan
cial
res
ourc
es’
item
, zre
sc =
z-s
core
of
the
‘saf
ety
reso
urce
s’ i
tem
, zhm
nr =
z-s
core
of
the
‘hum
an r
esou
rces
’ ite
m, z
trng
= z
-sco
re o
f th
e ‘t
rain
ing’
item
, zri
sk =
z-s
core
of
the
‘ris
k as
sess
men
t’ it
em, z
fdbk
= z
-sco
re o
f th
e ‘f
eedb
ack’
item
, zno
bm =
z-s
core
of
the
‘no-
blam
e ap
proa
ch’
item
, zhs
kp =
z-s
core
of
the
‘hou
seke
epin
g’ it
em, z
docu
= z
-sco
re o
f th
e ‘s
afet
y do
cum
enta
tion’
item
, zbn
mk
= z-
scor
e of
the
‘ben
chm
arki
ng s
yste
m’
item
, zjs
tf =
z-sc
ore
of
the
‘job
satis
fact
ion’
item
, zsw
bh =
z-s
core
of
the
‘saf
e w
ork
beha
viou
r’ it
em, z
acci
= z
-sco
re o
f th
e ‘n
umbe
r of
acc
iden
ts’
item
, zcs
tm =
z-s
core
of
the
‘cus
tom
ers’
exp
ecta
tions
’ ite
m,
zim
ge =
z-s
core
of t
he ‘i
ndus
tria
l im
age’
item
, zm
rle
= z-
scor
e of
the
‘wor
kfor
ce m
oral
e’ it
em, z
cost
= z
-sco
re o
f the
‘cos
t of a
ccid
ents
’ ite
m
A System Dynamics Approach to Construction Safety Culture
228
case zcoop zfinc zresc zhmnr ztrng zrisk zfdbk znobm zhskp zdocu zbnmk zjstf zswbh zacci zcstm zimge zmrle zcost 91 0.4 1.4 0.1 0.1 0.2 0.5 1.3 -1.1 0.3 0.1 0.7 0.4 0.2 1.1 0.1 0.2 0.2 0.0 92 0.4 -0.8 -0.9 0.1 0.2 -1.4 -0.9 -0.2 -0.9 0.1 -1.3 -0.8 0.2 -0.1 -1.0 -0.9 -1.0 -1.1 93 -0.7 -0.8 -2.0 0.1 -0.9 -1.4 -0.9 -1.1 -0.9 0.1 -0.3 -0.8 -1.0 -1.4 -1.0 -0.9 0.2 0.0 94 0.4 -1.8 -2.0 1.2 -0.9 -2.4 0.2 -2.0 0.3 0.1 -2.2 1.7 1.4 1.1 1.3 0.2 0.2 -2.2 95 -0.7 -0.8 -2.0 -1.1 -0.9 -1.4 -0.9 -1.1 -0.9 -0.9 -1.3 0.4 0.2 -0.1 0.1 -0.9 0.2 0.0 96 -1.8 -0.8 -0.9 -3.3 -0.9 -1.4 -2.0 -1.1 -2.2 -2.0 -1.3 -0.8 -1.0 -1.4 0.1 -2.1 -1.0 1.1 97 0.4 -0.8 -2.0 0.1 -0.9 -1.4 -2.0 -2.0 -0.9 0.1 -1.3 0.4 -1.0 -0.1 0.1 -0.9 0.2 -1.1 98 -1.8 -1.8 -0.9 -2.2 1.3 -1.4 -0.9 -2.0 -2.2 -2.0 -0.3 -3.3 -2.3 -1.4 -2.2 -2.1 -1.0 -1.1 99 -0.7 -0.8 -2.0 0.1 -0.9 -1.4 -0.9 -1.1 -0.9 0.1 -0.3 -0.8 -1.0 -1.4 -1.0 -0.9 0.2 0.0
100 -0.7 -0.8 0.1 -1.1 0.2 -0.5 0.2 -0.2 0.3 0.1 0.7 0.4 0.2 -0.1 0.1 0.2 -1.0 -1.1 101 -0.7 0.3 0.1 -1.1 -2.0 -1.4 0.2 -0.2 -2.2 0.1 -1.3 -0.8 0.2 -0.1 -1.0 -0.9 -1.0 0.0 102 0.4 -0.8 0.1 -1.1 -0.9 -0.5 0.2 0.7 -0.9 0.1 0.7 0.4 0.2 -0.1 -1.0 0.2 -1.0 0.0 103 0.4 0.3 1.2 1.2 1.3 -0.5 -0.9 0.7 0.3 1.1 0.7 0.4 0.2 -0.1 1.3 1.4 1.5 1.1 104 0.4 0.3 0.1 0.1 0.2 0.5 0.2 0.7 -2.2 0.1 0.7 0.4 0.2 -0.1 0.1 0.2 0.2 0.0 105 0.4 1.4 1.2 0.1 -0.9 0.5 1.3 -0.2 0.3 1.1 0.7 0.4 1.4 -0.1 0.1 0.2 0.2 0.0 106 -0.7 -0.8 -2.0 -1.1 -0.9 -1.4 -0.9 -1.1 -0.9 -0.9 -1.3 0.4 0.2 -0.1 0.1 -0.9 0.2 0.0 107 1.5 1.4 1.2 1.2 1.3 1.5 1.3 -1.1 0.3 1.1 0.7 1.7 1.4 1.1 1.3 1.4 1.5 1.1 108 -0.7 0.3 -0.9 0.1 0.2 0.5 0.2 0.7 -0.9 0.1 -0.3 0.4 0.2 -1.4 -1.0 0.2 0.2 0.0 109 0.4 -1.8 -2.0 1.2 -0.9 -2.4 0.2 -2.0 0.3 0.1 -2.2 1.7 1.4 1.1 1.3 0.2 0.2 -2.2 110 0.4 1.4 0.1 0.1 0.2 0.5 0.2 -0.2 0.3 0.1 0.7 0.4 0.2 1.1 0.1 0.2 1.5 1.1 111 0.4 0.3 0.1 1.2 0.2 -0.5 0.2 -0.2 0.3 0.1 0.7 0.4 0.2 -0.1 0.1 0.2 0.2 0.0 112 -1.8 -0.8 -0.9 0.1 -0.9 -1.4 -2.0 -0.2 -0.9 -0.9 -1.3 -0.8 -1.0 -0.1 -1.0 -2.1 0.2 0.0 113 -0.7 -0.8 -0.9 -1.1 -2.0 -1.4 -2.0 -2.0 -0.9 -0.9 -2.2 -2.0 -2.3 -2.6 -1.0 -2.1 -2.2 -2.2 114 -1.8 -0.8 0.1 -1.1 -2.0 0.5 -2.0 -1.1 -2.2 -2.0 -1.3 -2.0 -2.3 -2.6 -2.2 -2.1 -2.2 -2.2 115 -1.8 -0.8 0.1 -1.1 -2.0 0.5 -2.0 -1.1 -2.2 -2.0 -1.3 -2.0 -2.3 -2.6 -2.2 -2.1 -2.2 -2.2 Note: zcoop = z-score of the ‘stakeholders’ cooperation’ item, zfinc = z-score of the ‘financial resources’ item, zresc = z-score of the ‘safety resources’ item, zhmnr = z-score of the ‘human resources’ item, ztrng = z-score of the ‘training’ item, zrisk = z-score of the ‘risk assessment’ item, zfdbk = z-score of the ‘feedback’ item, znobm = z-score of the ‘no-blame approach’ item, zhskp = z-score of the ‘housekeeping’ item, zdocu = z-score of the ‘safety documentation’ item, zbnmk = z-score of the ‘benchmarking system’ item, zjstf = z-score of the ‘job satisfaction’ item, zswbh = z-score of the ‘safe work behaviour’ item, zacci = z-score of the ‘number of accidents’ item, zcstm = z-score of the ‘customers’ expectations’ item, zimge = z-score of the ‘industrial image’ item, zmrle = z-score of the ‘workforce morale’ item, zcost = z-score of the ‘cost of accidents’ item
A System Dynamics Approach to Construction Safety Culture
229
AAppppeennddiixx 44
MMeeaassuurreemmeenntt MMooddeell RReessuullttss
A System Dynamics Approach to Construction Safety Culture
230
NOTES FOR MODEL (Default model)
Computation of degrees of freedom (Default model)
Number of distinct sample moments: 324
Number of distinct parameters to be estimated: 87
Degrees of freedom (324 - 87): 237
Result (Default model)
Minimum was achieved
Chi-square = 390.173
Degrees of freedom = 237
Probability level = .000
Estimates (Group number 1 - Default model)
Scalar Estimates (Group number 1 - Default model)
Maximum Likelihood Estimates
A System Dynamics Approach to Construction Safety Culture
231
Regression Weights: (Group number 1 - Default model)
Item Estimate S.E. C.R. P Label
Safety initiatives <--- Pol 1.00
Safety standards <--- Pol 0.93 0.09 10.41 ***
Number of accidents <--- Goals 1.03 0.17 6.21 ***
Industrial image <--- Goals 1.05 0.19 5.65 ***
Workforce morale <--- Goals 0.96 0.16 5.83 ***
Cost of accidents <--- Goals 1.00
Leading by example <--- Lds 0.82 0.12 6.84 ***
Accountability <--- Lds 1.00
Communication <--- Lds 1.07 0.11 9.37 ***
Commitment <--- Lds 0.87 0.10 8.51 ***
Safety awareness <--- Pol 1.05 0.11 9.95 ***
Safety and productivity alignment <--- Lds 0.88 0.12 7.15 ***
Workers’ relationships <--- Pol 0.64 0.09 6.85 ***
Human resources <--- Ppl 1.00
Stakeholders’ cooperation <--- Ppl 1.00 0.08 12.00 ***
Safety responsibilities <--- Ppl 0.44 0.06 7.08 ***
Financial resources <--- Prs 1.00
Safety resources <--- Prs 1.00 0.08 13.44 ***
Workers’ involvement <--- Prs 0.47 0.07 7.07 ***
Training <--- Prs 0.61 0.07 8.89 ***
Benchmarking system <--- Pro 0.85 0.11 7.82 ***
Safety integration in business goals <--- Pro 0.92 0.10 9.28 ***
Feedback <--- Pro 0.89 0.10 9.33 ***
Safety documentation <--- Pro 1.00
A System Dynamics Approach to Construction Safety Culture
232
Standardized Regression Weights: (Group number 1 - Default model)
Item Estimate
Safety initiatives <--- Pol .868
Safety standards <--- Pol .787
Number of accidents <--- Goals .742
Industrial image <--- Goals .651
Workforce morale <--- Goals .679
Cost of accidents <--- Goals .603
Leading by example <--- Lds .613
Accountability <--- Lds .837
Communication <--- Lds .782
Commitment <--- Lds .727
Safety awareness <--- Pol .766
Safety and productivity alignment <--- Lds .635
Workers’ relationships <--- Pol .588
Human resources <--- Ppl .848
Stakeholders’ cooperation <--- Ppl .874
Safety responsibilities <--- Ppl .609
Financial resources <--- Prs .898
Safety resources <--- Prs .869
Workers’ involvement <--- Prs .589
Training <--- Prs .691
Benchmarking system <--- Pro .673
Safety integration in business goals <--- Pro .766
Feedback <--- Pro .769
Safety documentation <--- Pro .813
A System Dynamics Approach to Construction Safety Culture
233
Intercepts: (Group number 1 - Default model)
Item Estimate S.E. C.R. P Label
Safety initiatives 3.82 0.08 48.52 ***
Safety standards 3.80 0.08 47.36 ***
Safety and productivity alignment 3.77 0.09 43.85 ***
Safety awareness 3.64 0.09 38.87 ***
Safety integration in business goals 3.78 0.09 44.30 ***
Safety responsibilities 3.76 0.06 58.98 ***
Workers’ involvement 3.73 0.07 50.82 ***
Workers’ relationships 3.72 0.08 49.84 ***
Human resources 3.72 0.11 35.38 ***
Safety resources 3.70 0.11 34.74 ***
Financial resources 3.55 0.10 34.52 ***
Stakeholders’ cooperation 3.50 0.10 34.31 ***
Feedback 3.78 0.08 45.76 ***
Training 3.79 0.08 46.54 ***
Number of accidents 4.10 0.07 57.75 ***
Industrial image 3.74 0.08 45.59 ***
Workforce morale 3.77 0.07 52.35 ***
Cost of accidents 3.97 0.09 46.86 ***
Leading by example 4.23 0.08 50.97 ***
Accountability 4.02 0.07 54.22 ***
Communication 3.65 0.09 43.11 ***
Commitment 4.16 0.07 56.41 ***
Benchmarking system 3.34 0.09 36.83 ***
Safety documentation 3.91 0.09 44.55 ***
A System Dynamics Approach to Construction Safety Culture
234
Covariances: (Group number 1 - Default model)
Estimate S.E. C.R. P Label
Pol <--> Lds .412 .071 5.829 ***
Lds <--> Ppl .421 .085 4.977 ***
Lds <--> Prs .465 .087 5.333 ***
Pro <--> Lds .395 .073 5.391 ***
Goals <--> Lds .261 .059 4.410 ***
