Safety Science 98 (2017) 124–135

Contents lists available at ScienceDirect

Safety Science
journal homepage: www.elsevier.com/locate/ssci

Structural equation model of integrated safety intervention practices

affecting the safety behaviour of workers in the construction industry
Mohammad Mazlina Zaira a,b,⇑, Bonaventura H.W. Hadikusumo a
a
School of Engineering and Technology, Asian Institute of Technology, P.O. Box 4, Klong Luang, Pathumthani 12120, Thailand
b
Faculty of Civil Engineering, Universiti Teknologi MARA, Shah Alam, Selangor 40450, Malaysia

a r t i c l e i n f o a b s t r a c t

Article history: Fatality rates at workplaces in the construction industry are high compared to other industries.
Received 23 November 2016 Tremendous effort is required to strive towards zero accidents. Managing foreign workers with different
Received in revised form 3 March 2017 cultural backgrounds at the workplace requires appropriate safety intervention practices to improve
Accepted 13 June 2017
workers’ safety behaviour. Based on the literature, the importance of safety intervention for changing
Available online 23 June 2017
unsafe to safe worker behaviour is known. For this reason, an integrated safety intervention model affect-
ing workers’ safety behaviour was developed and tested. This study was conducted by distributing a
Keywords:
questionnaire survey to construction companies. The survey was randomly distributed, with a total of
Integrated safety intervention
Construction industry
198 responses received. Exploratory factor analysis (EFA) was conducted to confirm the three safety
Safety behaviour intervention constructs. Structural equation modelling (SEM) was performed to identify the most signif-
Structural equation modelling icant intervention-related safety practices, which are to be the focus in handling safety management. The
results indicate that technical intervention has a positive influence by management and human interven-
tion. In addition, an improvement in workers’ safety behaviour can be achieved by focusing on the tech-
nical intervention with five important safety practices: workplace safety inspections, personal protective
equipment (PPE) programmes, safety equipment availability and maintenance, safe work practices, and
safety permits. These findings attempt to help construction management by identifying the appropriate
selection of safety practices with specific interventions to improve workers’ safety behaviour.
Ó 2017 Elsevier Ltd. All rights reserved.

1. Introduction and Zhang and Fang (2013) find that unsafe work behaviour is
the most common cause of accidents at construction sites.
The dynamic environment of the construction industry at the Research in the area of safety management systems related to
construction site workplace is highly potentially harmful to work- safety behaviour has become prominent in developing countries
ers. International statistics show that the majority of fatal acci- (Vinodkumar and Bhasi, 2010). A number of study findings have
dents occur in the construction industry (Aksorn and indicated that management commitment plays a vital role in han-
Hadikusumo, 2007; Mohamed, 2002; Sawacha et al., 1999; Tam dling safety at construction sites (Aksorn and Hadikusumo, 2007;
et al., 2004). Similarly, in Malaysia, according to occupational acci- Mohamed, 2002; Sawacha et al., 1999), finding that safety training
dent statistics in 2014, among all industries, 184 fatality cases is the most important safety management practice in terms of
were inspected by the Department of Occupational Safety and safety performance components. However, different countries
Health (DOSH). The highest number, 72 fatalities, representing probably have various important safety practices to focus on to
almost 40% of total cases, was recorded in the construction indus- improve worker safety behaviour. As Geller (2001, 2005) states,
try (DOSH, 2014). Based on these statistics, the construction indus- the research question refers to developing a guideline for use by
try in Malaysia is still incapable of effectively tackling safety issues safety personnel in determining the appropriate safety behaviour
at the construction site workplace. Chi et al. (2014) find unsafe interventions for safety controls. Wirth and Sigurdsson (2008)
behaviour to be the main cause of accidents in Taiwan’s construc- argue that there is limited information in the safety literature on
tion industry. Similarly, Choudhry (2014), Sawacha et al. (1999), how to integrate miscellaneous safety intervention controls. Abun-
dant safety research in the construction industry has been con-
ducted in terms of safety performance improvement, such as the
⇑ Corresponding author at: School of Engineering and Technology, Asian Institute
studies by Aksorn and Hadikusumo (2007), Chinda and Mohamed
of Technology, P.O. Box 4, Klong Luang, Pathumthani 12120, Thailand.
E-mail addresses: kusumo@ait.ac.th, st115568@ait.asia (M. Mazlina Zaira).
(2008), Choudhry (2012), Tam et al. (2004), and Teo and Ling

http://dx.doi.org/10.1016/j.ssci.2017.06.007
0925-7535/Ó 2017 Elsevier Ltd. All rights reserved.
 M. Mazlina Zaira, B.H.W. Hadikusumo / Safety Science 98 (2017) 124–135 125

(2006). However, previous safety researchers have focused on

safety management systems without differentiating the safety
intervention practices that may have different effects on worker
safety behaviour performance. For this reason, the present study
addresses the improvement of worker safety behaviour in the con-
struction industry by integrating safety intervention practices.
Identification of the relevant safety intervention components on
which to concentrate is an important strategy to change unsafe
worker behaviour. Meanwhile, the objective of the present study
is to develop a model of integrated safety intervention practices
to encourage positive worker safety behaviour. The most signifi- Fig. 2. Integrated management, human, and technical safety intervention practices.
cant intervention and the related safety practices that should be
focused on to improve safety worker behaviour have been identi-
fied by analysing the relationships. study to determine the type of intervention that is appropriate to
promote work safety behaviour. Ramli et al. (2014) conduct
research on the industrial safety management of Malaysian multi-
2. Safety intervention national firms. According to them, based on the opinions of
experts, several factors must be reviewed because every country
Research on construction companies considers that all the has its own peculiarities and differences that may exist. They argue
safety practices implemented are under one roof as a safety man- that different countries have different methods of establishing
agement system (Teo and Ling, 2006; Chan et al., 2004; Robson safety management that depend on the safety culture and the local
et al., 2007). Abundant safety practices implemented in construc- culture. Due to the uniqueness of the Malaysian construction
tion companies are important to enforce to develop positive industry’s characteristics such as foreign workers from Indonesia,
worker behaviour. Oyewole and Haigth (2009) and Robson et al. Bangladesh, and Nepal, this issue requires a wide range of research
(2001) define a safety intervention as a step to alter or implement to understand how to implement safety management, as interven-
methods to improve safety. For example, training programmes, tion works more effectively. Therefore, the present study is con-
engineering intervention, and administrative procedures are ducted to identify which safety intervention level group and
among the new safety practices, including programmes within safety practices are crucial to encourage safe behaviour among for-
the workplace or actions for purposely improving safety. In addi- eign construction workers.
tion, Shakioye and Haight (2010) state that the safety management Fundamental constructs (management safety intervention,
in the organisation comprises two main intervention levels, as technical safety intervention, human safety intervention, and
shown in Fig. 1. Thus, at the management level, safety intervention workers’ safety behaviours), as developed in Fig. 3, and hypotheses
refers to the top management’s policies and safety administrative associated with each path of the model are discussed below.
activities. At the technical level, safety intervention refers to any
practices that ensure a safe working environment, including safety
planning. At the human level, safety intervention refers to any
2.1. Management safety intervention
practices to change human knowledge and cognition, considering
safety practices that directly affect the worker.
Many studies have found that management commitment and
Thus, the present study considers the integration of safety inter-
involvement are key elements in achieving success in safety per-
vention, with the first level being management and the second
formance in the workplace (Langford et al., 2000; Sawacha et al.,
level being human and technical. Fig. 2 illustrates the safety inter-
1999; Tam et al., 2004). It is difficult to perform safety practices
vention that occurs at different levels; on the top is the manage-
and foster a culture of safety within an organisation without man-
ment level, followed by the technical level and the human level.
agement support (Teo and Ling, 2006), which indicates that safety
All three intervention components are considered to have a direct
management is the backbone of intervention in the implementa-
influence on workers’ safety behaviour. Human safety intervention
tion of effective safety in the workplace. Robson et al. (2001)
is considered to influence the technical safety intervention. The
explain that the management of the organisation can influence
present study classifies all of the important safety practices applied
the technical and human intervention levels.
to the three intervention components, as identified by Robson et al.
Based on these previous studies, the following hypotheses have
(2001) and Shakioye and Haight (2010). This safety intervention
been formulated:
levels perspective highlights the dynamic and integration of safety
management influences for safety practices implementation at
each different intervention level. The target for safety intervention
is to change unsafe worker behaviour to safe worker behaviour, as
Neal et al. (2000) consider workers’ safety behaviour to be one of
the safety outcomes.
H1a H3
The different characteristics of this industry, such as a majority
of the workers in the workplace being foreign, require the present
H1b H1c H2b

