Accident Analysis and Prevention 93 (2016) 267279

Contents lists available at ScienceDirect

Accident Analysis and Prevention

journal homepage: www.elsevier.com/locate/aap

Critical factors and paths inuencing construction workers safety

risk tolerances
Jiayuan Wang b , Patrick X.W. Zou a, , Penny P. Li a
a
Department of Civil and Construction Engineering and Centre for Sustainable Infrastructures, Swinburne University of Technology, Hawthorn, VIC 3122,
Australia
b
College of Civil Engineering, Shenzhen University, Shenzhen, China

a r t i c l e

i n f o

Article history:
Received 3 August 2015
Received in revised form 2 November 2015
Accepted 21 November 2015
Available online 13 January 2016
Keywords:
Risk tolerance
Construction
Safety risk
Safety management
Personal subjective perception
System thinking

a b s t r a c t
While workers safety risk tolerances have been regarded as a main reason for their unsafe behaviors,
little is known about why different people have different risk tolerances even when confronting the same
situation. The aim of this research is to identify the critical factors and paths that inuence workers
safety risk tolerance and to explore how they contribute to accident causal model from a system thinking
perceptive. A number of methods were carried out to analyze the data collected through interviews and
questionnaire surveys. In the rst and second steps of the research, factor identication, factor ranking
and factor analysis were carried out, and the results show that workers safety risk tolerance can be
inuenced by four groups of factors, namely: (1) personal subjective perception; (2) work knowledge
and experiences; (3) work characteristics; and (4) safety management. In the third step of the research,
hypothetical inuencing path model was developed and tested by using structural equation modeling
(SEM). It is found that the effects of external factors (safety management and work characteristics) on risk
tolerance are larger than that of internal factors (personal subjective perception and work knowledge
& experiences). Specically, safety management contributes the most to workers safety risk tolerance
through its direct effect and indirect effect; while personal subjective perception comes the second and
can act as an intermedia for work characteristics. This research provides an in-depth insight of workers
unsafe behaviors by depicting the contributing factors as shown in the accident causal model developed
in this research.
2015 Elsevier Ltd. All rights reserved.

1. Introduction
Practically, the development of construction industry has been
plagued by the accidents or injuries that are frequently occurred.
It is estimated that there are around 60,000 construction fatalities occurred worldwide each year, which equates to one accident
happens every nine minutes (ILO, 2006). Further, the construction
industry employs nearly 10% of the workforce but it accounts for
2040% of the occupational fatal accidents (Raheem and Hinze,
2014). These highly disproportionate numbers indicate a deteriorating situation of construction industry. A thorough understanding
of the accident causation mechanism is essential for accident prevention. Heinrich et al. (1950) advocated that accidents are caused
by an unsafe act (e.g., an individuals behavior or activity that
Corresponding author at: Department of Civil and Construction Engineering
and Centre for Sustainable Infrastructures, Swinburne University of Technology,
Hawthorn, VIC 3122, Australia.
E-mail address: pwzou@swin.edu.au (P.X.W. Zou).
http://dx.doi.org/10.1016/j.aap.2015.11.027
0001-4575/ 2015 Elsevier Ltd. All rights reserved.

deviates from normal accepted safety procedure) or an unsafe condition (e.g., deciency in machines and materials). Research has
found that, 88% of accidents are caused by the former, and 10%
by the later (Heinrich et al., 1950). Garrett and Teizer (2009) also
pointed out that human error is a main reason for up to 80% of all
incidents and accidents in complex high-risk industry such as mining, construction and nuclear power. Further, Fang (2012) asserted
that workers unsafe behaviors have been recognized as the direct
and common reason for construction accidents. Since the characteristics of construction work determine workers usually work on
separate sites, this decentralization makes it more difcult to identify and manage unsafe behaviors (Olson and Austin, 2001; Gould
and Joyce, 2003).
Theoretically, workers internal factors such as attitude, perception and efcacy play a vital role in safety performance. As shown in
Fig. 1, unsafe behavior and unsafe condition have been recognized
as the main reasons for construction accidents, and among these
two, unsafe conditions can resulted from misusing of equipment
(i.e., workers are not familiar with crane operating) or decient

268

J. Wang et al. / Accident Analysis and Prevention 93 (2016) 267279

Fig. 1. A general accident casual model.