Pol <--> Ppl .490 .093 5.252 ***
Pol <--> Prs .565 .098 5.777 ***
Pro <--> Pol .533 .087 6.127 ***
Pol <--> Goals .316 .068 4.673 ***
Ppl <--> Prs .901 .141 6.377 ***
Pro <--> Ppl .603 .107 5.626 ***
Goals <--> Ppl .478 .097 4.927 ***
Pro <--> Prs .664 .111 5.980 ***
Goals <--> Prs .475 .096 4.949 ***
Pro <--> Goals .353 .075 4.707 ***
Correlations: (Group number 1 - Default model)
Estimate Pol <--> Lds .855 Lds <--> Ppl .668
Lds <--> Prs .713
Pro <--> Lds .783
Goals <--> Lds .724
Pol <--> Ppl .706
Pol <--> Prs .786
Pro <--> Pol .960
Pol <--> Goals .796
Ppl <--> Prs .960
Pro <--> Ppl .831
Goals <--> Ppl .922
Pro <--> Prs .884
Goals <--> Prs .884
Pro <--> Goals .850
A System Dynamics Approach to Construction Safety Culture
235
Variances: (Group number 1 - Default model)
Estimate S.E. C.R. P Label
Pol .531 .093 5.707 ***
Goals .297 .088 3.369 ***
Lds .438 .083 5.292 ***
Ppl .906 .165 5.504 ***
Prs .971 .160 6.084 ***
Pro .581 .112 5.164 ***
e8 .174 .032 5.524 ***
e7 .279 .043 6.540 ***
e6 .502 .072 6.932 ***
e5 .412 .062 6.680 ***
e9 .343 .051 6.716 ***
e11 .291 .040 7.210 ***
e13 .401 .055 7.290 ***
e14 .416 .057 7.237 ***
e20 .355 .060 5.911 ***
e19 .316 .053 5.937 ***
e18 .233 .044 5.297 ***
e17 .280 .052 5.368 ***
e23 .318 .047 6.696 ***
e21 .395 .056 7.098 ***
e30 .258 .041 6.210 ***
e32 .442 .065 6.802 ***
e33 .319 .048 6.663 ***
e34 .519 .074 6.981 ***
e4 .489 .070 7.001 ***
e3 .188 .035 5.366 ***
e2 .318 .052 6.085 ***
e1 .291 .045 6.513 ***
e27 .513 .072 7.092 ***
e26 .299 .047 6.348 ***
A System Dynamics Approach to Construction Safety Culture
236
Squared Multiple Correlations: (Group number 1 - Default model)
Estimate
Safety documentation .660
Benchmarking system .453
Commitment .529
Communication .612
Accountability .700
Leading by example .375
Cost of accidents .364
Workforce morale .461
Industrial image .423
Number of accidents .551
Training .478
Feedback .592
Stakeholders’ cooperation .764
Financial resources .806
Safety resources .755
Human resources .718
Workers’ relationships .345
Workers’ involvement .347
Safety responsibilities .371
Safety integration in business goals .587
Safety awareness .586
Safety and productivity alignment .403
Safety standards .620
Safety initiatives .753
A System Dynamics Approach to Construction Safety Culture
237
MODIFICATION INDICES (Group number 1 - Default model)
Covariances: (Group number 1 - Default model)
M.I. Par Change
e26 <--> Lds 6.805 -.066
e4 <--> Goals 4.990 .056
e4 <--> e26 4.799 -.087
e33 <--> Lds 5.023 .058
e32 <--> e1 6.810 -.098
e30 <--> e27 8.256 -.108
e21 <--> e27 4.431 .094
e23 <--> Ppl 5.288 .064
e23 <--> e34 4.272 .085
e18 <--> Lds 4.741 .053
e19 <--> Lds 6.205 -.067
e19 <--> e33 6.141 -.086
e19 <--> e30 4.876 .070
e20 <--> e23 4.663 .079
e14 <--> e17 4.486 .078
e13 <--> e21 6.667 .101
e11 <--> e13 5.968 .081
e9 <--> Ppl 4.076 -.058
e9 <--> Lds 9.345 .081
e9 <--> e3 8.482 .084
e5 <--> e27 7.203 .125
e5 <--> e14 7.143 .111
e6 <--> e26 5.214 -.092
e6 <--> e3 7.911 -.096
e7 <--> Goals 4.205 .040
e7 <--> e30 6.123 .071
e7 <--> e19 4.521 .070
e7 <--> e5 4.177 -.072
e8 <--> e7 9.689 .076
A System Dynamics Approach to Construction Safety Culture
238
Variances: (Group number 1 - Default model)
M.I. Par Change
Regression Weights: (Group number 1 - Default model)
M.I. Par Change
Safety responsibilities <--- Pol 6.088 .181
Means: (Group number 1 - Default model)
M.I. Par Change
Intercepts: (Group number 1 - Default model)
M.I. Par Change
MODEL FIT SUMMARY
CMIN
Model NPAR CMIN DF P CMIN/DF
Default model 87 390.173 237 .000 1.646
Saturated model 324 .000 0
Independence model 48 2022.641 276 .000 7.328
Baseline Comparisons
Model NFI Delta1
RFI rho1
IFI Delta2
TLI rho2 CFI
Default model .807 .775 .914 .898 .912
Saturated model 1.000 1.000 1.000
Independence model .000 .000 .000 .000 .000
A System Dynamics Approach to Construction Safety Culture
239
Parsimony-Adjusted Measures
Model PRATIO PNFI PCFI
Default model .859 .693 .783
Saturated model .000 .000 .000
Independence model 1.000 .000 .000
NCP
Model NCP LO 90 HI 90
Default model 153.173 102.862 211.385
Saturated model .000 .000 .000
Independence model 1746.641 1607.495 1893.216
FMIN
Model FMIN F0 LO 90 HI 90
Default model 3.423 1.344 .902 1.854
Saturated model .000 .000 .000 .000
Independence model 17.742 15.321 14.101 16.607
RMSEA
Model RMSEA LO 90 HI 90 PCLOSE
Default model .075 .062 .088 .002
Independence model .236 .226 .245 .000
AIC
Model AIC BCC
Default model 564.173 613.050
Saturated model 648.000 830.022
Independence model 2118.641 2145.607
A System Dynamics Approach to Construction Safety Culture
240
ECVI
Model ECVI LO 90 HI 90 MECVI
Default model 4.949 4.508 5.460 5.378
Saturated model 5.684 5.684 5.684 7.281
Independence model 18.585 17.364 19.870 18.821
HOELTER
Model HOELTER .05
HOELTER .01
Default model 81 85
Independence model 18 19
A System Dynamics Approach to Construction Safety Culture
241
AAppppeennddiixx 55
SSttrruuccttuurraall MMooddeell RReessuullttss
A System Dynamics Approach to Construction Safety Culture
242
NOTES FOR MODEL (Default model)
Computation of degrees of freedom (Default model)
Number of distinct sample moments: 299
Number of distinct parameters to be estimated: 77
Degrees of freedom (299 - 77): 222
Result (Default model)
Minimum was achieved
Chi-square = 373.328
Degrees of freedom = 222
Probability level = .000
Estimates (Group number 1 - Default model)
Scalar Estimates (Group number 1 - Default model)
Maximum Likelihood Estimates
A System Dynamics Approach to Construction Safety Culture
243
Regression Weights: (Group number 1 - Default model)
Estimate S.E. C.R. P Label
Ppl <--- Lds .910 .148 6.151 ***
Prs <--- Lds .236 .118 2.003 .045
Prs <--- Ppl .893 .100 8.969 ***
Pol <--- Lds .647 .124 5.220 ***
Pol <--- Prs .268 .076 3.535 ***
Pro <--- Ppl .356 .068 5.259 ***
Pro <--- Pol .619 .094 6.589 ***
Goals <--- Pro .662 .108 6.147 ***
Safety initiatives <--- Pol 1.000
Safety standards <--- Pol .929 .087 10.660 ***
Industrial image <--- Goals 1.045 .190 5.509 ***
Workforce morale <--- Goals .978 .170 5.763 ***
Cost of accidents <--- Goals 1.000
Accountability <--- Lds 1.000
Communication <--- Lds 1.071 .114 9.404 ***
Commitment <--- Lds .863 .101 8.512 ***
Safety awareness <--- Pol 1.030 .105 9.816 ***
Safety and productivity alignment <--- Lds .855 .123 6.924 ***
Workers’ relationships <--- Pol .635 .093 6.819 ***
Human resources <--- Ppl 1.000
Stakeholders’ cooperation <--- Ppl .996 .085 11.775 ***
Safety responsibilities <--- Ppl .442 .061 7.193 ***
Financial resources <--- Prs 1.000
Safety resources <--- Prs .996 .074 13.500 ***
Training <--- Prs .606 .068 8.889 ***
Safety integration in business goals <--- Pro .927 .105 8.863 ***
Feedback <--- Pro .928 .100 9.257 ***
Safety documentation <--- Pro 1.000
Number of accidents <--- Goals 1.041 .171 6.074 ***
Benchmarking system <--- Pro .838 .115 7.271 ***
Workers’ involvement <--- Prs .465 .066 7.045 ***
A System Dynamics Approach to Construction Safety Culture
244
Standardized Regression Weights: (Group number 1 - Default model)
Estimate
Ppl <--- Lds .637
Prs <--- Lds .159
Prs <--- Ppl .858
Pol <--- Lds .586
Pol <--- Prs .361
Pro <--- Ppl .458
Pro <--- Pol .615
Goals <--- Pro .904
Safety initiatives <--- Pol .875
Safety standards <--- Pol .798
Industrial image <--- Goals .645
Workforce morale <--- Goals .687
Cost of accidents <--- Goals .599
Accountability <--- Lds .842
Communication <--- Lds .789
Commitment <--- Lds .731
Safety awareness <--- Pol .758
Safety and productivity alignment <--- Lds .621
Workers’ relationships <--- Pol .586
Human resources <--- Ppl .847
Stakeholders’ cooperation <--- Ppl .869
Safety responsibilities <--- Ppl .618
Financial resources <--- Prs .902
Safety resources <--- Prs .868
Training <--- Prs .690
Safety integration in business goals <--- Pro .752
Feedback <--- Pro .778
Safety documentation <--- Pro .788
Number of accidents <--- Goals .743
Benchmarking system <--- Pro .640
Workers’ involvement <--- Prs .587
A System Dynamics Approach to Construction Safety Culture
245
Intercepts: (Group number 1 - Default model)
Estimate S.E. C.R. P Label
Safety initiatives 3.817 .079 48.516 ***
Safety standards 3.800 .080 47.360 ***
Safety and productivity alignment 3.765 .086 43.850 ***