H2a
H1b H2b

Fig. 1. Levels of safety intervention in the organisational structure (Shakioye and

Haight, 2010). Fig. 3. Hypothesised integrated safety intervention structural equation model.
126 M. Mazlina Zaira, B.H.W. Hadikusumo / Safety Science 98 (2017) 124–135

H1a. Management safety intervention positively influences According to Cooper (1998), an initiative in the implementation
technical safety intervention. of safety behaviour intervention leads to a reduction in accident
H1b. Management safety intervention positively influences rates and safety-related expenses. Based on scientific evaluations,
human safety intervention. positive outcomes, such as improved safety performance, a
H1c. Management safety intervention has a direct effect on strengthened safety climate, safety management systems that are
workers’ safety behaviour. increasingly improved by the workforce, increased cooperation
between management and labour in involvement and communica-
2.2. Human safety intervention tion, augmented acceptance of safety responsibilities, and a better
understanding of the relationship between accidents and safety
Robson et al. (2001) explain that this level includes interven- behaviour, are shown. Zohar and Luria (2003) state that, during
tions to change human knowledge, competence, attitude, motiva- many routine jobs, a lack of cautious behaviour still prevails;
tion, or behaviour related to safety. In addition, human safety hence, it has become a managerial challenge to encourage safe
intervention is also considered to indirectly influence technical behaviour among workers. Ismail et al. (2012) state that most
safety intervention. For example, the implementation of safety safety professionals think that the main purpose of a good safety
inspections (a technical safety intervention item) should be programme is to encourage and modify safety behaviour. Safety
affected by safety training (a human safety intervention item). outcomes are one indicator for identifying safety performance in
Accordingly, the following hypotheses are proposed as follows: the construction industry. In safety research, Neal et al. (2000)
use safety behaviour to measure safety outcomes. In addition, Lai
H2a. Human safety intervention positively influences technical et al. (2011) state that, after the implementation of safety pro-
safety intervention. grammes, worker attitudes change from safety concern at the indi-
H2b. Human safety intervention has a direct effect on workers’ vidual level to safety concern at the group level, which encourages
safety behaviour. worker participation. This proposal suggests that safety behaviour
is involved with individual and group commitment, particularly at
2.3. Technical safety intervention construction site workplaces.
In the broad safety behaviour research literature and in con-
Intervention changes an organisation, design, or environment at struction safety management and safety culture studies, there is
work (Robson et al., 2001). As Shakioye and Haight (2010) explain, a gap in specific research on different levels of safety intervention
a technical system forms the greater part of management planning practices. The present study aims to identify and create a model of
for the reduction of incident rates. This includes safe equipment integrated safety intervention practices that includes the manage-
maintenance, job procedures, and work permits as related to tech- ment, technical, and human levels that directly and indirectly
nical aspects that purposely create safe work behaviour. Thus, an affect safety behaviour by utilising SEM. This model attempts to
additional hypothesis is proposed as follows: demystify safety management implementation with the necessary
safety practices on site. In this research, it is assumed that the qual-
H3. Technical safety intervention has a direct effect on workers’ ity of intervention is constant, following the research by Shakioye
safety behaviour. and Haight (2010). Due to cost and time constraints, it is difficult to
implement all safety intervention practices at the workplace.
All of the aforementioned hypotheses have been identified as Therefore, the present study obtains information about how safety
integrated safety intervention practices that affect workers’ safety practices at every intervention level are practically implemented in
behaviour in the construction industry and the relationships the workplace from construction companies. The benefit of the
among them. present study is that it may guide construction companies in
Assume that integrated safety intervention practices have a sig- implementing the most significant safety practices in each inter-
nificant impact on safety behaviour performance. Safety interven- vention group.
tion, as determined from theory, comprises three pathways to
intervene, with the first level through management and the second
level through technical and human aspects. Hence, the effect of 3. Research method
management safety intervention on safety behaviour is partially
mediated by technical and human safety intervention. A hypothe- 3.1. Questionnaire structure
sised structural equation model with four latent variables and six
correlation paths is illustrated in Fig. 3. Thus, a research model Questionnaire development was based on safety intervention
based on management, technical, and human safety intervention practices and safety behaviour criteria from the literature review,
(Neal et al., 2000) affecting safety behaviour has been developed; referring to 23 journal articles, as shown in Table 1; expert valida-
the research model by Robson et al. (2001) and Shakioye and tion was also conducted. A preliminary questionnaire comprised
Haight (2010) indicates that technical and human safety interven- four constructs with 47 content items: management safety inter-
tion is influenced by management safety intervention. All of these vention (13 practices), technical safety intervention (13 practices),
have been identified as latent variables. Management safety inter- human safety intervention (15 practices), and safety behaviour (6
vention is an exogenous variable, whereas the technical and items). A set of questionnaires was completed anonymously using
human factors are mediating variables. Safety behaviour is the a three-point rating: (1) disagree, (2) not sure, and (3) agree. The
dependent variable. There are many different data scenarios utilis- purpose of this validation was to ensure that the questions were
ing the structural equation modelling (SEM) approach; this is truly representative of actual practices and conditions and to
known as mediation analysis, a popular use of SEM, as noted by determine how the contents were applicable for the current con-
Iacobucci (2010). struction industry in Malaysia. After verification of the content
by five expert safety officers (one from oil and gas construction,
2.4. Safety behaviour one from infrastructure construction, and three from building con-
struction), who had at least 10 years’ experience and had been
Intervention is required to improve safety performance, partic- involved in safety and handling construction projects, the degree
ularly worker safety behaviour (Wirth and Sigurdsson, 2008). of agreement was rated. This aspect depended on tacit knowledge
 M. Mazlina Zaira, B.H.W. Hadikusumo / Safety Science 98 (2017) 124–135 127

Table 1
List of the multilevel safety intervention practices in the construction industry.