management (i.e., sloppy material test). It has been proven that the
reasons of unsafe behaviors can be explained from two aspects:
(1) internal factors: which means personal characteristics, such as
risk perception, risk attitude, risk tolerance, self-efcacy and stress
(Hallowell, 2010; Wang and Yuan, 2011; Wang, 2014; Dixit et al.,
2014; Leung et al., 2012); and (2) external factors, which refers to
the environment in which individuals are living, such as culture,
regulations and weather (de Camprieu et al., 2007; Liu et al., 2014;
Acar and Gc, 2011). After an accident happens, project safety performance is negatively inuenced. Workers who experienced or
witnessed injures tend to be more risk aversion and less optimistic
(Shin et al., 2014). As a result, unsafe behaviors can be reduced.
Then, less accidents and better safety performance would occur.
Accordingly, the changes of individuals internal factors will determine the performance of construction safety. In other words, they
act as the input of accident causal model. And thus, exploration of
factors contributing different internal factors is vital for successful
accident prevention from a system thinking perceptive (Fung et al.,
2012). Having said so, it is worth noting that from a system perspective, there is normally a feedback loop which represents that
the input factors (i.e., the causes) are also affected by the outputs
(i.e., safety performance). This will be discussed in details at later
sections.
Previous studies have shown that non-objective risk assessment would happen if taking no consideration of individuals
risk tolerance (Hopkinson, 2012; Mu et al., 2014). Lichtenstein
et al. (1978) pointed out that people tend to overestimate their
ability to control or prevent accidents, thus leads to an underestimation of the risks. Balaz and Williams (2011) emphasized
the effect of risk tolerance on immigrants risk perceptions, and
showed that the more uncertainties immigrants can accept, the
more likely they underestimate the seriousness of potential risks.
Basically, individuals with higher (lower) risk tolerance are more
(less) likely to take a risk. Therefore, objective assessment of
risk tolerance plays a critical role in an effective and successful
safe behavior. Decisions made without considering risk tolerance might not be persuasive or reliable. Nevertheless, which
factors inuence workers safety risk tolerance and to what
extent the factors can affect remained as important and unsolved
problems.
System thinking, as a discipline of seeing systems holistically
(Goh et al., 2010) has been used for construction safety improvement. For example, Leveson (2004) developed a new accident
model based on system thinking, and proved that it can provide
better and less subjective understanding of the reasons and preventions of accidents. Goh et al. (2010) built a causal loop diagram
to simulate the relationships between safety culture and major
accidents. Their results illustrated that an amalgamation reaction
from each part can promote poor safety culture even though each
of them are necessary from each partys point of view. Shin et al.
(2014) developed a system dynamics model to analyze the feedback mechanisms and the resultant dynamics regarding workers
safety attitudes and safe behaviors. It is summarized that systematic thinking are effective in representing the complex interactions
between factors and making the accident be more readily understood.

Accordingly, investigating reasons behind different assessments

of risk tolerance and introducing them into accident causal model
would be helpful to reveal the causes of workers unsafe behaviors, and tailored prevention actions can be developed. This paper
starts from a review of the literature depicting what risk tolerance
is and why it plays an important role in risk management. Then
the research instruments and methods adopted in this study are
introduced. Afterwards, it is the identication of critical factors
inuencing workers safety risk tolerance and the analysis of the
interactions between critical factors and recognizing the key inuencing paths. The critical inuencing factors are then introduced
into the accident causal model. It is expected that the ndings of
this research not only contributes to the knowledge body of systematic safety risk management, but also serves as practical guidance
for construction safety management.
2. Risk tolerance and its function in risk management
A major issue within the literature of risk tolerance is the
lack of general agreement regarding the denition of this concept.
From the perceptive of governmental agencies, the Securities and
Exchange Commission (SEC) made an explanation of risk tolerance
as your ability and willingness to lose some or all of your original
investment in exchange for greater potential returns (SEC, 2010).
It is applied by the Financial Industry Regulatory Authority (FINRA)
(2011) as a guideline to dene risk tolerance. The ISO31000 Guide
73:2009 denes a more general risk tolerance as organizations or
stakeholders readiness to bear the risk after risk treatment in order
to achieve its objective. In addition, some researchers also give
their explanations of risk tolerance. Irwin (1993) dened nancial
risk tolerance as the willingness to engage in behaviors in which
the outcomes remain uncertain with the possibility of an identiable negative outcome. Grable (2000) explained the conception as
investors tolerance toward nancial risk refers to the amount of
uncertainty or investment return volatility that an investor is willing to accept when making a nancial decision. It should be noticed
that the denitions present above are focused on the nancial eld.
From the perspective of risk behavior and decision making, Hunter
(2002) gave a denition as the amount of risks that individuals
are willing to accept in the pursuit of some goal. Roszkowski and
Davey (2010) also agreed that this denition can provide better
understanding of decision makers risk tolerance.
Based on the above denitions, there are some key words that
indicate risk tolerance contains two aspects: (1) subjective, e.g.,
willingness and readiness, which address whether people want to
take risks; and (2) objective, e.g., ability, amount of uncertainties
and amount of risks, which refers that how many losses can individuals tolerant. Some researchers have realized the complicacy in risk
tolerance. Burton (1996) proposed that risk tolerance can be in two
parts: subjective risk tolerance, which was based on the economic
concept of risk aversion; objective risk tolerance, which refers to
individuals nancial situation, including investment horizon for
each goal. Similarly, Cordell (2002) contended that risk tolerance
should be separated into risk capacity and risk attitude, among
them, risk capacity is more an objective measure which includes
income and nancial stability. In contrast, risk attitude is more a