Safety awareness 3.635 .094 38.866 ***
Safety integration in business goals 3.783 .085 44.289 ***
Safety responsibilities 3.757 .064 58.980 ***
Workers’ relationships 3.722 .075 49.843 ***
Human resources 3.722 .105 35.377 ***
Safety resources 3.696 .106 34.742 ***
Financial resources 3.548 .103 34.519 ***
Stakeholders’ cooperation 3.504 .102 34.312 ***
Feedback 3.783 .083 45.747 ***
Training 3.791 .081 46.543 ***
Industrial image 3.739 .082 45.584 ***
Workforce morale 3.774 .072 52.342 ***
Cost of accidents 3.965 .085 46.854 ***
Accountability 4.017 .074 54.223 ***
Communication 3.652 .085 43.111 ***
Commitment 4.157 .074 56.408 ***
Safety documentation 3.913 .088 44.536 ***
Number of accidents 4.096 .071 57.731 ***
Benchmarking system 3.339 .091 36.819 ***
Workers’ involvement 3.730 .073 50.819 ***
A System Dynamics Approach to Construction Safety Culture
246
Variances: (Group number 1 - Default model)
Estimate S.E. C.R. P Label
Lds .443 .084 5.305 ***
ee2 .538 .112 4.801 ***
ee3 .063 .040 1.561 .118
ee1 .123 .033 3.740 ***
ee4 .003 .017 .205 .838
ee5 .053 .025 2.116 .034
e8 .165 .031 5.340 ***
e7 .267 .042 6.399 ***
e6 .517 .074 6.935 ***
e5 .424 .064 6.667 ***
e9 .361 .051 7.025 ***
e11 .286 .040 7.171 ***
e14 .417 .058 7.214 ***
e20 .357 .060 5.906 ***
e19 .318 .053 5.943 ***
e18 .224 .043 5.186 ***
e17 .291 .053 5.514 ***
e23 .308 .045 6.917 ***
e21 .396 .056 7.093 ***
e32 .448 .067 6.725 ***
e33 .313 .048 6.495 ***
e34 .524 .076 6.912 ***
e3 .182 .036 5.094 ***
e2 .309 .052 5.895 ***
e1 .288 .045 6.410 ***
e26 .333 .049 6.863 ***
e30 .257 .042 6.046 ***
e27 .554 .076 7.288 ***
e13 .403 .055 7.288 ***
A System Dynamics Approach to Construction Safety Culture
247
Squared Multiple Correlations: (Group number 1 - Default model)
Estimate
Ppl .406
Prs .936
Pol .773
Pro .994
Goals .818
Workers’ involvement .345
Benchmarking system .409
Number of accidents .552
Safety documentation .621
Commitment .534
Communication .622
Accountability .709
Cost of accidents .358
Workforce morale .473
Industrial image .416
Training .476
Feedback .605
Stakeholders’ cooperation .755
Financial resources .814
Safety resources .754
Human resources .717
Workers’ relationships .343
Safety responsibilities .383
Safety integration in business goals .565
Safety awareness .575
Safety and productivity alignment .385
Safety standards .636
Safety initiatives .766
A System Dynamics Approach to Construction Safety Culture
248
MODIFICATION INDICES (Group number 1 - Default model)
Covariances: (Group number 1 - Default model)
M.I. Par Change ee5 <--> ee2 4.439 .063 ee5 <--> ee1 4.034 -.034
e13 <--> ee1 9.077 .085
e27 <--> ee1 6.662 .085
e27 <--> e13 4.760 .100
e30 <--> e13 4.339 -.069
e30 <--> e27 11.985 -.135
e26 <--> ee1 4.469 .055
e26 <--> e13 5.879 .088
e26 <--> e27 4.941 .094
e1 <--> e30 4.696 .064
e34 <--> ee2 5.979 .142
e33 <--> e2 4.124 .070
e32 <--> e1 6.572 -.097
e21 <--> e13 6.840 .102
e21 <--> e27 4.781 .100
e23 <--> e34 4.596 .087
e17 <--> ee1 4.552 -.057
e17 <--> e13 4.340 -.077
e17 <--> e32 4.410 .085
e18 <--> ee1 8.593 -.070
e18 <--> e26 4.963 -.069
e19 <--> e30 4.789 .071
e19 <--> e33 6.775 -.090
e20 <--> e13 4.068 -.081
e20 <--> e27 4.968 -.104
e14 <--> ee3 4.070 -.064
e14 <--> e17 4.509 .081
e11 <--> ee1 13.404 .087
e11 <--> e13 5.710 .079
e9 <--> ee2 6.823 -.125
e9 <--> e3 7.679 .081
e9 <--> e17 4.922 -.079
A System Dynamics Approach to Construction Safety Culture
249
Covariances: (Group number 1 - Default model) (Cont.)
M.I. Par Change e5 <--> e27 8.034 .138 e5 <--> e14 7.668 .117
e5 <--> e9 4.601 .085
e6 <--> e26 6.102 -.103
e6 <--> e2 4.405 .090
e6 <--> e3 6.688 -.089
e7 <--> e30 5.637 .068
e7 <--> e19 4.832 .072
e7 <--> e5 4.567 -.076
e8 <--> ee5 4.138 -.035
e8 <--> e7 7.310 .065
Variances: (Group number 1 - Default model)
M.I. Par Change
Regression Weights: (Group number 1 - Default model)
M.I. Par Change
Safety responsibilities <--- Pol 5.437 .169
Means: (Group number 1 - Default model)
M.I. Par Change
Intercepts: (Group number 1 - Default model)
M.I. Par Change
A System Dynamics Approach to Construction Safety Culture
250
MODEL FIT SUMMARY
CMIN
Model NPAR CMIN DF P CMIN/DF
Default model 77 373.328 222 .000 1.682
Saturated model 299 .000 0
Independence model 46 1944.668 253 .000 7.686
Baseline Comparisons
Model NFI Delta1
RFI rho1
IFI Delta2
TLI rho2 CFI
Default model .808 .781 .912 .898 .911
Saturated model 1.000 1.000 1.000
Independence model .000 .000 .000 .000 .000
Parsimony-Adjusted Measures
Model PRATIO PNFI PCFI
Default model .877 .709 .799
Saturated model .000 .000 .000
Independence model 1.000 .000 .000
NCP
Model NCP LO 90 HI 90
Default model 151.328 101.949 208.593
Saturated model .000 .000 .000
Independence model 1691.668 1555.060 1835.701
FMIN
Model FMIN F0 LO 90 HI 90
Default model 3.275 1.327 .894 1.830
Saturated model .000 .000 .000 .000
Independence model 17.058 14.839 13.641 16.103
A System Dynamics Approach to Construction Safety Culture
251
RMSEA
Model RMSEA LO 90 HI 90 PCLOSE
Default model .077 .063 .091 .001
Independence model .242 .232 .252 .000
AIC
Model AIC BCC BIC CAIC
Default model 527.328 568.395
Saturated model 598.000 757.467
Independence model 2036.668 2061.202
ECVI
Model ECVI LO 90 HI 90 MECVI
Default model 4.626 4.193 5.128 4.986
Saturated model 5.246 5.246 5.246 6.644
Independence model 17.866 16.667 19.129 18.081
HOELTER
Model HOELTER .05
HOELTER .01
Default model 79 84
Independence model 18 19
A System Dynamics Approach to Construction Safety Culture
252
�
AAppppeennddiixx 66
SSDD EEqquuaattiioonnss ooff BBaassee RRuunn SSiimmuullaattiioonn
A System Dynamics Approach to Construction Safety Culture
253
GOALS(t) = GOALS(t - dt) + (rgoals)*dt
INIT GOALS = 68
Inflows:
rgoals = used_pro*DF_goals_pro
LEADERSHIP(t) = LEADERSHIP(t - dt) + (rlds)*dt
INIT LEADERSHIP = 0
Inflows:
rlds = ((used_lds + ggoals)*rldsf) + (glds*plds)
PARTNERSHIPS_&_RESOURCES(t) = PARTNERSHIPS_&_RESOURCES(t - dt)
+ (rprs)*dt
INIT PARTNERSHIPS_&_RESOURCES = 0
Inflows:
rprs = (used_lds*DF_prs_lds) + (used_ppl*DF_prs_ppl) + (gprs*pprs)
PEOPLE(t) = PEOPLE(t - dt) + (rppl)*dt
INIT PEOPLE = 0
Inflows:
rppl = (used_lds*DF_ppl_lds) + (gppl*pppl)
POLICY_&_STRATEGY(t) = POLICY_&_STRATEGY(t - dt) + (rpol)*dt
INIT POLICY_&_STRATEGY = 0
Inflows:
rpol = (used_lds*DF_pol_lds) + (used_prs*DF_pol_prs) + (gpol*ppol)
A System Dynamics Approach to Construction Safety Culture
254
PROCESSES(t) = PROCESSES(t - dt) + (rpro)*dt
INIT PROCESSES = 0
Inflows:
rpro = (used_ppl*DF_pro_ppl) + (used_pol*DF_pro_pol) + (gpro*ppro)
ENABLERS = used_lds + used_pol + used_ppl + used_prs + used_pro
CSC_INDEX = ENABLERS + used_goals
Co_lds_pol = 0.59
Co_lds_ppl = 0.64
Co_lds_prs = 0.16
Co_pol_pro = 0.62
Co_ppl_pro = 0.46
Co_ppl_prs = 0.86
Co_pro_goals = 0.90
Co_prs_pol = 0.36
DF_goals_pro = (ggoals*Co_pro_goals)/100
DF_pol_lds = (gpol*Co_lds_pol)/100
DF_pol_prs = (gpol*Co_prs_pol)/100
DF_ppl_lds = (gppl*Co_lds_ppl)/100
DF_pro_pol = (gpro*Co_pol_pro)/100
DF_pro_ppl = (gpro*Co_ppl_pro)/100
DF_prs_lds = (gprs*Co_lds_prs)/100
DF_prs_ppl = (gprs*Co_ppl_prs)/100
A System Dynamics Approach to Construction Safety Culture
255
goals = 500
dlds = 100
doll = 80
dppl = 90
dero = 140
dors = 90
ggoals = IF(CSC_INDEX <= 200)THEN(100 - used_goals)ELSE(IF(CSC_INDEX <=
400)THEN(200 - used_goals)ELSE(IF(CSC_INDEX <= 600)THEN(300 - used_goals)
ELSE(IF(CSC_INDEX <= 800)THEN(400 - used_goals)ELSE(500 - used_goals))))
glds = dlds - used_lds
gpol = doll - used_pol
gppl = dppl - used_ppl
gpro = dpro - used_pro
gprs = dprs - used_prs
plds = 0
ppol = 0
pppl = 0
ppro = 0
pprs = 0
used_goals = MIN(GOALS,dgoals)
used_lds = MIN(LEADERSHIP,dlds)
used_pol = MIN(POLICY_&_STRATEGY,dpol)
used_ppl = MIN(PEOPLE,dppl)
used_pro = MIN(PROCESSES,dpro)
A System Dynamics Approach to Construction Safety Culture
256
used_prs = MIN(PARTNERSHIPS_&_RESOURCES,dprs)
rldsf = 0.08
�
A System Dynamics Approach to Construction Safety Culture
257
AAppppeennddiixx 77
LLiinneeaarr RReeggrreessssiioonn RReessuullttss
A System Dynamics Approach to Construction Safety Culture
258
REGRESSION
Descriptive Statistics
Mean Std. Deviation N
Enablers 377.3158 68.14106 114
Goals 392.7632 68.05806 114
Correlations
Enablers Goals
Pearson Correlation Enablers
Goals
1.000
.787
.787
1.000
Sig. (1-tailed) Enablers
Goals
.