Constructs Label Contents Supporting sources

M1 Safety policy Fernández-Muñiz et al. (2007) and Mullen (2004)
M2 Safety objectives Nielsen (2014)
M3 Safety organisation, Safety committee Teo and Ling (2006), Chan et al. (2004) and Nielsen (2014)
M4 Safety standard Tam et al. (2004)
M5 Management worker interaction (periodic safety meetings, regular site Iyer et al. (2004)
visits)
M6 Safety records Yu et al. (2014)
M7 Incident and accident investigation, analysis, and prevention Teo and Ling (2006), Chan et al. (2004) and Nielsen (2014)
M8 In-house safety rules and regulations Teo and Ling (2006) and Chan et al. (2004)
M9 Contracting strategy (safety requirement and capability) Teo and Ling (2006) and Chan et al. (2004)
M10 Safety information management and feedback Iyer et al. (2004) and Laitinen and Ruohomäki (1996)
M11 Safety audit on overall safety management system Mohamed (2002) and Oyewole et al. (2010)
M12 Reviewing and implementing safety programmes Iyer et al. (2004)
T1 Inspecting hazardous conditions, facilities, plant, and equipment to Teo and Ling (2006), Chan et al. (2004), Choudhry et al. (2008) and Iyer
affirm a safe workplace (workplace safety inspections) et al. (2004)
T2 Personal protective equipment (PPE) programme Teo and Ling (2006), Chan et al. (2004) and Choudhry et al. (2008)
T3 Safe work practices, safe operation procedures Teo and Ling (2006)
T4 Safety equipment availability and maintenance Iyer et al. (2004)
T5 Implementation of safety inspections Teo and Ling (2006), Chan et al. (2004) and Yu et al. (2014)
T6 Scheduled maintenance for all machinery and equipment Teo and Ling (2006)
T7 Movement control and use of hazardous substances and chemicals, Teo and Ling (2006) and Chan et al. (2004)
safety process control programme
T8 Emergency response preparedness Teo and Ling (2006) and Chan et al. (2004)
T9 Designing safe temporary structure for construction Gambatese (2000) and Toole (2002)
T10 Implementation of safety permits for high-risk operations Oyewole et al. (2010)
H1 Behavioural-based safety (BBS) programme Choudhry (2012) and Ismail et al. (2012)
H2 Safety training Teo and Ling (2006), Chan et al. (2004), Choudhry et al. (2008), Iyer
et al. (2004) and Oyewole et al. (2010)
H3 Safety inductions for new workers Makin and Winder (2008)
H4 Safety awards, safety promotion, safety incentives Teo and Ling (2006), Chan et al. (2004), Choudhry et al. (2008), Iyer
et al. (2004), Oyewole et al. (2010) and Zohar (2002)
H5 Safety supervision Oyewole et al. (2010), Zohar and Luria (2003) and Zohar (2002)
H6 Safety awareness programme, safety campaigns, safety knowledge Nielsen (2014) and Yu et al. (2014)
programme, safety education
H7 Safety information, safety bulletin boards Nielsen (2014) and Yu et al. (2014)
H8 Requisite safety expertise for high-risk operations Yu et al. (2014)
H9 Job hazard analysis (JHA), job safety analysis (JSA) Teo and Ling (2006), Chan et al. (2004) and Iyer et al. (2004)
H10 Daily tailgate, toolbox meeting Iyer et al. (2004) and Oyewole et al. (2010)
H11 Penalty, accident repeater punishment programme Hopkins (2006)
I1 Safety awareness Neal et al. (2000)
I2 Safety competency Neal et al. (2000)
I3 Safety understanding Neal et al. (2000)
G1 Peers who actively care Neal et al. (2000)
G2 Sharing safety concerns Neal et al. (2000)
G3 Working safely together Törner and Pousette (2009)

and perception specifically influencing appropriateness to finalise background, (ii) questions about how safety intervention practices
the questionnaire. The latest simplified table of content validity at each level have been applied, and (iii) questions about the per-
ratio (CVR) critical values produced by Ayre and Scally (2014) ception of the influence of safety behaviour on the implemented
was used as a guideline to validate the preliminary questionnaire’s safety intervention practices.
content. The group of experts required a minimum CVR value of 1
for each element. Table 1 shows that 39 of the 47 items remained 3.2. Construct measurement
in the final research instrument as part of the questionnaire survey
structure. Each intervention classification consisted of several 3.2.1. Management safety intervention practices
safety practices and safety outcomes with safety behaviour criteria. Safety policy development demonstrates an organisation’s
The results were evaluated on a nine-point Likert scale (1- commitment to safety with a clear sense of responsibility
Extremely poor to 9-Extremely good), which tends to produce bet- (Fernández-Muñiz et al., 2007). Moreover, to encourage an
ter distributions of data. Leung (2011) states that a Likert scale may improvement of workers’ safety behaviour, a clear mission and
be used to expand an instrument for the measurement of opinions goals must be set by the organisation (Nielsen, 2014). Hence, safety
and preferences. This method of evaluation measured construction objectives must be determined. Both the safety policy and safety
companies’ perceptions of and commitment to each construct to be objectives are mainly established for a safety system based on
performed. Before the ground analysis, to ensure the reliability of the concerns of management (Janssens et al., 1995). Key elements
the questionnaire items, a pilot study was conducted. The next sec- in a safety management system (Teo and Ling, 2006; Chan et al.,
tion shows the detailed items and the Cronbach’s alpha value for 2004) include safety organisations and committees; incident/acci-
each construct above 0.80, indicating high internal consistency. A dent investigation, analysis, and prevention; in-house safety rules
questionnaire survey was designed by incorporating the validated and regulations; and evaluation, selection, and control subcontrac-
safety intervention practices and safety behaviour criteria. The tors as a contract strategy. Safety standards are set by management
final questionnaire comprises three parts: (i) questions about to ensure the safety of equipment, processes, activities, and special
the general information of the respondents and the company’s safety conditions (Tam et al., 2004) that are normally established
128 M. Mazlina Zaira, B.H.W. Hadikusumo / Safety Science 98 (2017) 124–135

by a regulatory body. Other safety intervention practices by man- supervision by assigning competent safety personnel to protect
agement are management-worker interactions such as periodic workers from workplace hazards. Zohar and Luria (2003) state that
safety meetings and regular site visits to update and monitor supervision enhances safety behaviour. Safety campaigns, safety
worker safety issues (Iyer et al., 2004). A safety record is a neces- information, and safety bulletin boards are safety intervention
sary document for storing all statistical data and details relevant activities that directly affect workers, as specified by Nielsen
to safety. Safety information and feedback are required in safety (2014). The requisite safety expertise for each task is required
management at the workplace, and they have a positive impact (Toole, 2002), particularly for high-risk operations, to ensure that
on safety performance. As Laitinen and Ruohomäki (1996) employees work under safe control. Pre-task hazard assessment,
shows, the safety index is significantly higher when feedback is i.e., job safety analysis (JSA), also known as job hazard analysis
given after goals are issued compared to when no feedback is (JHA), is a tool that must be handled by an employee in the execu-
given. Mohamed (2002) states that safety management audits tion of a job task. Supervisors must study and record the steps of a
assist a company in focusing on target responsibility on safety. job task to identify possible hazards that may occur and to deter-
Moreover, the effectiveness of safety management systems that mine the best method of performing the job task correctly and
are performed in real practice should be assessed by auditing safely (Ismail et al., 2012). Poon et al. (2000) have shown that com-
(Teo and Ling, 2006). A safety audit provides managers with panies may reduce accident frequency rates by implementing JHA.
clear information about the overall status of safety management Daily tailgate or toolbox safety meetings are a type of safety inter-
performance. Hence, it becomes easier to determine whether the vention activity to be considered (Iyer et al., 2004; Oyewole and
current safety programmes are effective. Without reviewing and Haigth, 2009). Thus, one should expect workers to be more aware
implementing safety programmes, a company will be unable to of their safety. Punishment similar to accident repeater pro-
create a safety culture for a long-term and/or short-term safety grammes (Hopkins, 2006) is also a tool for stopping unsafe beha-
climate. viours. All of these safety practices directly affect the worker.
Therefore, these interventions are considered part of the human
3.2.2. Technical safety intervention practices component.
Technical safety intervention in a safety management system
(Teo and Ling, 2006; Chan et al., 2004) includes inspecting haz- 3.2.4. Safety behaviour criterion
ardous conditions such as the facilities, plants, and equipment as Hadjimanolis and Boustras (2013) argue that the entire range of
to whether the conditions are safe to start the work. In other risks and hazards at the workplace is faced by a workgroup. In
words, it considers workplace safety inspections; housekeeping terms of behaviourally oriented techniques, worker involvement
practices to ensure a safe workplace; a personal protective equip- comprises individual or group involvement (Vinodkumar and
ment (PPE) programme; safe work practices; maintenance for all Bhasi, 2010). Thus, the safety behaviours of individual and group
machinery and equipment; safety inspections; safety process con- workers are counted. Neal et al. (2000) conduct research based
trol programmes such as movement control and the manner in on safety performance related to safety behaviour. Six components
which hazardous substances and chemicals are used; and emer- are considered to be included in safety behaviour performance.
gency response preparedness. Gambatese (2000) states that the Referring to the research of Neal et al. and Törner and Pousette,
design stage should proactively consider site safety. Hence, the the components found are (i) the safety awareness of the individ-
design of a safe temporary structure for construction is one of ual worker, considering that the worker is voluntarily performing
the technical practices that ensures that the work site is safe and tasks or activities that help improve workplace safety; (ii) the
that workers are able to work safely. For example, temporary load- safety competency of the individual worker, considering that the
ings are safe on temporary work platforms (Toole, 2002). The worker ensures the highest levels of safety when performing the
implementation of safety permits is necessary for high-risk opera- job; (iii) the safety understanding of the individual worker, consid-
tions (Oyewole and Haigth, 2009). Workers will be aware of the ering that the worker uses the correct safety procedures for per-
existence of safety permits, that safer work is required and that forming the job; (iv) peers who actively care for the group of
they need to be extra careful. All of these safety practices have workers, considering that workers help their co-workers when
been included in the technical level because they are more related they are working under risky conditions; (v) sharing the safety
to the intervention of workers through the equipment, the envi- concern of workers in a group, considering that workers help co-
ronment, or systems. workers in safety learning and implementation; (vi) working safely
together, in which the group of workers is considered, given that
3.2.3. Human safety intervention practices nobody ever works alone and workers should work together safely
Choudhry (2012) states that the process of safety engagement if they are concerned about safety.
between management and workers is regularly subject to daily
safety behaviour concerns about workers’ actions and attention, 3.3. Data collection
which is known as behaviour-based safety (BBS). It has been noted
that training plays the most vital role in safety at work in the con- The unit analysis set in the present study is construction com-
struction industry (Hinze and Gambatese, 2003). In addition, Tam panies. Duff (2000) shows that problems in construction safety
et al. (2004) state that safety training is a main safety practice in are associated with management control by the contractors
reducing accident rates. The purpose of safety training is to provide involved in the process of construction. Hence, it is shown that
the necessary knowledge related to safety that should be known by accident causation by the construction management contractor is
workers to guide them in how to work in a safe manner. Employers related to the failure to manage unsafe acts or conditions
must provide direct safety practices for all new workers to com- (Abdelhamid and Everett, 2000). Construction companies regis-
plete safety inductions. Makin and Winder (2008) describe induc- tered under the Construction Industry Development Board (CIDB)
tion into a safe system as one of the building blocks of an in Malaysia at Grade 7 that are eligible to undertake a project that
occupational health and safety management system. One method costs more than 10 million MYR were targeted. As stated in Sec-
of encouraging workers’ concern for safety and health is a safety tion 29 of the Occupational Safety and Health Act of 1994, any con-
incentive practice that rewards workers for positive behaviours, struction company that has the ability to receive a total project
such as safe acts or reporting hazards, as related to safety issues contract of above twenty million ringgits must employ a safety
(Teo and Ling, 2006). Employers are required to provide sufficient and health officer. Therefore, G7 contractor companies are
 M. Mazlina Zaira, B.H.W. Hadikusumo / Safety Science 98 (2017) 124–135 129