J. Wang et al. / Accident Analysis and Prevention 93 (2016) 267279

269

Fig. 3. Risk probability, impact and tolerance.

Fig. 2. A typical risk management process (adopted from ISO 31000:2009).

subjective measure which refers whether investors are risk aversion, risk neutral or risk seeking. These explanations provide a more
direct way to understand willingness and ability. Specically,
willingness refers whether an individual wants to take a risk, and
the ability means if a person is capable enough to bear losses.
This paper selected the objective part to represent individuals
risk tolerance. First, no matter whether an individual is risk aversion or risk seeking, one of the most critical things in risk decision
making is to measure whether they have enough real things to pay
potential losses, such as money. This means there exist interactions between the two dimensions of risk tolerance, the objective
assessment can inuence its subjective part. Accordingly, if they
are taken as a whole, confusions could be formed about the difference between risk tolerance and risk attitude. Thus, in this paper,
risk tolerance means individuals capability of bear losses.
Risk tolerance plays a critical role in risk management. Based on
Fig. 2, which is adopted from ISO 31000:2009, risk assessment is
the core of the process of risk management as it provides a basis for
decisions about the most appropriate approach to be used to treat
risks by offering decision-makers with an improved understanding
of how potential risks could affect achievement of objectives (ISO,
2009a). In other words, the output of risk assessment is an input
of risk-based decision-making (Karimiazari et al., 2011). However,
what is the output of risk assessment? After risk identication and
risk analysis, risk evaluation is processed to determine whether a
risk is acceptable and/or tolerable. As showed in Fig. 3, risks located
above the intolerable line are regarded as cannot be tolerated no
matter what benets the risk may bring. For these kinds of risks,
risk treatment is essential whatever their costs; for risks belong to
middle band, it means these risks can be tolerated; in the lower
band, the level of risks is regarded as negligible or so small that no
risk treatment measures are needed (ISO, 2009b).
It is noteworthy that the intolerable line plays a critical role in
risks ranking and resources distributing. Kwak and Laplace (2005)
pointed out that managing risk is resource consuming, and not
only prioritizing identied risks and address the most crucial ones
are of great importance, but also to know how much to reduce
them so that the risk is acceptable is crux for improving decisionmaking. However, the intolerable line is difcult to determine
as acceptance of risk may change throughout the duration of a
project (Kwak and Laplace, 2005) and it is unreliable to make it just
according to historical gures and facts pertaining to the project

at hand (Kahneman and Lovallo, 1993). As for construction workers, ajavascript:void(0); large proportion of them never received
professional risk management training, let alone having awareness of how much safety risk they can tolerate. It is expected that
this research can help workers to understand how their risk tolerance are formed by exploring the critical inuencing factors and
paths.
3. Research methodology
3.1. Research design
Fig. 4 shows the research process and methods employed for
data collection and analysis in this study. The whole research comprises three steps.
3.2. Data collection
In step one, a literature review, including Frank (2011), Moreschi
(2004), Rawlinson and Farrell (2009), Abdelhamid and Everett
(2000) and Acar and Gc (2011), was undertaken to develop an
in-depth understanding of the inuencing factors that related to
workers behaviors and the risks construction workers may confront. Then, semi-structured interviews with 10 workers from 2
construction companies were carried out to develop the tentativecritical factor list in which all factors can inuence workers safety
risk tolerance.
In step two, based on the factor list obtained by step one, survey questionnaire was adopted for data collection, which contains
two main sections: the rst section is designed to collect the background information of respondents; the second section consists of
27 factors with potential effects on workers safety risk tolerance.
The Likert scale was used to quantify these effects and was ranged
as 1, 2, 3, 4 and 5, which is correspondingly represent respondents
attitude from (1) = this factor has the least effect on risk tolerance,
and (5) = this factor has the strongest effect on risk tolerance (the
effects grow with number increases). Three hundred and eightythree (383) sets of questionnaires were distributed to workers, who
were employed in 11 construction companies in China. Finally, 297
valid responses were received. The software of Statistic Package
for the Social Sciences (SPSS) version 23.0 was used to process the
empirical data and make statistical analysis. Factor ranking and factor analysis methods were adopted to determine the critical factors
inuencing workers safety risk tolerance.