.000
.000
.
N Enablers
Goals
114
114
114
114
Variables Entered/Removedb
Model Variables Entered Variables Removed Method
1 Goalsa Enter
a. All requested variables entered. b. Dependent variable: Enablers
Model Summary
Model R R Square Adjusted R Square Std. Error of the Estimate
1 .787 a .619 .615 42.25812
a. Predictors: (Constant), Goals
ANOVAb
Model Sum of Squares df Mean Square F Sig.
1 Regression
Residual
Total
324678.3
200003.8
524682.1
1
112
113
324678.322
1785.748
181.816 .000 a
a. Predictors: (Constant), Goals. b. Dependent variable: Enablers
A System Dynamics Approach to Construction Safety Culture
259
Coefficients a
Model Unstandardized Coefficients Standardized Coefficients t Sig.
B Std. Error Beta
1 (Constant)
Goals
67.974
.788
23.280
.058
.787
2.920
13.484
.004
.000
a. Dependent variable: Enablers
A System Dynamics Approach to Construction Safety Culture
260
AAppppeennddiixx 88
SSeennssiittiivviittyy AAnnaallyysseess RReessuullttss wwhheenn tthhee IInniittiiaall VVaalluuee ooff
eeaacchh EEnnaabblleerr iiss CChhaannggeedd
A System Dynamics Approach to Construction Safety Culture
261
Sensitivity results of the ‘used_pol’ value when the initial values of Pol are changed
Page 3
1.00 5.00 9.00 13.00 17.00
Years
1:
1:
1:
0
40
80
used pol score: 1 - 2 - 3 - 4 -
1
1
1
1
2
2
2
2
3
3
33
4
4
4 4
Note: Numbers 1, 2, 3 and 4 represent the initial value of Policy and Strategy of zero, 20, 40, and 60 points, respectively
Sensitivity results of the CSC index when the initial values of Pol are changed
Page 8
1.00 5.00 9.00 13.00 17.00
Years
1:
1:
1:
0
500
1000
CSC Index Score: 1 - 2 - 3 - 4 -
1
1
1
1
2
2
2
2
3
3
3
3
4
4
44
Note: Numbers 1, 2, 3 and 4 represent the initial value of Policy and Strategy of zero, 20, 40, and 60 points, respectively
A System Dynamics Approach to Construction Safety Culture
262
Sensitivity results of the ‘used_ppl’ value when the initial values of Ppl are changed
Page 4
1.00 5.00 9.00 13.00 17.00
Years
1:
1:
1:
0
45
90
used ppl score: 1 - 2 - 3 - 4 -
1
1
1
1
2
2
2
2
3
3
33
4
4
4 4
Note: Numbers 1, 2, 3 and 4 represent the initial value of People of zero, 22.5, 45, and 67.5 points, respectively
Sensitivity results of the CSC index when the initial values of Ppl are changed
Page 8
1.00 5.00 9.00 13.00 17.00
Years
1:
1:
1:
0
500
1000
CSC Index Score: 1 - 2 - 3 - 4 -
1
1
1
1
2
2
2
2
3
3
33
4
44
4
Note: Numbers 1, 2, 3 and 4 represent the initial value of People of zero, 22.5, 45, and 67.5 points, respectively
A System Dynamics Approach to Construction Safety Culture
263
Sensitivity results of the ‘used_prs’ value when the initial values of Prs are changed
Page 5
1.00 5.00 9.00 13.00 17.00
Years
1:
1:
1:
0
45
90
used prs score: 1 - 2 - 3 - 4 -
1
1
1
1
2
2
2
2
3
3
3 3
4
4
4 4
Note: Numbers 1, 2, 3 and 4 represent the initial value of Partnerships and Resources of zero, 22.5, 45, and 67.5 points, respectively
Sensitivity results of the CSC index when the initial values of Prs are changed
Page 8
1.00 5.00 9.00 13.00 17.00
Years
1:
1:
1:
0
500
1000
CSC Index Score: 1 - 2 - 3 - 4 -
1
1
1
1
2
2
2
2
3
3
33
4
4
44
Note: Numbers 1, 2, 3 and 4 represent the initial value of Partnerships and Resources of zero, 22.5, 45, and 67.5 points, respectively
A System Dynamics Approach to Construction Safety Culture
264
Sensitivity results of the ‘used_pro’ value when the initial values of Pro are
changed
Page 7
1.00 5.00 9.00 13.00 17.00
Years
1:
1:
1:
0
70
140
used pro score: 1 - 2 - 3 - 4 -
1
1
1
1
2
2
2
2
3
3
33
4
4
4 4
Note: Numbers 1, 2, 3 and 4 represent the initial value of Processes of zero, 35, 70, and 105 points, respectively
Sensitivity results of the CSC index when the initial values of Pro are changed
Page 8
1.00 5.00 9.00 13.00 17.00
Years
1:
1:
1:
0
500
1000
CSC Index Score: 1 - 2 - 3 - 4 -
1
1
1
1
2
2
2
2
3
3
3
3
4
4
4
4
Note: Numbers 1, 2, 3 and 4 represent the initial value of Processes of zero, 35, 70, and 105 points, respectively
A System Dynamics Approach to Construction Safety Culture
265
AAppppeennddiixx 99
SSeennssiittiivviittyy AAnnaallyysseess RReessuullttss wwhheenn tthhee EExxttrraa EEffffoorrtt GGiivveenn ttoo
IImmpprroovvee eeaacchh EEnnaabblleerr iiss CChhaannggeedd
A System Dynamics Approach to Construction Safety Culture
266
Sensitivity results of the ‘used_pol’ value when the ‘ppol’ values are changed
Page 3
1.00 5.00 9.00 13.00 17.00
Years
1:
1:
1:
0
40
80
used pol score: 1 - 2 - 3 - 4 -
1
1
1
1
2
2
22
3
3
3 3
4
44 4
Note: Numbers 1, 2, 3 and 4 represent ‘ppol’ of zero, 0.1, 0.2, and 0.3, respectively
Sensitivity results of the CSC index when the ‘ppol’ values are changed
Page 8
1.00 5.00 9.00 13.00 17.00
Years
1:
1:
1:
0
500
1000
CSC Index Score: 1 - 2 - 3 - 4 -
1
1
1
1
2
2
2
2
3
3
3
3
4
4
44
Note: Numbers 1, 2, 3 and 4 represent ‘ppol’ of zero, 0.1, 0.2, and 0.3, respectively
A System Dynamics Approach to Construction Safety Culture
267
Sensitivity results of the ‘used_ppl’ value when the ‘pppl’ values are changed
Page 4
1.00 5.00 9.00 13.00 17.00
Years
1:
1:
1:
0
45
90
used ppl score: 1 - 2 - 3 - 4 -
1
1
1
1
2
2
2
2
3
3
3 3
4
44 4
Note: Numbers 1, 2, 3 and 4 represent ‘pppl’ of zero, 0.1, 0.2, and 0.3, respectively
Sensitivity results of CSC index when ‘pppl’ values are changed
Page 8
1.00 5.00 9.00 13.00 17.00
Years
1:
1:
1:
0
500
1000
CSC Index Score: 1 - 2 - 3 - 4 -
1
1
1
1
2
2
2
2
3
3
33
4
4
44
Note: Numbers 1, 2, 3 and 4 represent ‘pppl’ of zero, 0.1, 0.2, and 0.3, respectively
A System Dynamics Approach to Construction Safety Culture
268
Sensitivity results of the ‘used_prs’ value when the ‘pprs’ values are changed
Page 5
1.00 5.00 9.00 13.00 17.00
Years
1:
1:
1:
0
45
90
used prs score: 1 - 2 - 3 - 4 -
1
1
1
1
2
2
22
3
3
3 3
4
44 4
Note: Numbers 1, 2, 3 and 4 represent ‘pprs’ of zero, 0.1, 0.2, and 0.3, respectively
Sensitivity results of the CSC index when the ‘pprs’ values are changed
Page 8
1.00 5.00 9.00 13.00 17.00
Years
1:
1:
1:
0
500
1000
CSC Index Score: 1 - 2 - 3 - 4 -
1
1
1
1
2
2
2
2
3
3
3
3
4
4
44
Note: Numbers 1, 2, 3 and 4 represent ‘pprs’ of zero, 0.1, 0.2, and 0.3, respectively
A System Dynamics Approach to Construction Safety Culture
269
Sensitivity results of the ‘used_pro’ value when the ‘ppro’ values are changed
Page 7
1.00 5.00 9.00 13.00 17.00
Years
1:
1:
1:
0
70
140
used pro score: 1 - 2 - 3 - 4 -
1
1
1
1
2
2
22
3
3
3 3
4
44 4
Note: Numbers 1, 2, 3 and 4 represent ‘ppro’ of zero, 0.1, 0.2, and 0.3, respectively
Sensitivity results of the CSC index when the ‘ppro’ values are changed
Page 8
1.00 5.00 9.00 13.00 17.00
Years
1:
1:
1:
0
500
1000
CSC Index Score: 1 - 2 - 3 - 4 -
1
1
1
1
2
2
2
2
3
3
3
3
4
4
4
4
Note: Numbers 1, 2, 3 and 4 represent ‘ppro’ of zero, 0.1, 0.2, and 0.3, respectively
A System Dynamics Approach to Construction Safety Culture
270
AAppppeennddiixx 1100
SSDD EEqquuaattiioonnss ooff OOrrggaanniizzaattiioonn ‘‘AA’’
A System Dynamics Approach to Construction Safety Culture
271
GOALS(t) = GOALS(t - dt) + (rgoals)*dt
INIT GOALS = 129
Inflows:
rgoals = used_pro*DF_goals_pro
LEADERSHIP(t) = LEADERSHIP(t - dt) + (rlds)*dt
INIT LEADERSHIP = 20
Inflows:
rlds = ((used_lds + ggoals)*rldsf) + (glds*plds)
PARTNERSHIPS_&_RESOURCES(t) = PARTNERSHIPS_&_RESOURCES(t - dt)
+ (rprs)*dt
INIT PARTNERSHIPS_&_RESOURCES = 18
Inflows:
rprs = (used_lds*DF_prs_lds) + (used_ppl*DF_prs_ppl) + (gprs*pprs)
PEOPLE(t) = PEOPLE(t - dt) + (rppl)*dt