expected to have safety capabilities. Company contact details were 3.4. Analysis methods
extracted from the CIDB Malaysia database website. The present
study focuses only on one group of respondents (the safety man- A structural equation model (SEM) of the hypothesised con-
agement team: the project manager, the safety manager, the safety structs utilised Analysis of Moment Structure (AMOS), Version
officer, the project engineer, the site engineer, the site safety super- 21. Two main factor analyses were used. First, exploratory factor
visor, the site supervisor, and those in similar positions). It consid- analysis (EFA) using Statistical Package for the Social Sciences
ers the respondents’ perceptions of impacts on individuals and (SPSS) was conducted to explore and confirm that all of the safety
groups of workers based on the selected safety behaviour criteria. practices were within the related construct structures of a group of
Four hundred questionnaires were circulated to construction com- three constructs from the data set. The main applications of EFA
panies by email and were followed up by phone calls. Due to the are to detect structures in the relationships between variables to
passive response rate after a month, the data collection process categories and reduce the variables into smaller numbers based
was preceded by visiting the construction site personally and on the factors (Thompson, 2004). Second, confirmatory factor anal-
meeting the respondents to complete the questionnaire. A total ysis (CFA) by AMOS 21 was conducted on a data set to test the cor-
of 212 company responses were received, for a response rate of relation structure against a hypothesised structure based on the
53%. After data screening, 14 responses that were deemed unus- ‘‘goodness of fit” values (Schermelleh-Engel et al., 2003).
able due to missing data and unengaged respondents were
removed. One hundred ninety-eight data samples were retained. 4. Results
Iacobucci (2010) describes that a sample size of 150 is adequate
for the SEM approach. Thus, 198 data samples were sufficient to 4.1. EFA (preliminary analysis)
conduct the analysis. The construction companies were randomly
selected from six province locations and three main company pro- According to Henson and Roberts (2006), EFA is a useful tool for
ject types (see Table 2). These were selected based on rapid devel- testing for variation and manifest variables to generate a new the-
opment and having many construction projects underway. The ory by exploring latent factors. To confirm all 33 safety practice
specific respondents were based on positions that directly engaged components underlying the three safety intervention constructs,
with safety management and their companies’ project activities EFA is conducted. According to Matsunaga (2010), this analysis is
(see Table 2). The respondents over 30 years old constitute almost able to reconstruct the complex observed data in an essential form
half the total respondents (43.5%). Ages ranging from 25 to by identifying a set of unobserved factors. After an essential form is
30 years old were the majority (56.5%). The respondents were created, all of the important information from the original data by
divided into those with less than 10 years’ work experience the factor solution extracted from an EFA should remain. Unneces-
(75.3%) and those with more than 10 years’ work experience sary factors induced by measurement errors have been removed.
(24.7%). More than 90% of the respondents were bachelor’s degree According to Comrey and Lee’s (1992) guidelines, 0.63 is the cut-
or diploma holders. During the survey, the researcher observed off selection for primary target factor loadings. Principle compo-
that many young middle-aged people work as site engineers and nent analysis and varimax rotation are conducted in EFA. In total,
safety supervisors after having graduated with a bachelor’s degree by selecting a fixed number of factors, three constructs were
or diploma. The construction company project types were 71.2% extracted. The 33 total items were reduced to 25 after 8 items
buildings, 21.7% infrastructure, and 7.1% oil and gas. Table 2 shows (M6, H2, H3, H7, H8, H9, T6, and T8) were removed due to factor
the demographic details from the data collection. loadings of less than 0.63 (see Table 3). The results indicate that

Table 2
Respondents’ and companies’ demographics. Table 3
Reliability testing for the final model.
Characteristic Category Frequency Percentage
(N = 198) Items Factor loadings Cronbach’s alpha
Age 25–30 years old 112 56.5 Management Technical Human Α
>30 years old 86 43.5
M5 0.817 0.964
Gender Male 162 81.8 M2 0.803
Female 36 18.2 M9 0.792
Designation Project manager/ 54 27.3 M4 0.780
engineer M11 0.770
Safety manager 10 5.1 M8 0.757
Site engineer 14 7.1 M1 0.756
Safety officer 30 15.2 M3 0.755
Site safety supervisor 57 22.7 M12 0.753
Site supervisor 33 16.7 M10 0.698
M7 0.655
Work experience <10 years 149 75.3
>10 years 49 24.7 T2 0.802 0.925
T5 0.765
Educational PhD/Master’s 15 7.6
T9 0.702
background Bachelor’s 101 51.0
T3 0.692
Diploma/others 82 41.4
T1 0.662
Location Wilayah Persekutuan 33 16.7 T4 0.658
Selangor 71 35.9 T10 0.655
Johor 57 28.8 T7 0.655
Pulau Pinang 11 5.6
H10 0.778 0.919
Sabah 13 6.6
H11 0.744
Sarawak 13 6.6
H4 0.716
Company’s project Building 141 71.2 H1 0.680
type Infrastructure 43 21.7 H5 0.653
Oil and gas 14 7.1 H6 0.651
130 M. Mazlina Zaira, B.H.W. Hadikusumo / Safety Science 98 (2017) 124–135

Table 4
Final model fitted criteria.