270

J. Wang et al. / Accident Analysis and Prevention 93 (2016) 267279

Fig. 4. Research process.

In step three, hypothetical mode, indicating the relationships

between the factors and potential inuencing paths, was built
based on the ndings of step two. Then, measurement scales
for each factor were developed. Before data collection, interviews with experienced construction practitioners are conducted
to ensure the accuracy and completeness of the measurement
scales and the questionnaire contents, as well as expression style.
Three practitioners participated in this process, including two
project managers with 10 or more years of site-management
experiences; and one professor, who has been engaged in construction safety research for more than 10 years. Based on their
comments, some items were revised to avoid ambiguity and
inaccuracy.
In the process of data collection, workers, which come from
11 construction projects in China, were gathered together in the
project meeting room and were given instructions related to the
questionnaire by the researchers, including the research aims,
the questionnaire structure and content, and ethical matters. No
communication between participants was allowed during this process. In case some participants had difculties in answering the
questionnaire, individual explanation was provided for them to
guarantee the high quality data. This survey process lasted for
three months and took place during the daily safety briengs of the
projects, in order to minimize the disturbance the research brought
to the construction project in case over-disturbance caused biases
to the data collection.

3.3. Data analysis

In the process of data analysis, conrmatory factor analysis
(CFA) and structural equation modeling (SEM) were applied to test
the reliability and validity of the research reported in this paper. In
CFA, the Cronbach was used to evaluate the internal consistency
reliability which reects correlations between questionnaire items
belonging to one dimension (Flynn et al., 1994). After the reliability
of the factorial structure of constructs reported and their measurement scales conrmed, the SEM was carried out to provide support
for the hypotheses, especially the links among different constructs.
The software Analysis of Moment Structure (AMOS) version 18 was
used to process the empirical data and make statistical analysis.
AMOS software is a widely-used tool for SEM (Zou and Sunindijo,
2013). During the testing process, various t indices were used to
assess the t of the model: goodness-of-t index (GFI), adjusted
goodness-of-t index (AGFI), comparative t index (CFI) and root
mean square error of approximation (RMSEA). Values of GFI, AGFI

and CFI of 0.9 or above, and an RMSEA of 0.08 or less also indicate
a good t.
4. Identication of critical factors and their grouping
4.1. Possible factors
Based on the results of literature review and interviews, a list of
27 factors was obtained, which formed the main basis for questionnaire design. All factors are listed in Table 1 with detail descriptions.
4.2. Critical factors
This section identies the important factors affecting workers
safety risk tolerance. The mean and standard deviation of each factor are derived from the total sample to determine the level of
importance. Factors with higher mean values and lower standard
deviation indicate they have greater impacts on workers risk tolerance. The factors with mean values that are greater than the average
value of all mean values (3.1178) are classied as critical factors in
affecting workers safety risk tolerance. The ranking results of these
factors are shown in Table 2.
Based on Table 2, there are 14 factors among the initial 27
factors receive a mean value of greater than 3.1178, which are
therefore determined as critical factors inuencing workers safety
risk tolerance in construction project. The top-ve critical factors
are Time constraints for decision-making, Relevant safety regulations, and Behaviors from peers, Safety knowledge and
Relevant project information, each of which has a mean value
above 3.680.
4.3. Grouping of critical factors
This section explores the groupings that might exist among
the 14 critical factors. Factors analysis was applied to address the
problem of analyzing the structure of the interrelationships (correlations) among a large number of variables.
The survey data was analyzed by principal component analysis,
which is a common method in factor analysis. The results show that
the Bartlett test of sphericity is 4733.668 with signicance level of
0.000, and the value of the KaiserMayerOlkin measure of sampling adequacy is 0.809 (higher than 0.50). These demonstrate that
the sample meets the fundamental requirements for factor analysis (Hair, 2010). The principal component analysis generated a
four-factor solution with eigenvalues greater than 1.0. The component matrix after rotation is presented in Table 3. Each factor

J. Wang et al. / Accident Analysis and Prevention 93 (2016) 267279

271

Table 1
Descriptions of factors that may inuence safety risk tolerance.
No.