INIT PEOPLE = 43.2
Inflows:
rppl = (used_lds*DF_ppl_lds) + (gppl*pppl)
POLICY_&_STRATEGY(t) = POLICY_&_STRATEGY(t - dt) + (rpol)*dt
INIT POLICY_&_STRATEGY = 19.2
Inflows:
rpol = (used_lds*DF_pol_lds) + (used_prs*DF_pol_prs) + (gpol*ppol)
A System Dynamics Approach to Construction Safety Culture
272
PROCESSES(t) = PROCESSES(t - dt) + (rpro)*dt
INIT PROCESSES = 37.3
Inflows:
rpro = (used_ppl*DF_pro_ppl) + (used_pol*DF_pro_pol) + (gpro*ppro)
ENABLERS = used_lds + used_pol + used_ppl + used_prs + used_pro
CSC_INDEX = ENABLERS + used_goals
Co_lds_pol = 0.59
Co_lds_ppl = 0.64
Co_lds_prs = 0.16
Co_pol_pro = 0.62
Co_ppl_pro = 0.46
Co_ppl_prs = 0.86
Co_pro_goals = 0.90
Co_prs_pol = 0.36
DF_goals_pro = (ggoals*Co_pro_goals)/100
DF_pol_lds = (gpol*Co_lds_pol)/100
DF_pol_prs = (gpol*Co_prs_pol)/100
DF_ppl_lds = (gppl*Co_lds_ppl)/100
DF_pro_pol = (gpro*Co_pol_pro)/100
DF_pro_ppl = (gpro*Co_ppl_pro)/100
DF_prs_lds = (gprs*Co_lds_prs)/100
DF_prs_ppl = (gprs*Co_ppl_prs)/100
A System Dynamics Approach to Construction Safety Culture
273
dgoals = 500
dlds = 100
dpol = 80
dppl = 90
dpro = 140
dprs = 90
ggoals = IF(CSC_INDEX <= 200)THEN(100 - used_goals)ELSE(IF(CSC_INDEX <=
400)THEN(200 - used_goals)ELSE(IF(CSC_INDEX <= 600)THEN(300 - used_goals)
ELSE(IF(CSC_INDEX <= 800)THEN(400 - used_goals)ELSE(500 - used_goals))))
glds = dlds - used_lds
gpol = dpol - used_pol
gppl = dppl - used_ppl
gpro = dpro - used_pro
gprs = dprs - used_prs
plds = 0
ppol = 0
pppl = 0
ppro = 0
pprs = 0
used_goals = MIN(GOALS,dgoals)
used_lds = MIN(LEADERSHIP,dlds)
used_pol = MIN(POLICY_&_STRATEGY,dpol)
used_ppl = MIN(PEOPLE,dppl)
used_pro = MIN(PROCESSES,dpro)
A System Dynamics Approach to Construction Safety Culture
274
used_prs = MIN(PARTNERSHIPS_&_RESOURCES,dprs)
rldsf = 0.08
A System Dynamics Approach to Construction Safety Culture
275
AAppppeennddiixx 1111
SSDD EEqquuaattiioonnss ooff OOrrggaanniizzaattiioonn ‘‘BB’’
A System Dynamics Approach to Construction Safety Culture
276
GOALS(t) = GOALS(t - dt) + (rgoals)*dt
INIT GOALS = 200
Inflows:
rgoals = used_pro*DF_goals_pro
LEADERSHIP(t) = LEADERSHIP(t - dt) + (rlds)*dt
INIT LEADERSHIP = 85
Inflows:
rlds = ((used_lds + ggoals)*rldsf) + (glds*plds)
PARTNERSHIPS_&_RESOURCES(t) = PARTNERSHIPS_&_RESOURCES(t - dt)
+ (rprs)*dt
INIT PARTNERSHIPS_&_RESOURCES = 40.5
Inflows:
rprs = (used_lds*DF_prs_lds) + (used_ppl*DF_prs_ppl) + (gprs*pprs)
PEOPLE(t) = PEOPLE(t - dt) + (rppl)*dt
INIT PEOPLE = 43.2
Inflows:
rppl = (used_lds*DF_ppl_lds) + (gppl*pppl)
POLICY_&_STRATEGY(t) = POLICY_&_STRATEGY(t - dt) + (rpol)*dt
INIT POLICY_&_STRATEGY = 35.2
Inflows:
rpol = (used_lds*DF_pol_lds) + (used_prs*DF_pol_prs) + (gpol*ppol)
A System Dynamics Approach to Construction Safety Culture
277
PROCESSES(t) = PROCESSES(t - dt) + (rpro)*dt
INIT PROCESSES = 56
Inflows:
rpro = (used_ppl*DF_pro_ppl) + (used_pol*DF_pro_pol) + (gpro*ppro)
ENABLERS = used_lds + used_pol + used_ppl + used_prs + used_pro
CSC_INDEX = ENABLERS + used_goals
Co_lds_pol = 0.59
Co_lds_ppl = 0.64
Co_lds_prs = 0.16
Co_pol_pro = 0.62
Co_ppl_pro = 0.46
Co_ppl_prs = 0.86
Co_pro_goals = 0.90
Co_prs_pol = 0.36
DF_goals_pro = (ggoals*Co_pro_goals)/100
DF_pol_lds = (gpol*Co_lds_pol)/100
DF_pol_prs = (gpol*Co_prs_pol)/100
DF_ppl_lds = (gppl*Co_lds_ppl)/100
DF_pro_pol = (gpro*Co_pol_pro)/100
DF_pro_ppl = (gpro*Co_ppl_pro)/100
DF_prs_lds = (gprs*Co_lds_prs)/100
DF_prs_ppl = (gprs*Co_ppl_prs)/100
A System Dynamics Approach to Construction Safety Culture
278
dgoals = 500
dlds = 100
dpol = 80
dppl = 90
dpro = 140
dprs = 90
ggoals = IF(CSC_INDEX <= 200)THEN(100 - used_goals)ELSE(IF(CSC_INDEX <=
400)THEN(200 - used_goals)ELSE(IF(CSC_INDEX <= 600)THEN(300 - used_goals)
ELSE(IF(CSC_INDEX <= 800)THEN(400 - used_goals)ELSE(500 - used_goals))))
glds = dlds - used_lds
gpol = dpol - used_pol
gppl = dppl - used_ppl
gpro = dpro - used_pro
gprs = dprs - used_prs
plds = 0
ppol = 0
pppl = 0
ppro = 0
pprs = 0
used_goals = MIN(GOALS,dgoals)
used_lds = MIN(LEADERSHIP,dlds)
used_pol = MIN(POLICY_&_STRATEGY,dpol)
used_ppl = MIN(PEOPLE,dppl)
used_pro = MIN(PROCESSES,dpro)
A System Dynamics Approach to Construction Safety Culture
279
used_prs = MIN(PARTNERSHIPS_&_RESOURCES,dprs)
rldsf = 0.08
A System Dynamics Approach to Construction Safety Culture
280
AAppppeennddiixx 1122
SSDD EEqquuaattiioonnss ooff tthhee CCyycclliiccaall SSttyyllee ooff SSaaffeettyy MMaannaaggeemmeenntt
A System Dynamics Approach to Construction Safety Culture
281
GOALS(t) = GOALS(t - dt) + (rgoals - rgoals2)*dt
INIT GOALS = 381.64
Inflows:
rgoals = used_pro*DF_goals_pro
Outflows:
rgoals2 = IF (rlds2 > 0) OR (rlds3 > 0) THEN (rgoals + (rgoals*Co_lds_pol*
Co_pol_pro*Co_pro_goals) + (rgoals*Co_lds_ppl*Co_ppl_pro*Co_pro_goals)) ELSE
(0)
LEADERSHIP(t) = LEADERSHIP(t - dt) + (rlds - rlds2 - rlds3)*dt
INIT LEADERSHIP = 60.37
Inflows:
rlds = IF (CSC_INDEX > dCSC_INDEX) OR ((800 < CSC_INDEX < dCSC_INDEX)
AND (slope < 0)) THEN (0) ELSE ((used_lds + ggoals)*rldsf)
Outflows:
rlds2 = IF (CSC_INDEX < 800) OR ((800 < CSC_INDEX < dCSC_INDEX) AND
(slope > 0)) THEN (0) ELSE ((glds + ggoals)*rldsf)
rlds3 = IF (CSC_INDEX > dCSC_INDEX) AND (rlds = 0) AND (rlds2 = 0) THEN
((glds + ggoals)*rldsf) ELSE (0)
PARTNERSHIPS_&_RESOURCES(t) = PARTNERSHIPS_&_RESOURCES(t - dt)
+ (rprs - rprs2)*dt
INIT PARTNERSHIPS_&_RESOURCES = 82.23
Inflows:
rprs = (used_lds*DF_prs_lds) + (used_ppl*DF_prs_ppl)
Outflows:
rprs2 = IF (rlds2 > 0) OR (rlds3 > 0) THEN (rprs + (rprs*Co_lds_prs) +
(rprs*Co_lds_ppl *Co_ppl_prs)) ELSE (0)
A System Dynamics Approach to Construction Safety Culture
282
PEOPLE(t) = PEOPLE(t - dt) + (rppl - rppl2)*dt
INIT PEOPLE = 68.07
Inflows:
rppl = (used_lds*DF_ppl_lds)
Outflows:
rppl2 = IF (rlds2 > 0) OR (rlds3 > 0) THEN (rppl + (rppl*Co_lds_ppl)) ELSE (0)
POLICY_&_STRATEGY(t) = POLICY_&_STRATEGY(t - dt) + (rpol - rpol2)*dt
INIT POLICY_&_STRATEGY = 73.17
Inflows:
rpol = (used_lds*DF_pol_lds) + (used_prs*DF_pol_prs)
Outflows:
rpol2 = IF (rlds2 > 0) OR (rlds3 > 0) THEN (rpol + (rpol*Co_lds_pol) +
(rpol*Co_lds_prs *Co_prs_pol)) ELSE (0)
PROCESSES(t) = PROCESSES(t - dt) + (rpro - rpro2)*dt
INIT PROCESSES = 132.44
Inflows:
rpro = (used_ppl*DF_pro_ppl) + (used_pol*DF_pro_pol)
Outflows:
rpro2 = IF (rlds2 > 0) OR (rlds3 > 0) THEN (rpro + (rpro*Co_lds_pol*Co_pol_pro) +
(rpro*Co_lds_ppl*Co_ppl_pro)) ELSE (0)
desired_CSC_INDEX(t) = desired_CSC_INDEX(t - dt) + (CSC_flow)*dt
INIT desired_CSC_INDEX = 950
A System Dynamics Approach to Construction Safety Culture
283
Inflows:
CSC_flow = PULSE (10,(slope < 0) AND (CSC_INDEX >= desired_CSC_INDEX),
(CSC_INDEX >= desired_CSC_INDEX))
ENABLERS = used_lds + used_pol + used_ppl + used_prs + used_pro
CSC_INDEX = ENABLERS + used_goals
Co_lds_pol = 0.59
Co_lds_ppl = 0.64
Co_lds_prs = 0.16
Co_pol_pro = 0.62
Co_ppl_pro = 0.46
Co_ppl_prs = 0.86
Co_pro_goals = 0.90
Co_prs_pol = 0.36
DF_goals_pro = (ggoals*Co_pro_goals)/100
DF_pol_lds = (gpol*Co_lds_pol)/100
DF_pol_prs = (gpol*Co_prs_pol)/100
DF_ppl_lds = (gppl*Co_lds_ppl)/100
DF_pro_pol = (gpro*Co_pol_pro)/100
DF_pro_ppl = (gpro*Co_ppl_pro)/100
DF_prs_lds = (gprs*Co_lds_prs)/100
DF_prs_ppl = (gprs*Co_ppl_prs)/100
A System Dynamics Approach to Construction Safety Culture
284
dgoals = 500
dlds = 100
dpol = 80
dppl = 90
dpro = 140
dprs = 90
ggoals = IF(CSC_INDEX <= 200)THEN(100 - used_goals)ELSE(IF(CSC_INDEX <=