Fit index Evaluation criteria CFA Model SEM Model

Base Fitted Base Fitted
Chi-square test X2/df <3.00 (Arbuckle, 2011) 2.979 2.340 3.026 2.117
Absolute fit RMSEA <0.10 (Byrne, 2009) 0.100 0.082 0.101 0.075
Incremental fit CFI >0.90 (Hair et al., 2010) 0.890 0.952 0.902 0.948
Parsimonious fit PNFI >0.50 (Chen and Fong, 2012) 0.765 0.751 0.763 0.777

the list of safety practices was gathered in the expected construct The most commonly used criteria have been reported in this study,
and that there was no change in the content of the constructs, as shown in Table 4. The first analysis shows that the CFA base
including the following: management safety intervention (11 prac- model does not meet the criteria of the root mean square error
tices), technical safety intervention (8 practices), and human safety of approximation (RMSEA) and the comparative fit index (CFI).
intervention (6 practices). The adequacy and strength of the mea- Hence, it is necessary to modify the model until it fits. According
surement model was determined by a Cronbach’s alpha reliability to Arbuckle (2011), there are three methods of modifying the
test. The acceptable level of internal consistency is above the cut- model: deleting the minimum loading factor variable, combining
off value of 0.60 as the Cronbach’s alpha coefficient (Nunnaly, the two variables into one variable, or adding a covariance relation.
1978). The internal consistency of three exogenous variables that In this model, only the method of combining two variables into one
result in values above 0.60 has been shown to be adequately reli- variable remains unused.
able for analysis. Subsequently, CFAs with three constructs and
with one construct require at least three or more observed vari- 4.3. SEM (structural model)
ables to diminish bias in estimating the parameters (Iacobucci,
2010). Independent variables are known as exogenous variables.
Endogenous variables are similar to dependent variables. These
4.2. CFA (measurement model) are two other terms that are associated with SEM (Schreiber
et al., 2006). The purpose of the SEM model is to explore the inter-
For many researchers, the choice of model fit indices is perplex- relationship among latent variables, as described by Xiong et al.
ing, as noted by Iacobucci (2010). Further, according to Xiong et al. (2015). The SEM model must be fitted with four fit index cate-
(2015), SEM has an application in construction research and spec- gories: the chi-square test X2/df < 3.00, the absolute fit
ified goodness of fit (GOF) evaluation criteria and practical results. RMSEA < 0.10, the incremental fit CFI > 0.90, and the parsimonious

0.70 0.66 0.68 0.61 0.58

0.82

0.76

0.72
0.77

0.68
0.82
0.65

0.79
0.75
0.12

0.79 0.69

0.76 0.72

0.67 0.71

0.62

0.71 0.59 0.72 0.76

Fig. 4. Final multilevel safety intervention structural equation model.

 M. Mazlina Zaira, B.H.W. Hadikusumo / Safety Science 98 (2017) 124–135 131

Table 5
The final output relationship of multilevel safety intervention with safety behaviour.

Hypothesis Estimate S.E. C.R. p Hypothesis testing result

H1a Management ? Technical 0.264 0.072 3.684 0.000 Direct influence
H1b Management ? Human 0.719 0.076 9.445 0.000 Direct influence
H1c Management ? Safety behaviour 0.113 0.088 1.283 0.199 Indirect effect
H2a Human ? Technical 0.630 0.087 7.260 0.000 Direct influence
H2b Human ? Safety behaviour 0.182 0.139 1.304 0.192 Indirect effect
H3 Technical ? Safety behaviour 0.902 0.164 5.514 0.000 Direct effect

Note: p < 0.05 significant level.

fit PNFI > 0.50, as shown in detail in Table 4. According to the val- as safe operation procedures, workers will not perform their work
ues in Table 4, the base SEM model does not fit three fit index cat- properly or safely. The expected direct effect of human safety
egories: the X2/df, RMSEA, and CFI criteria. Modifying the SEM intervention on safety behaviour (H2b) was not supported because
model is similar to CFA analysis in three ways. In this process, the human safety intervention construct was not significantly
ten covariance relations are added for the SEM model to meet all related to safety behaviour. A list of important safety practices
of the criteria. Hence, the final SEM model is illustrated in Fig. 4 results in a model: safety supervision, safety campaign, toolbox
below, and the final output results are tabulated in Table 5. meetings, and punishment. This safety intervention level shows
that it is not effective to focus the implementation due to the
characteristics of construction workers in Malaysia. It can be
5. Discussion assumed that human safety intervention does not directly affect
safety behaviour. Perhaps the practice of earlier exposure to
5.1. Discuss on hypotheses results workers is not consistent throughout the job because, in terms of
communication, difficulty in understanding is a challenge for
Hypothesis 1. The predicted positive influence of management foreign workers. For example, a toolbox meeting may be held
safety intervention on technical safety intervention (H1a) and on before starting work. Afterwards, workers simply do their work
human safety intervention (H1b) was supported because the without considering what was discussed. This includes safety
management safety intervention construct significantly influenced campaigns, of which only a few are presented to workers. Workers
technical and human safety intervention. Referring to the impor- will not remember to practice them. As Choudhry et al. (2008)
tant list of safety practice results from the model discussed in H1, reported on a toolbox meetings programme, the company should
these management safety intervention practices support technical obtain feedback from workers on safety matters. It seems that
safety intervention implementation. It is shown that the results safety practices under human intervention are better for develop-
support the argument of Robson et al. (2001); that is, management ing a safety culture. According to Kines et al. (2010), the safety
can influence the human and technical intervention levels in an concern among workers is more lasting with safety practices with
organisation. direct coaching of the worker on site. This shows the importance of
daily verbal safety communication among workers. Furthermore,
The predicted direct effect of management safety intervention Bust et al. (2008) argue that different nationalities have different
on safety behaviour (H1c) was not supported because the manage- ways of expressing themselves and understanding their beha-
ment safety intervention construct was not significantly related to viours. Hence, the safety management system requires a new
safety behaviour. A list of important safety practices results in a approach to address the cross-cultural misunderstandings that
model: safety policies, safety objectives, safety committees, safety occur among foreign workers and that lead to safety problems. This
standards, safety meetings, in-house safety rules, safety feedback, is the challenge of transforming the safety management system to
and safety audits. This indicates that the intervention of manage- accommodate multinational or multicultural workers on site.
ment does not directly affect workers’ safety behaviour. However, Therefore, it is difficult to implement human intervention safety
it affects safety behaviour through one level of the other two safety practices among foreign workers of various nationalities, who are
interventions (human and technical). The management safety characteristic of the Malaysian construction industry.
intervention level practices focus more on establishing the imple- Moreover, the safety practices included in human safety interven-
mentation of safety in a construction company. Hence, the effect tion seemingly require more effective communication to deliver
on workers’ safety behaviour is not significant. However, there safety information to foreign workers on site. Based on the
are many other safety performance criteria that most likely have research findings by Ismail et al. (2012), one of the main factors
a positive effect on this intervention level. The findings by influencing the implementation of safety management systems in
Aksorn and Hadikusumo (2008) indicate that management support Malaysian construction sites is communication. Because all foreign
is the key contributor to a successful safety programme. This is rel- workers are from other countries and thus speak different
evant to the findings of this study because the management safety languages, they do not fully understand the local language.
intervention level has a direct effect on both of the other interven- The author’s perception is that perhaps this issue is one of the
tions, which in turn will change workers’ safety behaviour. reasons why the findings of human safety intervention practices
have no direct effect on improving workers’ safety behaviour
Hypothesis 2. The predicted positive influence of human safety among foreign workers.
intervention on technical safety intervention (H2a) was supported.
It is shown that human safety intervention practices support
technical safety intervention practices. For example, the toolbox Hypothesis 3. The predicted direct effect of technical safety inter-
meeting (a human safety intervention item) should affect the vention on safety behaviour (H3) was supported because the tech-
implementation of safe operation procedures (a technical safety nical safety intervention construct was significantly related to
intervention item). The reason is that toolbox meetings should be safety behaviour. A list of important safety practices results in a
directly implemented and performed by safety personnel for model: workplace safety inspections, PPE programmes, safety
workers. Therefore, without further technical interventions such equipment availability and maintenance, safe work practices, and
132 M. Mazlina Zaira, B.H.W. Hadikusumo / Safety Science 98 (2017) 124–135