Factors

Descriptions

F01

Physical health

F02

Emotion

F03

Judgment ability

F04

Professional knowledge

F05

Working experiences

F06

Weather

F07

Noise

F08

Interested in their work

F09

Workload

F10

Availability of safety protection equipment

F11

Safety training

F12

Supervision

F13

Safety regulations

F14

Peers behaviors

F15
F16

Age
Education background

F17

Perception of risks

F18

Trust of self-capability

F19

Time constraints for decision making

F20

Completeness of relevant project

information

F21

Boldness

F22
F23
F24

Attitude of Immediate supervisor

Emphasis on safety
Decision motivation

F25

Safety knowledge

F26

Risk preference

F27

Sensitivity to potential risks

It means workers physical condition. This factor inuences the pressure workers can endure, the
working quality, and the corresponding ability to confront risks
It means whether workers are happy or not, sometimes working with anger or sadness may result
in irrational of risk decision making (Segal et al., 2005)
It refers to workers abilities of analyzing and judging problems according to their own knowledge
and experience. This ability plays an important role within the process of risk tolerance assessment
It refers to the degree of professional knowledge will affect workers directly while dealing with
professional project issues, result with different risk tolerance (Wang and Yuan, 2011)
Rich working experiences make workers more familiar with potential risks regarding projects,
thus could likely reduce risk taking behavior
High temperature in summer and stronger wind may inuence workers judgment of their risk
tolerance (Rowlinson et al., 2014)
Loud noise from operation of machines may make workers feel dget, then, unreasonable
assessment of self-risk tolerance may happen
To what degree workers are interested in their construction job can inuence their willingness to
nish it with good quality rather take risk. Thus inuence their degree of risk tolerance
Too much work may make workers feel boring, thus want to nish the job as soon as possible,
then higher level of risk tolerance could happen
This means employers provide enough safety protection equipment and workers can get them at
any time. Enough protection equipment enhance their condence in dealing with risks
Effective and comprehensive safety training make workers aware of the seriousness and
consequences of accidents, thus lower their risk tolerance
Supervision from government or supervisor may lower workers willingness to take risks, then
lower their risk tolerance
Clear and specic safety regulation help workers realize the punishment of against safety
regulation, then lower their risk tolerance
The effect of peers behavior refers to workers would do as same as what their peer workers do. If
other workers complete work earlier by taking risks, it will enhance individuals risk tolerance to
take the same risks
Older worker seems have less willingness to take risk which resulted in lower risk tolerance
Workers with higher education tend to be more rational and cautious, while those who received
less education tend to be more fearless and impulsive
It means how serious do workers perceive the confronted risks. If workers believe the confronted
risks are more serious, they will not want to take the risks
Workers with stronger belief of they can handle most of the uncertainties would have higher
degree of risk tolerance
In some abrupt cases, quick response and decision making are required, as little time left for
thorough discussion and consideration. In these cases, workers risk tolerance vary depending on
time permission for making a decision
Relevant project information is vital in making the right decision, many risk assessment are
generally carried out with reference to the limited available historical accident data but other
considerations are ignored (Fung et al., 2012)
Boldness means the traits of being willing to undertake tasks that involve risk or danger. Decision
makers with this trait always have the ability to determine the right scheme timely and resolutely
This immediate supervisors safety attitude would have effect on workers safety risk tolerance
This means how much workers realize and emphasize on their safety
With specic decision motivation, the decision is of signicant directivity, which results in the fact
that the decision activity will move on toward expected direction and objective
The more safety knowledge the workers has, the more clear about the seriousness of risk taking in
construction project, then lower risk tolerance may happen (Pohjola, 2003)
The more willingness to take a risk, the stronger risk tolerance an individual may have (Hunter,
2006)
It refers to the ability that workers can make quick response and judgment to potential risks by
analyzing relevant information. More sensitive means the workers emphasis more on safety
issues, thus they have lower willingness to take risks (Wang, 2014)

Table 2
Ranking of critical factors affecting workers safety risk tolerance.
Factors

Mean

Standard deviation

Ranking

F19 Time constraints for decision making

F13 Safety regulations
F26 Peers behaviors
F25 Safety knowledge
F20 Completeness of relevant project information
F14 Risk preference
F11 Safety training
F04 Professional knowledge
F12 Supervision
F22 Attitude of immediate supervisor
F03 Weather
F10 Availability of safety protection equipment
F27 Sensitivity to the potential danger
F21 Working experiences

4.1414
3.7677
3.6936
3.6835
3.6800
3.5286
3.5219
3.5017
3.4579
3.3131
3.2054
3.2054
3.1825
3.1178

0.962
1.002
0.978
0.897
0.860
0.818
0.825
0.832
0.898
1.047
0.934
1.198
1.101
0.942