400)THEN(200 - used_goals)ELSE(IF(CSC_INDEX <= 600)THEN(300 - used_goals)
ELSE(IF(CSC_INDEX <= 800)THEN(400 - used_goals)ELSE(500 - used_goals))))
glds = dlds - used_lds
gpol = dpol - used_pol
gppl = dppl - used_ppl
gpro = dpro - used_pro
gprs = dprs - used_prs
dCSC_INDEX = MIN(desired_CSC_INDEX,1000)
used_goals = MIN(GOALS,dgoals)
used_lds = MIN(LEADERSHIP,dlds)
used_pol = MIN(POLICY_&_STRATEGY,dpol)
used_ppl = MIN(PEOPLE,dppl)
used_pro = MIN(PROCESSES,dpro)
used_prs = MIN(PARTNERSHIPS_&_RESOURCES,dprs)
rldsf = 0.08
slope = DERIVN(CSC_INDEX,1)
A System Dynamics Approach to Construction Safety Culture
285
RREEFFEERREENNCCEESS
ACSNI, 1993. Study group on human factors. Third report: organizing for safety.
Advisory Committee on the Safety of Nuclear Installations, London: HMSO.
Ahmed, A.M., Yang, J.B. and Dale, B.G., 2003. Self-assessment methodology: the
route to business excellence. The quality management journal, 10(1), 43-57.
Ahmed, S.M., Tang, P., Azhar, S. and Ahmad, I., 2004. An evaluation of safety
measures in the Hong Kong construction industry based on total quality management
principals [online]. Available from: http://www.fiu.edu/~sazha002/ research/tqmpaper.
pdf [Accessed 9 September 2004].
Aksorn, T., and Hadikusumo, B.H.W., 2004. Investigating unsafe acts and decision to
err factors of Thai construction workers. CIB international symposium on globalisation
and construction, meeting the challenges, reaping the benefits, 17-19, November 2004,
Thailand.
Aksorn, T. and Hadikusumo, B.H.W., 2006. Critical success factors of safety programs
implementation in Thai construction projects. In: D. Fang, R.M. Choudhry and J.W.
Hinze, eds. Proceedings of the CIB W99 2006 international conference on global unity
for safety and health in construction, 28-30 June 2006, Beijing, China. Beijing:
Tsinghua University Press, 328-336.
Aksorn, T. and Hadikusumo, B.H.W., 2007. Critical success factors influencing safety
program performance in Thai construction projects. Safety science, In press, Corrected
proof, Available online 6 August 2007.
A System Dynamics Approach to Construction Safety Culture
286
Ali, T.H., 2006. Influence of national culture on construction safety climate in Pakistan.
Thesis (PhD), School of Engineering, Griffith University, Australia.
Arboleda, A., Morrow, P.C., Crum, M.R., and Shelley II, M.C., 2003. Management
practices as antecedents of safety culture within the trucking industry: similarities and
differences by hierarchical level. Journal of safety research, 34, 189-197.
Australian Bureau of Statistics, 1993. Australian and New Zealand standard industrial
classification 1993 ANZSIC, Australia, Cat No. 1292.0.
Barnes, J., Gaskin, J., Park, T., Saengdaeng, S. and Wilson, A., 2005. System dynamics
and its use in an organization [online]. Available from: http://www.mngt.waikato.ac.nz/
depts/mnss/courses/511/Ass1pdf/99Gp1.pdf [Accessed 14 March 2005].
Berg, K.S., 2006. Synergy for success effective construction safety and health
management. In: D. Fang, R.M. Choudhry and J.W. Hinze, eds. Proceedings of the CIB
W99 2006 international conference on global unity for safety and health in
construction, 28-30 June 2006, Beijing, China. Beijing: Tsinghua University Press, 34-
37.
Blockley, D., 1995. Process re-engineering for safety. Proceedings of risk engineering
and management in civil, mechanical and structural engineering. ICE, London, 51-66.
Boonrod, C., Kittiwisit, C., and Laokhongthavorn, L., 1998. Feasibility study of TIS
18000 for construction industry. Paper presented at the 6thnational civil engineering
conference, 10-12 May, 1998, Petchaburi, Thailand, 1-6.
Byrne, B.M., 2001. Structural equation modelling with AMOS: basic concepts,
applications and programming. Mahwah, NJ: Lawrence Erlbaum Associations, Inc.
A System Dynamics Approach to Construction Safety Culture
287
Camison, C., 1996. Total quality management in hospitality: an application of the
EFQM model. Tourism management, 17(3), 191-201.
Caravatta, M., 1997. Conducting an organizational self-assessment using the 1997
Baldrige award criteria. Quality progress, 30(10), 87-91.
Chritamara, S. and Ogunlana, S.O., 2002. System dynamic modelling of design and
build construction projects. Construction innovation, 2, 269-295.
Clarke, S., 1999. Perceptions of organizational safety: implications for the development
of safety culture. Journal of organizational behaviour, 20(2), 185-198.
Clissold, G., 2004. Understanding safety performance using safety climate and
psychological climate. Working Paper 65/04, Department of Management, Monash
University, Australia.
Coakes, S.J. and Steed, L.G., 2003. SPSS analysis without anguish version 11.0 for
Windows. Australia: John Wiley & Sons.
Cohen, A., 1977. Factors in successful occupational safety programs. Journal of safety
research, 9, 168-178.
Cohen, J.M., 2002. Measuring safety performance in construction. Occupational
hazards, 64(6), 41-44.
Cooper, M.D., 2000. Towards a model of safety culture. Safety science, 36, 111-136.
A System Dynamics Approach to Construction Safety Culture
288
Cooper, M.D., 2002. Safety culture: a model for understanding and quantifying a
difficult concept. Professional safety, 47(6), 30-35.
Cortada, J.W. and Woods, J.A., 1994. The Quality year 1994. New York: McGraw-Hill.
Curran, P.J., West, S.G., and Finch, J.F., 1996. The robustness of test statistics to
nonnormality and specification error in confirmatory factor analysis. Psychological
methods, 1(1), 16-29.
Dale, B.G. and Smith, M., 1997. Spectrum of quality management implementation grid:
development and use. Managing service quality, 7(6), 307-311.
Damodaran, R., 2006. Safety in construction organizations [online]. Property bytes: the
Indian real estate blog. Available from: http://propertybytes.com/?p=193 [Accessed 20
February 2007].
Dane, F., 1990. Research methods. California, U.S.A.: Brooks/Cole Publishing
Company.
Dedobbeleer, N. and Beland, F., 1991. A safety climate measure for construction sites.
Journal of safety research, 22, 97-103.
Dias, L.M.A. and Coble, R.J., 1996. Implementation of safety and health on
construction sites. Rotterdam, Netherlands: A.A. Balkema.
Dunlap, S., 2004. The role of management: influencing a complete safety culture
[online]. Available from: http://www.geaps.com/proceedings/2004/Dunlap.cfm
[Accessed 2 July 2004].
A System Dynamics Approach to Construction Safety Culture
289
Duynisveld, J.L., 1999. A dynamic model of the N flows on a dairy farm. Thesis
(Master). Nova Scotia Agricultural College, Truro, Nova Scotia, Canada.
Eberlein, B., 2007. System dynamics software info [online]. Available from:
http://www.vensim.com/sdmail/sdsoft.html [Accessed 14 March 2007].
EFQM, 1998. Self-assessment: guidelines for companies, Brussels, Belgium: The
European Foundation for Quality Management, EFQM.
EFQM, 2000. Introducing excellence, Brussels, Belgium: The European Foundation for
Quality Management, EFQM.
Embassy of Denmark, Bangkok, 2006. Establishment of a regional workers’ institute
for occupational health, safety, and environment project [online]. Available from:
http://www.ambbangkok.um.dk/en/menu/DevelopmentCooperation/ProgrammeCompon
ents/UrbanEnvironmentalManagement/EstablishmentofaRegionalWorkersInstituteforOc
cupationalHealthSafetyandEnvironmentProject.htm [Accessed 26 September 2007].