safety permits. In the author’s opinion, these findings are logically several limitations and requires further research for the compar-
why technical safety intervention directly affects worker safety ison of findings from different conditions. First, an intervention
behaviour. One reason is the characteristics of the construction process was not involved due to cost and time constraints. The
industry in Malaysia, in which many foreign workers are study focuses only on safety intervention practices applied in the
employed. They are more easily controlled by technical means. current construction industry without considering other safety
The three most significant safety practices among the technical practices that will be implemented in the future. Second, the pre-
aspect are safety inspections, PPE requirements, and safety proce- sent survey targeted only contractor companies at Grade 7 and
dures. For example, workplace inspections make workers aware of did not include small and medium-sized contractor companies.
performing a job safely within a safe area. Moreover, workers Arocena and Núñez (2010) state that small firms have a higher risk
immediately think that they should constantly be alert by wearing of occupational accidents than large firms. Moreover, the list of
PPE while working. Safety procedures also directly guide workers safety intervention practices implemented in small and medium-
to work safely. Choudhry and Fang (2008) show that workers sized firms is less than or different from those implemented in
observe that working at heights is dangerous. The safety awareness large contractor companies. Therefore, it would be beneficial for
of workers causes them to wear the required PPE and check that future research to cover small and medium-sized construction
the equipment is safe to use. It is shown that the recent findings companies in determining which intervention is more appropriate
support part of the previous research outcomes. The author con- to tackle. Third, this study was conducted primarily within the
tends that technical safety intervention leads to the development context of Malaysia, and the surveys focused on the Malaysian con-
of a safety climate among the workers as well. struction industry. Because the characteristics of workers in this
country are different, further research would need to be conducted
in other countries, given that there are most likely different safety
5.2. Significance of the findings intervention levels for improving safety behaviour depending on
the characteristics of workers in different countries.
Most safety management research findings indicate that man-
agement commitment is the vital key. However, occasionally, a site
6. Conclusions
safety supervisor is required to prepare an abundance of safety
documents. This leads to frustration in the process of completing
The fatality rate at workplaces in the Malaysian construction
an inspection form with improper checking (Ding et al., 2000).
industry is high. There have been tremendous efforts to strive
Technical safety intervention has a direct effect on workers’ safety
towards zero accidents. To provide more effective guidelines for
behaviour. Technical intervention safety practices have a positive
safety personnel in implementing safety management in the con-
effect on safety behaviour. The most important safety practices
struction workplace, references for safety intervention practices
for managing workers’ safety behaviour are workplace safety
are provided in this paper. The safety intervention practices
inspections, PPE programmes, the availability and maintenance of
approach to improving workers’ safety behaviour in this paper is
safety equipment, safe work practices, and safety permits. Consci-
firmly rooted. Specifically, this study illuminates safety interven-
entious action by the contractor is required for implementation at
tion, in which the contractor companies should focus more on
the construction site. In addition, the management safety interven-
improving workers’ safety behaviour. This study shows that man-
tion plays an important role in the implementation of safety prac-
agement intervention influences the other two safety intervention
tices at the technical and human intervention levels. Previous
components, the technical and the human aspects. In addition,
studies indicate that management commitment is essential in han-
human intervention also influences technical intervention. How-
dling safety in construction without specific recommended safety
ever, only one technical intervention has a direct effect on the
intervention practices (Langford et al., 2000; Sawacha et al.,
safety behaviour performance of workers through five safety prac-
1999; Tam et al., 2004). Therefore, this model provides some guid-
tices (PPE programmes, safe work practices, safety equipment
ance to construction companies to better focus on specific safety
availability, safety permits, and hazardous conditions inspections).
intervention practices that improve workers’ safety behaviour.
The findings of our study show that, for managing foreign workers
Acknowledgement
with different cultural backgrounds, technical interventions are
more effective with the implementation of PPE programmes, safety
The first author would like to express thanks for fund alloca-
permits, workplace safety inspections, the availability and mainte-
tions provided by the Ministry of Education of Malaysia and the
nance of safety equipment, and safe work practices, whereas
Faculty of Civil Engineering, Universiti Teknologi MARA of
human interventions are less effective. Hence, the findings of this
Malaysia.
study reveal the relationship of management, technical, and
human interventions in safety behaviour. Technical interventions
are the most effective intervention directly affecting safety beha- Appendix A. Content of questionnaire
viour, whereas management and human interventions are impor-
tant for supporting the technical interventions. A.1. Part 1 general information

5.3. Recommendations and further research Please tick U the appropriate answer to respond the general
information
The practical application of this study is the provision of guide-
lines that contractor companies, particularly those with foreign Age: Less than 25; 25–30; More than 30.
workers at construction sites, can use to improve safety behaviour Gender: Male; Female.
by focusing on the technical safety intervention component, with Designation: Project manager; Project engineer; Safety man-
support from management safety intervention as well. The present ager; safety engineer; safety officer; site safety supervisor; site
study addresses improvements in safety behaviour in the Malay- supervisor.
sian construction industry through integrated safety interventions, Work experience:<10 years; >10 years.
which certainly provide guidance for the safety personnel and Educational background: PhD; Master; Bachelor; Diploma;
management team in construction companies. This study has Others.
 M. Mazlina Zaira, B.H.W. Hadikusumo / Safety Science 98 (2017) 124–135 133

Location: Wilayah persekutuan; Selangor; Johor; Pulau Pinang; Technical

Sabah; Sarawak.
Company’s project type (Major type): Building; Infrastructure; Oil 2.1 Inspecting hazardous conditions, facilities, plant and
and gas. equipment to affirm a safe workplace (workplace safety
inspections)
A.2. Part 2 construction safety intervention practices 2.2 Personal protective equipment (PPE) programme
2.3 Safe work practices, safe operation procedures
Please tick U the appropriate scale accordingly how well the 2.4 Safety equipment availability and maintenance
safety practices intervention has been applied to the company 2.5 Implementation of safety inspections
according to the following statements about construction safety 2.6 Scheduled maintenance for all machinery and
intervention practices (1-Extremely very poor; 5-fair; 9-Exremely equipment
very good).