1
2
3
4
5
6
7
8
9
10
11
12
13
14

272

J. Wang et al. / Accident Analysis and Prevention 93 (2016) 267279

Table 3
Component matrix after varimax rotation.
Factors

Components
1

Sensitivity to the potential danger

Risk preference
Professional knowledge
Safety knowledge
Working experience
Time constraints for decision making
Completeness of relevant project information
Weather
Safety regulations
Available of safety protection equipment
Safety training
Supervision
Attitude of Immediate supervisor
Behaviors from peers

1
2
3
4

0.751
0.588
0.668
0.784
0.719
0.746
0.779
0.765
0.704
0.645
0.582
0.752

Table 4
Total variance explained for the critical factor groups.
Groups

0.732
0.669

Extraction sums of squared loadings

Total

% of variance

Cumulative (%)

6.373
4.724
3.372
2.237

25.495
17.896
13.947
8.829

25.492
43.391
57.338
66.167

belongs only to one of the four clusters generated by factor analysis, with the loading on each factor exceeding 0.50. Table 4 shows
the total variance explained, from which it can be seen that the four
components account for 66.17% of the variance.
Based on the critical factors in each group, the four components can be renamed as: (1) personal subjective perception; (2)
knowledge and experiences; (3) work characteristics; (4) safety
management, with the rst two items are internal and the latter
two are external, as shown in Fig. 5.
5. Critical paths
5.1. Hypotheses and model development
Based on the results of factor analysis, four groups, which have
effects on risk tolerance, have been identied. Two of them are characteristics of individuals which are grouped as internal factors; the

other two are external which come from the work environment that
individuals cannot control. Accordingly, four hypotheses (H1H4)
are proposed to investigate to what extent these four groups can
have effects on risk tolerance. Further, it is believed that workers with rich experiences and profession knowledge tend to be
more rational and objective when assessing their own capability to
tolerant safety risks. Thus, hypothesis (H21) regarding the relationships between the internal factors is proposed. Taking both internal
and external factors into consideration, good working environment
may shape individuals attitude and behavior (Wang and Yuan,
2011). As a result, hypotheses (H31, H32, H41, H42) regarding to
relationships between internal and external factors are developed.
All hypotheses can be described as below, and their relationships
with risk tolerance are shown in Fig. 6.
H1: Workers subjective perception has signicant effect on their
safety risk tolerance.
H2: Workers knowledge and experiences have signicant effect
on their safety risk tolerance.
H21: Workers knowledge and experiences have signicant effect
on their subjective perception.
H3: Work characteristics have signicant effect on workers safety
risk tolerance.
H31: Works characteristics have signicant effect on workers
subjective perception.
H32: Work characteristics have signicant effect on workers
knowledge and experiences.

Fig. 5. Factors and groupings inuencing workers safety risk tolerance.

J. Wang et al. / Accident Analysis and Prevention 93 (2016) 267279

273

Table 5
Cronbachs of each factor.

Fig. 6. The hypothetical safety risk tolerance model.

H4: Safety management has signicant effect on worker safety

risk tolerance.
H41: Safety management has signicant effect on worker subjective perception.
H42: Safety management has signicant effect on worker safety
knowledge and experiences.
5.2. Measurement scales
For safety risk tolerance, a self-reported assessment method
has been shown with good reliability in relation to risk tolerance
measurements (Hunter, 2002; Hallahan et al., 2004). For example,
Hunter (2002) described aviation risk scenarios in which participants were asked questions like To what extent are you willing to
accept the risks in this risk scenario? to rate the level of their risk
tolerance. Following this logic, construction safety risk scenarios
were designed for workers risk tolerance. Based on Haslam et al.
(2005), eight construction safety risk situations, which cover the
most common construction accidents and injuries (falling hazard,
electric accident, unsafe environment, etc.) were developed for risk
tolerance measurement. A ve-point Likert scale (for risk tolerance
from 1 = totally unacceptable to 5 = totally acceptable) were used for
response. Detail descriptions of questionnaire items can be found in
Appendix I. The questionnaire items for risk tolerance were coded
as Items RT01RT08.
For subjective perception, knowledge and experiences, safety
management and work characteristics, there are no measurements
scales can be directly applied. Based on the results of factor analysis, the four groups are measured by the factors that belong to them
with self-reported assessment method. All of the items are measured by a ve-point Likert scale (rating from 1 = strongly disagree
to 5 = strongly agree).
For personal subjective preference, it includes two inuencing
factors: risk preference and sensitivity to potential danger, both of
them refer to individuals reection to safety risks. Therefore, a four
item scale is designed which includes two items for risk preference,
for example, I dont like to get higher pay by nishing working
in a hurry, and two items for danger sensitivity, example item is
unsafe behaviors of other workers always remind me of checking
behaviors of myself. Details can be seen in Appendix I, the items
were coded as Item SS1 to SS4.
Regarding knowledge and experiences, Neal et al. (2000) has
developed a four items scale to assess employees knowledge about
safety practices and procedures. Following this logic, the four items
are used by adding construction safety context. Example items are