Erdos, P. L., 1970. Professional mail surveys. New York: McGraw-Hill.
Eskildsen, J.K. and Dahlgaard, J.J., 2000. A causal model for employee satisfaction.
Total quality management, 11(8), 1081-1094.
Fang, D., Chen, Y. and Wong, L., 2006. Safety climate in construction industry: a case
study in Hong Kong. Journal of construction engineering and management, June, 573-
584.
Flannery, J.A., 2001. Safety culture and its measurement in aviation. Thesis (Master of
Aviation Management), University of Newcastle, Australia.
A System Dynamics Approach to Construction Safety Culture
290
Flin, R. and Mearns, K.J., 1999. Assessing the state of organizational safety: culture or
climate? Current psychology, 18(1), 13-17.
Ford, A., 1999. Modelling the environment: an introduction to system dynamic
modelling of environment systems, USA: Island Press.
Forrester, J.W., 1961. Industrial dynamics, Cambridge, MA: MIT Press.
Forrester, J.W., 1985. Industrial dynamics, Cambridge, MA: MIT Press.
Forrester, J.W. and Senge, P.M., 1980. Tests for building confidence in system dynamic
models. TIMS studies in management science, 14, 209-228.
Fung, I.W.H., Tam, C.M., Tung, K.C.F. and Man, D.S.K., 2005. Safety cultural
divergences among management, supervisory and worker groups in Hong Kong
construction industry. International journal of project management, 23(7), 495-572.
Gadd, S. and Collins, A.M., 2002. Safety culture: a review of the literature. Report by
the Health & Safety Laboratory, UK: Sheffield University.
Garson, G.D., 2006. Structural equation modelling [online]. Available from:
http://www2.chass.ncsu.edu/garson/pa765/structur.htm [Accessed 4 August 2006].
Geller, E.S., 2000. 10 leadership qualities for a total safety culture. Professional safety,
45(5), 38-41.
Geller, E.S., 2001. A total safety culture: from a corporate achievement to a global
vision. Behaviour and social issues. 11(1), 18-20.
A System Dynamics Approach to Construction Safety Culture
291
Gibb, A.G.F. and Foster, M., 1996. Safety motivation: evaluation of incentive schemes.
In: L.M.A. Dias and R.J. Coble, eds. Proceedings of the first international conference of
CIB working commission, 4-7 September 1996, Portugal. Rotterdam. Netherlands: A.A.
Balkema, 405-415.
Gillen, M., Baltz, D., Gassel, M., Kirsch, L., and Vaccaro, D., 2002. Perceived safety
climate, job demands, and co-worker support among union and non-union injured
construction workers. Journal of safety research, 33, 33-51.
Glendon, A.I. and Litherland, D.K., 2001. Safety climate factors, group differences and
safety behaviour in road construction. Safety science, 39, 157-188.
Goetsch, D.L., 2003. Construction safety and health, New Jersey: Pearson Education,
Inc.
Grote, G. and Kunzler, C., 2000. Diagnosis of safety culture in safety management
audits. Safety science, 34, 131-150.
Guldenmund, F.W., 2000. The nature of safety culture: a review of theory and research.
Safety science, 34(1-3), 215-257.
Hair, J.F., Anderson, R.E., Tatham, R.L. and Black, W.C., 1998. Multivariate data
analysis, 5th edition, U.S.A.: Prentice-Hall International.
Hale, A.R., 2000. Culture’ s confusions. Safety science, 34(1-3), 1-14.
Harvey, J., Erdos, G., Bolam, H., Cox, M.A.A., Kennedy, J.N.P. and Gregory, D.T.,
2002. An analysis of safety culture attitudes in a highly regulated environment. Work &
stress, 16(1), 18-36.
A System Dynamics Approach to Construction Safety Culture
292
Hinze, J.W. and Raboud, P., 1988. Safety on large building construction projects.
Journal of construction engineering and management, 114(2), 286-293.
Ho, J.K.L. and Zeta, K.C., 2004. Cultural factors and their significance to the Hong
Kong construction industry [online]. Available from: http://www.ic.ployu.edu.hk/
esh/kb/culture/Ho&Zeta.pdf [Accessed 20 September 2004].
Hofstede, G., 1980. Framework for describing cultural differences [online]. Available
from: http://www.changecommblog.com/2006/09/26/framework-for-describing-
cultural-differences/ [Accessed 17 January 2008].
HSC, 2003. Health and safety statistics highlights 2002/03, HSE Books, Health and
Safety Commission, Sudbury, Suffolk.
Hsu, C.H., 2002. A structural equation modelling analysis of transformational
leadership, organizational culture and organizational effectiveness in Taiwanese
sport/fitness organizations. Thesis (PhD), United Stated Sports Academy, Alabama,
USA.
Hudson, P., 2001. Safety management and safety culture: the long and winding road.
Presented to CASA, September, 10, 2001, Canberra.
Huy, V.V., 2002. A multiple perspectives approach to organizational problem
formulation: two case studies. Thesis (PhD), Information and Operations Management,
Texas A&M University, USA.
ICAO, 1992. Human factors digest no.10: human factors, management and
organization. Montreal, Canada: International Civil Aviation Organization, ICAO.
A System Dynamics Approach to Construction Safety Culture
293
INEEL, 2004. Idaho national engineering and environmental laboratory voluntary
protection program: total safety culture [online]. Available from: http://www.inel.gov/
vpp/total-safety-culture.shtml [Accessed 20 December 2004].
International Labour Organization, 2005. Thailand – occupational safety and health in
the construction industry [online]. Available from:http://www.ilo.org/public/english/
region/asro/bangkok/download/background/osh/conth05.pdf [Accessed 13 August
2007].
Ithink, 2003. Software reference guide: Ithink 8 technical documentation. USA: Isee
systems, Inc.
Jaafari, A., 1996. Human factors in the Australian construction industry: towards total
quality management. Australian journal of management, 21(2), 159-186.
Jackson, S.L., 2003. Research methods and statistics a critical thinking approach.
U.S.A.: Thomson Wadsworth.
Jaselskis, E.J., Anderson, S.D. and Russell, J.S., 1996. Strategies for achieving
excellence in construction safety performance. Journal of construction engineering and
management, March, 61-70.
Jones, M., 2007. Risks to high consequence [online]. Available at: http://www.awe.
co.uk/Images/Discovery%2013%20article%2002_tcm6-4325.pdf [Accessed 8
September 2007].
Joreskog, K. and Sorbom, D., 1993. LISREL 8: structural equation modelling with the
SIMPLIS command language. Chicago: SSI Inc.
A System Dynamics Approach to Construction Safety Culture
294
Kartam, N., 1997. Integrating safety and health performance into construction CPM.
Journal of construction engineering and management, 123(2), 121-126.
Kartam, N.A., Flood, I. and Koushki, P., 2000. Construction safety in Kuwait: issues,
procedures, problems and recommendations. Safety science, 36, 163-184.
Khanna, V.K., Vrat, P., Shankar, R., and Sahay, B.S., 2004. Managing the transition
phases in the TQM journey: a system dynamics approach. International journal of
quality & reliability management, 21(5), 518-544.
Kline, R.B., 2005. Principles and practice of structural equation modelling, 2nd edition,
New York: Guilford Press.
Kristensen, K. and Juhl, H.J., 1999. Beyond the bottom line-measuring stakeholder
value. In: B., Edvardsson, and A., Gustafsson, eds. The Nordic School of Quality
Management. Studentlitteratur: Lund.
Kumar, R., 2005. Research methodology: a step-by-step guide for beginners, 2nd
edition, London: Thousand Oaks.
Lamotte, G. and Carter, G., 2000. Are the balanced scorecard and the EFQM
excellence model mutually exclusive or do they work together to bring added value to a
company? Prepared for the EFQM common interest day, March 17, 2000, UK:
Balanced Scorecard Collaborative Europe.
Langford, D., Rowlinson, S. and Sawacha, E., 2000. Safety behaviour and safety
management: its influence on the attitudes of workers in the UK construction industry.
Engineering, construction and architectural management, 7(2), 133-140.
A System Dynamics Approach to Construction Safety Culture
295
Lardner, R., Fleming, M. and Joyner, P., 2001. Towards a mature safety culture.
IChemE symposium series, 148, 635-642.
Lee, T., 1998. Assessment of safety culture at a nuclear reprocessing plant. Work &
stress, 12(3), 217-237.
Lingard, H. and Blismas, N., 2006. Building a safety culture: the importance of shared
mental models in the Australian construction industry. In: D. Fang, R.M. Choudhry and
J.W. Hinze, eds. Proceedings of the CIB W99 2006 international conference on global
unity for safety and health in construction, 28-30 June 2006, Beijing, China. Beijing:
Tsinghua University Press, 201-208.
Little, A.D., 2002. Improving safety culture in the construction industry. A workshop
for senior management in construction contracting and client companies. Cambridge:
University Press.
Love, P.E.D., Mandal, P., Smith, J. and Li, H., 2000. Modelling the dynamics of design
error induced rework in construction. Construction management and economics, 18,
567-574.
Maloney, W.F., 2003. Employee involvement, consultation and information sharing in
health and safety in construction. Report submitted on the work performed under
engineering physical science research Council. University of Kentucky and Glasgow
Caledonian University.
Mbuya, E. and Lema, N.M., 2004. Towards development of a framework for integration
of safety and quality management techniques in construction project delivery process
[online]. Available from: http://buildnet.csir.co.za/w107/Authors/Accepted% 20Papers/
012p%20-%20final.doc [Accessed 5 May 2004].
A System Dynamics Approach to Construction Safety Culture
296
McBurney, D.H., 1994. Research methods. Pacific Grove, California, U.S.A.:
Brooks/Cole Publishing Company.
McDougall, M., 2004. Developing a positive safety culture. Environment, Health and
Safety Office, University of California, San Diego, U.S.A.
McLucas, A.C., 2005. System dynamics applications: a modular approach to modelling
complex world behaviour. Canberra, Australia: Argos Press.
Melville, S. and Goddard, W., 1996. Research methodology an introduction for science
& engineering students. Kenwyn: Juta & Co., Ltd.
Miller, N.G., 1990. A methodology for introducing technology into organizations.
Thesis (PhD), University of New South Wales, Australia.
Mohamed, S., 1999. Empirical investigation of construction safety management
activities and performance in Australia. Safety science, 33, 129-142.
Mohamed, S., 2002. Safety climate in construction site environments. Journal of
construction engineering and management, 128(5), 375-384.
Mohamed, S., 2003. Scorecard approach to benchmarking organizational safety culture
in construction. Journal of construction engineering and management, 129(1), 80-88.
Molenaar, K., Brown, H., Caile, S. and Smith, R., 2002. Corporate culture: a study of
firms with outstanding construction safety. Professional safety, July, 18-27.