2.7 Movement control & use of hazardous substances and

chemicals, safety process control programme
1. Extremely very poor mean ‘‘Not been practiced” 2.8 Emergency response preparedness
2. Extremely poor mean ‘‘One time practiced” 2.9 Designing safe temporary structure for construction
3. Very poor mean ‘‘Rarely practiced” 2.10 Implementation of safety permits for high risk
4. Poor mean ‘‘Inconsistent practiced” operations
5. Fair mean ‘‘Occasionally practiced”
6. Good mean ‘‘Usually practiced”
7. Very good mean ‘‘Consistent practiced” Human
8. Extremely good mean ‘‘Frequently practiced”
9. Extremely very good mean ‘‘Well established practiced” 3.1 Behavioural-based safety (BBS) programme
3.2 Safety training
3.3 Safety inductions for new workers
Management 3.4 Safety awards, safety promotion, safety incentives
3.5 Safety supervision
1.1 Safety policy 3.6 Safety awareness programme, safety campaigns, safety
1.2 Safety objectives knowledge programme, safety education
1.3 Safety organisation, safety committee 3.7 Safety information, safety bulletin boards
1.4 Safety standard 3.8 Requisite safety expertise for high-risk operations
1.5 Management worker interaction (periodic safety 3.9 Job hazard analysis (JHA), job safety analysis (JSA)
meetings, regular site visits) 3.10 Daily tailgate, toolbox Meeting
1.6 Safety records 3.11 Penalty, accident repeater punishment programme
1.7 Incident and accident investigation, analysis, and
prevention
1.8 In-house safety rules and regulations
1.9 Contracting strategy (safety requirement and A.3. Part 3 Behavioural safety performance outcome
capability)
1.10 Safety information management and feedback Please tick U the appropriate scale accordingly how would you
1.11 Safety audit on overall safety management system rate the quality level of the following behavioural safety perfor-
1.12 Reviewing and implementing safety programmes mance of worker based on your experience in the company accord-
ing to the following statements about level of workers’ safety
behaviour performance (1-Extremely very poor; 5-fair; 9-
Exremely very good).
134 M. Mazlina Zaira, B.H.W. Hadikusumo / Safety Science 98 (2017) 124–135

Chinda, T., Mohamed, S., 2008. Structural equation model of construction safety
culture. Eng. Constr. Architect. Manage. 15, 114–131. http://dx.doi.org/10.1108/
09699980810852655.
1. Extremely very poor mean ‘‘None of the worker” Choudhry, R.M., 2012. Implementation of BBS and the impact of site-level
2. Extremely poor mean ‘‘25% of the worker” commitment. J. Profess. Iss. Eng. Educ. Pract. 138 (4), 296–304.
Choudhry, R.M., 2014. Behavior-based safety on construction sites: a case study.
3. Very poor mean ‘‘35% of the worker” Accid. Anal. Prev. 70, 14–23.
4. Poor mean ‘‘45% of the worker” Choudhry, R.M., Fang, D., 2008. Why operatives engage in unsafe work behavior:
5. Fair mean ‘‘50% of the worker” investigating factors on construction sites. Saf. Sci. 46 (4), 566–584.
Choudhry, R.M., Fang, D., Ahmed, S.M., 2008. Safety management in construction:
6. Good mean ‘‘65% of the worker”
best practices in Hong Kong. J. Profess. Iss. Eng. Educ. Pract. 134 (1), 20–32.
7. Very good mean ‘‘75% of the worker” Comrey, A.L., Lee, H.B., 1992. A First Course in Factor Analysis. Psychology Press.
8. Extremely good mean ‘‘85% of the worker” Cooper, D., 1998. Improving Safety Culture: A Practical Guide. Wiley.
Department of Occupational Safety and Health. <http://www.dosh.gov.my/index.
9. Extremely very good mean ‘‘100% of the worker”
php?option=com_content&view=article&id=1225:occupational-accidents-
statistics-by-sector-2&catid=502:2014-archive-statistics&Itemid=1280&lang=
en> (Retrieved 9 Jun 2015).
Ding, C.B., Deng, Q., Yao, C.W., 2000. Safety Construction in China. China Building
Safety behaviour Industry Yearbook Press, Beijing.
Duff, A.R., 2000. Behavior measurement for continuous improvement in
Individual worker behaviour construction safety and quality. Manage. Constr. Safety Health, 1–18.
Fernández-Muñiz, B., Montes-Peón, J.M., Vázquez-Ordás, C.J., 2007. Safety
management system: development and validation of a multidimensional
1.1 Safety awareness scale. J. Loss Prev. Process Ind. 20, 52–68. http://dx.doi.org/10.1016/j.
Voluntarily conducting tasks or activities that help to jlp.2006.10.002.
improve workplace safety Gambatese, J.A., 2000. Safety in a designer’s hands. Civ. Eng. 70 (6), 56.
Geller, E.S., 2001. Working Safe: How to Help People Actively Care for Health and
1.2 Safe competency
Safety. CRC Press.
Ensure the highest levels of safety when they conduct the Geller, E.S., 2005. Behavior-based safety and occupational risk management. Behav.
job Modif. 29 (3), 539–561.
1.3 Safety understanding Hadjimanolis, A., Boustras, G., 2013. Health and safety policies and work attitudes in
Cypriot companies. Saf. Sci. 52, 50–56.
Use the correct safety procedures for conducting the job and Hair, J.F., Black, W.C., Babin, B.J., Anderson, R.E., 2010. Multivariate data analysis.
know safety issues Vectors 816. <http://www.emeraldinsight.com/doi/abs/10.1108/eb021138>.
Henson, R.K., Roberts, J.K., 2006. Use of exploratory factor analysis in published
research common errors and some comment on improved practice. Educ.
Psychol. Measur. 66 (3), 393–416.
Hinze, J., Gambatese, J., 2003. Factors that influence safety performance of specialty
Group worker behaviour contractors. J. Constr. Eng. Manage. 129, 159–164. http://dx.doi.org/10.1061/
(ASCE)0733-9364(2003) 129:2(159).
Hopkins, A., 2006. What are we to make of safe behavior programs? Saf. Sci. 44,
2.1 Peers who actively care 583–597. http://dx.doi.org/10.1016/j.ssci.2006.01.001.
Help their co-workers when they are working under risky Iacobucci, D., 2010. Structural equations modeling: fit indices, sample size, and
condition advanced topics. J. Consum. Psychol. 20, 90–98. http://dx.doi.org/10.1016/j.
jcps.2009.09.003.
2.2 Sharing safety concerns Ismail, F., Ahmad, N., Janipha, N.A.I., Ismail, R., 2012a. Assessing the behavioral
Helping co-workers in safety learning and implementation factors’ of safety culture for the Malaysian construction companies. Proc. – Soc.
2.3 Workingly safe together Behav. Sci. 36, 573–582. http://dx.doi.org/10.1016/j.sbspro.2012.03.063chan.
Ismail, Z., Doostdar, S., Harun, Z., 2012b. Factors influencing the implementation
Nobody ever works alone in construction industry, should of a safety management system for construction sites. Saf. Sci. 50 (3),
work together safely 418–423.
Iyer, P.S., Haight, J.M., Del Castillo, E., Tink, B.W., Hawkins, P.W., 2004. Intervention
effectiveness research: understanding and optimizing industrial safety
programs using leading indicators. Chem. Health Saf. 11 (2), 9–19.
Janssens, M., Brett, J.M., Smith, F.J., 1995. Confirmatory cross-cultural research:
testing the viability of a corporation-wide safety policy. Acad. Manage. J. 38 (2),
References 364–382.
Kines, P., Andersen, L.P., Spangenberg, S., Mikkelsen, K.L., Dyreborg, J., Zohar, D.,
2010. Improving construction site safety through leader-based verbal safety
Abdelhamid, T.S., Everett, J.G., 2000. Identifying root causes of construction
communication. J. Saf. Res. 41 (5), 399–406.
accidents. J. Constr. Eng. Manage. 126 (1), 52–60.
Lai, D.N.C., Liu, M., Ling, F.Y.Y., 2011. A comparative study on adopting human
Aksorn, T., Hadikusumo, B.H.W., 2007. The unsafe acts and the decision-to-err
resource practices for safety management on construction projects in the
factors of Thai construction workers. J. Constr. Dev. Count. 12 (1), 1–25.
United States and Singapore. Int. J. Project Manage. 29, 1018–1032. http://dx.
Aksorn, T., Hadikusumo, B.H.W., 2008. Critical success factors influencing safety
doi.org/10.1016/j.ijproman.2010.11.004.
program performance in Thai construction projects. Saf. Sci. 46 (4), 709–727.
Laitinen, H., Ruohomäki, I., 1996. The effects of feedback and goal setting on safety
Arbuckle, J.L., 2011. Amos (20.0). Crawfordville, FL: Amos Development
performance at two construction sites. Saf. Sci. 24, 61–73. http://dx.doi.org/
Corporation.
10.1016/S0925-7535(96)00070-7.
Arocena, P., Núñez, I., 2010. An empirical analysis of the effectiveness of
Langford, D., Rowlinson, S., Sawacha, E., 2000. Safety behavior and safety
occupational health and safety management systems in SMEs. Int. Small Bus.
management: its influence on the attitudes of workers in the UK construction
J. 28 (4), 398–419.
industry. Eng. Constr. Architect. Manage. 7, 133–140. http://dx.doi.org/10.1108/
Ayre, C., Scally, A.J., 2014. Critical Values for Lawshe’s Content Validity Ratio:
eb021138.
Revisiting the Original Methods of Calculation. Measurement and Evaluation in
Leung, S.O., 2011. A comparison of psychometric properties and normality in 4-, 5-,
Counseling and Development, vol. 47, Sage Publ. Inc., pp. 79–86. doi:http://
6-, and 11-point Likert scales. J. Soc. Serv. Res. 37 (4), 412–421.
dx.doi.org/10.1177/0748175613513808.
Makin, A.M., Winder, C., 2008. A new conceptual framework to improve the
Bust, P.D., Gibb, A.G., Pink, S., 2008. Managing construction health and safety:
application of occupational health and safety management systems. Saf. Sci. 46
migrant workers and communicating safety messages. Saf. Sci. 46 (4), 585–602.
(6), 935–948.
Byrne, B., 2009. Structural Equation Modeling with AMOS - Basic Concepts,
Matsunaga, M., 2010. How to factor-analyze your data right: do’s, don’ts, and how-
Applications, and Programming.
to’s. Int. J. Psychol. Res. 3 (1), 97–110.
Chan, A.H.S., Kwok, W.Y., Duffy, V.G., 2004. Using AHP for determining priority in a
Mohamed, S., 2002. Safety climate in construction site environments. J. Constr. Eng.
safety management system. Ind. Manage. Data Syst. 104, 430–445. http://dx.
Manage. 128 (5), 375–384. http://dx.doi.org/10.1061/(ASCE)0733-9364(2002)
doi.org/10.1108/02635570410537516.
128:5(375).
Chen, L., Fong, P.S.W., 2012. Revealing performance heterogeneity through
Mullen, J., 2004. Investigating factors that influence individual safety behavior at
knowledge management maturity evaluation: a capability-based approach.
work. J. Saf. Res. 35 (3), 275–285.
Expert Syst. Appl. 39 (18), 13523–13539. http://dx.doi.org/10.1016/j.
Neal, A., Griffin, M., Hart, P., 2000. The impact of organizational climate on safety
eswa.2012.07.005.
climate and individual behavior. Saf. Sci. 34, 99–109. http://dx.doi.org/10.1016/
Chi, C.F., Lin, S.Z., Dewi, R.S., 2014. Graphical fault tree analysis for fatal falls in the
S0925-7535(00)00008-4.
construction industry. Accid. Anal. Prev. 72, 359–369.
 M. Mazlina Zaira, B.H.W. Hadikusumo / Safety Science 98 (2017) 124–135 135