Factors

Cronbachs

Acceptable criteria

Risk tolerance
Subjective perception
Knowledge & experiences
Safety management
Working characteristics

0.856
0.713
0.707
0.809
0.711

0.600
0.600
0.600
0.600
0.600

I know how to perform my job in a safe manner, and I know how

to maintain or improve workplace health and safety. Two items
are applied to assess individual working experience. Example item
are I am very clear about all types of construction accidents, and
I have rich construction work experiences. The items were coded
as Item SK1 to SK6 in Appendix I.
In terms of work characteristics, it includes completeness of relevant project information, time constraints for decision making and
weather. Taking their meanings into account, the measurement
scale contains one item for weather (high temperature in summer and low temperature in winter will affect normal work), two
items for information (e.g., technicians always tell me all relevant
information about the work I am going to do) and two items for
decision-making time constraints (e.g., I can keep calm and behave
safely even when I am asked to nish the work as soon as possible).
Details can be seen in Appendix I, the items were coded as Item
SW1 to SW5.
Nine items are designed to assess safety management; including
two items for safety regulation, two items for safety training, one
item for supervision, two items for availability of safety protection
equipment, one item for immediate supervisors safety attitude and
one item for peers behavior. Details can be seen in Appendix I, the
items were coded as Item SM01 to SM09.
5.3. Respondents proles
After three months of controlled surveys, a total of 330 questionnaires were received, of which the valid surveys amounted to 317
(the 13 invalid ones were incompletely lled in). The respondents
proles are shown in Fig. 7.
5.4. Conrmatory factor analysis
Cronbachs was used to conrm the internal consistency reliability which reects correlations between questionnaire items
belonging to one dimension. When one questionnaire contains
more than one dimension, the internal consistency of each dimension should be respectively tested. The results are shown in Table 5.
Since 0.6 is generally accepted as the bottom line of the desired
value of internal consistency, an value lower than 0.6 indicates
unreliable questionnaire design (Flynn et al., 1994); thus, the measurement scales of the identied factors can be seen as reliable.
5.5. Structural equation modeling (SEM) analysis
The AMOS software of 18 was employed to undertake the SEM
analysis for testing the hypotheses. During the testing process, it is
found some paths are not signicant with the standardized effect
coefcients. After improving the hypothetical model by eliminating
the insignicant paths, the model showed a good t; Fig. 8 shows
the hypotheses testing results and the tness indices are listed in
Table 6.
In Fig. 8, the solid lines represent signicant links which passed
the hypotheses tests, while the dashed line means the paths failed
to pass the hypotheses tests. The numbers on the arrows represent
the path coefcients (), which refer to the change proportion of

274

J. Wang et al. / Accident Analysis and Prevention 93 (2016) 267279

Fig. 7. Participants proles.

Table 6
Fit indexes.
2 /df

RMSEA

GFI

AGFI

NFI

IFI

CFI

PGFI

2.05

0.060

0.832

0.805

0.876

0.859

0.857

0.717

dependent variable in per single unit change of the predictor (independent) variable. For example, for every single unit of increase
in workers subjective perception, the level of safety risk tolerance decreases by 0.218. The three asterisks accompanying the
path coefcients mean p < 0.001. Fig. 8 clearly illustrates the proved
hypotheses model, which is depicted in the following paragraphs.
In relation to internal factors and safety risk tolerance, Hypotheses 1 and 2 proposed that workers subjective perception and
knowledge & experiences have signicant effects on their risk
tolerance, which are supported by the results ( = 0.218 and

= 0.190, respectively, p < 0.001). Hypothesis 21 proposed that

there are existed interactions between subjective perception and
knowledge & experiences is not vailed in the test. Moving on to the
Hypotheses concerning external factors and safety risk tolerance,
Hypothesis 3, which proposed that construction working characteristics has signicant effect on risk tolerance, not supported by the
results (p > 0.05). Hypothesis 4 proposed that safety management
is signicantly associated with safety risk tolerance is validated by
the testing results ( = 0.307, p < 0.001).
As for the relationships between internal and external factors,
Hypothesis 31 proposed that the group of work characteristics is
signicantly associated with workers safety risk tolerance, which
is supported by the testing results ( = 0.199, p < 0.001). Hypothesis 32 proposed that work characteristics can inuence workers
safety risk tolerance, which is failed to pass the test in this research
(p > 0.05). Hypothesis 41 proposed that safety management has a
signicant effect on workers subjective perception, which is not
supported (p > 0.05), however, Hypothesis 42, which proposed that
safety management has signicant effect on knowledge & experiences, is supported ( = 0.523, p < 0.001).
5.6. Key inuencing paths
Based on the statements above, there exist direct and indirect
effects of identied factors on construction workers safety risk tolerance. Which one is the key path for risk tolerance formation and
how much effects the identied factors contribute to risk tolerance
are questions of great importance for understanding the formation
of risk tolerance. Table 7 shows each paths effect in the model of

Table 7
Direct and indirect effect of identied factors in risk tolerance.