A System Dynamics Approach to Construction Safety Culture
297
Morecroft, J.D.W., 1988. System dynamics and microworlds for policymakers.
European journal of operational research, 35, 301-320.
Morgan, G.A. and Griego, O.V., 1998. Easy use and interpretation of SPSS for
Windows. USA: Lawrence Erlbaum Associates.
Mueller, R.O. and Hancock, G.R., 2004. Factor analysis and latent structure,
confirmatory. International Encyclopaedia of the Social & Behavioural Sciences, 5239-
5244.
Niskanen, T., 1994. Safety climate in the road administration. Safety science, 17, 237-
255.
NIST, 1993. 1993 award criteria, Malcolm Baldrige National Quality Award.
Gaithersburg, MD: National Institute of Standards and Technology, NIST.
Northouse, P.G., 1997. Leadership: theory and practice. Thousand Oaks, Calif: Sage
Publications.
NOSHC, 2005. Regulation impact statement, National Occupational Health and Safety
Commission, Sydney, Australia.
NPS Risk Management Division, 2006. Occupational safety and health overview for
NPS employees [online]. Available from: http://www.nps.gov/training/tel/Guides/
OSH_emp_ pguide_2006_0605.pdf [Accessed 3 August 2007].
O’ Dea, A. and Flin, R., 2001. Site manager and safety leadership in the offshore oil and
gas industry. Safety science, 37, 39-57.
A System Dynamics Approach to Construction Safety Culture
298
Ogunlana, S., Promkuntong, K., and Jearkjirm, V., 1996. Construction delays in a fast-
growing economy: comparing Thailand with other economies. International journal of
project management, 14(1), 37-45.
Ogunlana, S., Siddiqui, Z., Yisa, S., and Olomolaiye, P., 2002. Factors and procedures
used in matching project managers to construction projects in Bangkok. International
journal of project management, 20(5), 385-400.
Oklahoma Department of Labour, 1998. Essential elements of an effective safety &
health program [online]. Available from: http://www.state.ok.us/~okdol/peosh/stee.pdf
[Accessed 29 January 2007].
Olcott, J.W., 1997. Characteristics of safety cultures. Flight Safety Foundation
Corporate Aviation Safety Seminar, Phoenix, AZ, May 1, 1997.
Otley, D.T., 1999. Performance management: a framework for management control
system research, Management accounting research, 10, 363-382.
Packendorff, J., 1995. Inquiring into the temporary organization: new directions for
project management research. Scandinavian journal of management, 11(4), 319-333.
Pallant, J., 2005. SPSS survival manual: a step by step guide to data analysis using
SPSS for windows (version 12), New South Wales: Allen & Unwin.
Pannirselvam, G.P. and Ferguson, L.A., 2001. A study of the relationships between the
Baldrige categories. International journal of quality & reliability management, 18(1),
14-34.
A System Dynamics Approach to Construction Safety Culture
299
Pasman, H.J., 2000. Risk informed resource allocation policy: safety can save costs.
Journal of hazardous materials, 71, 375-394.
Paul, P.S. and Maiti, J., 2007. The role of behavioural factors on safety management in
underground mines. Safety science, 45(4), 449-471.
Pipitsupaphol, T. and Watanabe, T., 2000. Identification of root causes of labour
accidents in the construction industry. Proceedings of the 4th Asia Pacific Structural
Engineering and Construction Conference, Malaysia, pp. 93-98.
Potter, D.L., 2003. Research report organizational culture and safety: integrating for a
safe workplace [online]. Available from: http://www.debpotter.com/admin/files/files/
Organization%20safety%20and%20culture.pdf [Accessed 14 July 2004].
Ratikin, S.R., 2001. Software verification and validation for practitioners and
managers. Norwood : Artech House.
Rodrigues, A.G. and Bowers, J., 1996. The role of system dynamics in project
management. International journal of project management, 14(4), 213-220.
Rodrigues, A.G. and Williams, T.M., 1998. System dynamics in project management:
assessing the impacts of client behaviour on project performance. Journal of the
operational research society, 49(1), 2-15.
Rosenfeld, Y., Rozenfeld, O., Sacks, R., and Baum, H., 2006. Efficient and timely use
of safety resources in construction. In: D. Fang, R.M. Choudhry and J.W. Hinze, eds.
Proceedings of the CIB W99 2006 international conference on global unity for safety
and health in construction, 28-30 June 2006, Beijing, China. Beijing: Tsinghua
University Press, 290-297.
A System Dynamics Approach to Construction Safety Culture
300
Saeed, K. and Brooke, K., 1996. Contract design for profitability in macro-engineering
projects. System dynamics review, 12(3), 235-246.
Sarshar, M., Haigh, R., Finnemore, M., Aouad, G., Barrett, P., Baldry, M. and Sexton,
M., 2000. SPICE: a business process diagnostic tool for construction projects.
Engineering, construction and architectural management, 7(3), 241-250.
Sawacha, E., Naoum, S. and Fong, D., 1999. Factors affecting safety performance on
construction sites. International journal of project management, 17(5), 309-315.
Selvanathan, S. and Selvanathan, S., 2005. Sampling methods and survey &
questionnaire design. RHD statistics and research design support (STARDS) unit,
Griffith University, Gold Coast, Australia.
Seo, D.C., Torabi, M.R., Blair, E.H., and Ellis, N.T., 2004. A cross-validation of safety
climate scale using confirmatory factor analytic approach. Journal of safety research,
35, 427-445.
Sheffield Hallam University, 2003. Linking the EFQM excellence model to other
management models and tools. Centre for Integral Excellence, Sheffield Hallam
University, Howard, Sheffield, UK.
Simonovic, S., 2005. System simulation [online]. The University of Western Ontario,
Canada. Available from: http://www.engga.uwo.ca/research/iclr/simonovic/ES566/
Lecture4.pdf [Accessed 22 March 2005].
Siu, O., Phillips, D.R. and Leung, T., 2004. Safety climate and safety performance
among construction workers in Hong Kong: the role of psychological strains as
mediators. Accident analysis and prevention, 36, 359-366.
A System Dynamics Approach to Construction Safety Culture
301
Smith, G.R. and Roth, R.D., 1991. Safety programs and the construction manager.
Journal of construction engineering and management, 117(2), 360-371.
Sorensen, J.N., 2002. Safety culture: a survey of the state-of-the-art. Reliability
engineering and system safety, 76, 189-204.
Speirs, F. and Johnson, C.W., 2002. Safety culture in the face of industrial change: a
case study from the UK rail industry. Research Report, University of Glasgow,
Glasgow, Scotland, May 29, 2002.
Sterman, J.D., 1992. System dynamic modelling for project management. Cambridge:
MIT.
Suraji, A., Duff, R., and Peckitt, S.J., 2001. Development of causal model of
construction accident causation. Journal of construction engineering and management,
July/ August, 337-344.
Tabachnick, B.G. and Fidell, L.S., 2007. Using Multivariate Statistics. 5th edition, USA:
Pearson Education, Inc.
Tam, C.M., Zeng, S.X. and Deng, Z.M., 2004. Identifying elements of poor
construction safety management in China. Safety science, 42(7), 569-586.
Tang, S.L., De Saram D.D., Wang, Z.M. and Zhang, T.Q., 2003. Costs of construction
accidents in social and humanity context. Proceedings of the 9th east-pacific conference
on structural engineering & construction, Indonesia. Paper number: 1433.
Tang, Y.H. and Ogunlana, S.O., 2003a. Modelling the dynamic performance of a
construction organization. Construction management and economics, 21, 127-136.
A System Dynamics Approach to Construction Safety Culture
302
Tang, Y.H. and Ogunlana, S.O., 2003b. Selecting superior performance improvement
policies. Construction management and economics, 21, 247-256.
Taylor, R.H., 2003. Managing risks through developing a strong safety culture. Nuclear
energy, 42(6), 341-346.
Teo, A.L. and Fang, D., 2006. Measurement of safety climate in construction industry:
studies in Singapore and Hong Kong. In: D. Fang, R.M. Choudhry and J.W. Hinze, eds.
Proceedings of the CIB W99 2006 international conference on global unity for safety
and health in construction, 28-30 June 2006, Beijing, China. Beijing: Tsinghua
University Press, 157-164.
Teo, E.A.L., Ling, F.Y.Y. and Chong, A.F.W., 2005. Framework for project managers
to manage construction safety. International journal of project management, 23(4), 329-
341.
The UAE Ministry of Labour and Social Affairs, 2001. The reality of occupational
health and safety in UAE 2000-2001. Dubai, UAE: The Department of Research and
Studies, Ministry of Labour and Social Affairs.
Tummala, V.M.R. and Tang, C.L., 1995. Strategic quality management, Malcolm
Baldrige and European quality awards and ISO 9000 certification: core concepts and
comparative analysis. 1994-95 Annual Issue of IIE, Hong Kong.
Turner, B.A., 1991. The development of a safety culture. Chemistry and industry, 7,
241-247.
A System Dynamics Approach to Construction Safety Culture
303
Ullman, J.B., 2001. Structural equation modelling. In: B.G. Tabachnick, and L.S D.
Fidell. Using Multivariate Statistics. 5th edition. Needham Heights, MA: Allyn &
Bacon, 157-164.
Vecchio-Sadus, A.M. and Griffiths, S., 2004. Marketing strategies for enhancing safety
culture. Safety science, 42(2), 601-619.
Vennix, J.A.M., 1996. Group model-building: facilitating team learning using system
dynamics England: John Wiley & Sons.
Ventana Systems, Inc., 2001. System behaviour and causal loop diagrams [online].
Available from: http://www.public.asu.edu/~kirkwood/sysdyn/SDIntro/ch-1.pdf
[Accessed 24 July 2007].
Wallace, A. and Neal, A., 2000. A report on safety in the Queensland meat industry.
Report prepared for the Meat Industry Advisory Group and the Australian Meat
Industry Employees Union, Australia.
Wangniwetkul, K., 2007. Use of personal safety equipment in construction [online].
Available from: http://www.eit.or.th/article/data/01040026.pdf [Accessed 24 September
2007].
Williamson, A.M., Feyer, A., Cairns, D. and Biancotti, D., 1997. The development of a
measure of safety climate: the role of safety perceptions and attitudes. Safety
science, 25(1-3), 15-27.
Wong, P.S.P. and Cheung, S.O., 2005. Structural equation model of trust and partnering
success. Journal of management in engineering, April, 70-80.
A System Dynamics Approach to Construction Safety Culture
304
Wongrassamee, S., Gardiner, P.D. and Simmons, J.E.L., 2003. Performance
measurement tools: the balanced scorecard and the EFQM excellence model. Measuring
business excellence, 7(1), 14-29.
Wright, M.S., Brabazon, P., Tipping, A. and Talwalkar, M. 1999. Development of a
business excellence model of safety culture: safety culture improvement matrix. London:
Entec UK Ltd.
Zohar, D., 1980. Safety climate in industrial organizations: theoretical and applied
implications. Journal of applied psychology, 65(1), 96-102.