Nielsen, K.J., 2014. Improving safety culture through the health and safety Shakioye, S.O., Haight, J.M., 2010. Modeling using dynamic variables–an approach
organization: a case study. J. Saf. Res. 48, 7–17. for the design of loss prevention programs. Saf. Sci. 48 (1), 46–53.
Nunnaly, J., 1978. Psychometric Theory. McGraw-Hill, New York. Tam, C.M., Zeng, S.X., Deng, Z.M., 2004. Identifying elements of poor construction
Oyewole, S.A., Haight, J.M., 2009. Making the business case: assessment of safety safety management in China. Saf. Sci. 42 (7), 569–586.
intervention and optimization of resource allocation. In: ASSE Professional Teo, E.A.L., Ling, F.Y.Y., 2006. Developing a model to measure the effectiveness of
Development Conference and Exhibition. American Society of Safety Engineers. safety management systems of construction sites. Build. Environ. 41 (11), 1584–
Oyewole, S.A., Haight, J.M., Freivalds, A., Cannon, D.J., Rothrock, L., 2010. Statistical 1592.
evaluation and analysis of safety intervention in the determination of an Thompson, B., 2004. Exploratory and Confirmatory Factor Analysis: Understanding
effective resource allocation strategy. J. Loss Prevent. Proc. Ind. 23 (5), 585–593. Concepts and Applications. American Psychological Association.
Poon, W.F., Ma, C.H., Ho, K.L., 2000. Statistical analysis on factors in reducing Toole, T.M., 2002. Construction site safety roles. J. Constr. Eng. Manage. 128 (3),
construction site accident frequency rate in Hong Kong. In: Proceedings of the 203–210.
6th Annual Conference of the Australian and New Zealand Association of Törner, M., Pousette, A., 2009. Safety in construction - a comprehensive description
Occupational Health and Safety Educators, Hong Kong, pp. 341–355. of the characteristics of high safety standards in construction work, from the
Ramli, A., Mokhtar, M., Aziz, B.A., 2014. The development of an initial framework for combined perspective of supervisors and experienced workers. J. Safety Res. 40
multi-firm industrial safety management based on cooperative relationship: a (6), 399–409. http://dx.doi.org/10.1016/j.jsr.2009.09.005.
Malaysia case study. Int. J. Disaster Risk Reduct. 10, 349–361. Vinodkumar, M.N., Bhasi, M., 2010. Safety management practices and safety
Robson, L.S., Shannon, H.S., Goldenhar, L.M., Hale, A.R., 2001. Guide to evaluating the behavior: assessing the mediating role of safety knowledge and motivation.
effectiveness of strategies for preventing work injuries. Natl. Inst. Occup. Safety Accid. Anal. Prev. 42, 2082–2093. http://dx.doi.org/10.1016/j.aap.2010.06.021.
Health 121. Wirth, O., Sigurdsson, S.O., 2008. When workplace safety depends on behavior
Robson, L.S., Clarke, J.A., Cullen, K., Bielecky, A., Severin, C., Bigelow, P.L., Irvin, E., change: topics for behavioral safety research. J. Safety Res. 39 (6), 589–598.
Culyer, A., Mahood, Q., 2007. The effectiveness of occupational health and safety Xiong, B., Skitmore, M., Xia, B., 2015. A critical review of structural equation
management system interventions: a systematic review. Saf. Sci. 45, 329–353. modeling applications in construction research. Autom. Constr. 49, 59–70.
http://dx.doi.org/10.1016/j.ssci.2006.07.003. http://dx.doi.org/10.1016/j.autcon.2014.09.006.
Sawacha, E., Naoum, S., Fong, D., 1999. Factors affecting safety performance on Yu, Q.Z., Ding, L.Y., Zhou, C., Luo, H.B., 2014. Analysis of factors influencing safety
construction sites. Int. J. Project Manage. 17 (5), 309–315. management for metro construction in China. Accid. Anal. Prev. 68, 131–138.
Schermelleh-Engel, K., Moosbrugger, H., Müller, H., 2003. Evaluating the fit of Zhang, M., Fang, D., 2013. A cognitive analysis of why Chinese scaffolders do not use
structural equation models: tests of significance and descriptive goodness-of-fit safety harnesses in construction. Constr. Manage. Econ. 31 (3), 207–222.
measures. Meth. Psychol. Res. Online 8 (2), 23–74. Zohar, D., 2002. The effects of leadership dimensions, safety climate, and assigned
Schreiber, J.B., Nora, A., Stage, F.K., Barlow, E.A., King, J., 2006. Reporting structural priorities on minor injuries in work groups. J. Org. Behav. 23 (1), 75–92.
equation modeling and confirmatory factor analysis results: a review. J. Educ. Zohar, D., Luria, G., 2003. The use of supervisory practices as leverage to improve
Res. 99 (6), 323–338. safety behavior: a cross-level intervention model. J. Safety Res. 34 (5), 567–577.