Fig. 8. Modied model of construction workers safety risk tolerance. (Note: solid
line means relationship exists, and dashed line means relationship does not exist.)

Paths to risk tolerance

Effect

Subjective perception risk tolerance

Knowledge & experiences risk tolerance
Safety management risk tolerance

0.218
0.190
0.307

J. Wang et al. / Accident Analysis and Prevention 93 (2016) 267279

275

Table 8
Highest factor loadings of questionnaire items of identied factors in hypothetical model.
Groups
Subjective perception
Knowledge &
experiences
Work characteristics
Safety management

Questionnaire items

Factor loadings

SS01 I dont like to nish working in a hurry in order to get more pay
SS02 When I feel fatigue or tired, I dont want to do more work even if I could get more pay
SK05 I am very clear about all types of construction accidents
SK06 Thanks to my rich experiences, I know how to handle unforeseen accident situations
SW02 Supervisors/Technicians always tell me all specic instructions about the work I am going to do
SW03 It is necessary to attend safety meeting with managers and supervisors/technicians
SM06 There are sufcient resources available for safety in my company
SM07 The safety protection tools are available at any time I need

0.67
0.58
0.64
0.62
0.70
0.64
0.94
0.78

workers safety risk tolerance. Safety management has the largest

effect of 0.307 on risk tolerance among the factors; subjective perception comes next with a direct effect of 0.218, and the direct
effect from knowledge & experiences is 0.190, as the third. However, as shown in Fig. 6, there is also indirect effect from safety
management to risk tolerance, which is intermediated by knowledge & experiences; and from work characteristics to risk tolerance
which is intermediated by subjective perception. Taking both direct
and indirect effect into consideration, safety management has the
largest effect on workers risk tolerance, and the path of safety
management knowledge & experiences risk tolerance contributes the most to safety risk tolerance.

behaviors (Li et al., 2015). However, how does it affect behavior?

What logics were involved in the inuencing has not been clearly
revealed. Based on Fig. 9, it is known that safety management
has negative direct and indirect effects, which through inuencing workers knowledge & experiences, on risk tolerance. Then, if
workers have lower risk tolerance, more unsafe behaviors would
happen, which would result in more possibilities of accidents.
In addition, generally speaking, it is difcult to quantify the
actual effects in the causal loops within a system, especially when
the factors cannot be measured directly (Shin et al., 2014). In this
research, a new method SEM has been used to quantify such effects.
Therefore, it could be argued that it provides a method to quantify
the causal relationships in a system thinking context.

5.7. Further exploration of the questionnaire items

7. Discussion
The four groups of main inuencing factors are measured by
their own indicators (i.e., the items in the questionnaire); how
much the traits of a measuring indicator or item can be attributed to
its high-order factor is a question that will be addressed in this section. In addition the attributes are measured by actor loadings. For
example, the questionnaire Item SS01 has a factor loading of 0.67 on
subjective perception, which means 67% of the traits of Item SS01
can be explained by subjective perception. The higher the factor
loading values, the more consistent the questionnaire item is with
its high-order factor (Fang et al., 2015). This means the item with
the highest factor loadings should be given special attention, due
to it possesses the highest-level characteristics of its factor. Table 8
reveals the highest factor loadings of questionnaire items related
to each identied factor.
6. Integrating risk tolerance into the accident causal model
Fig. 9 shows a modied accident causation model, which is based
on the results of this research, and from a system thinking perspective. The combination of contributing factors of risk tolerance and
reasons behind accidents offers a big picture for the interactions
between different factors and safety performance. This circular
analysis avoids supercial linear analysis which may fail to provide
in-depth understandings of the feedback loop of accident causal
model. Specically, though workers unsafe behaviors has been recognized as the main reasons of accidents, based on the exploration
of risk tolerance, it is found that the reasons of unsafe behavior
come from four groups of factor, namely safety management, work
characteristics, knowledge & experiences and personal subjective
perception.
It should be pointed out that the feedback loops shown in Fig. 9
also mean that the system output (i.e., safety performance in this
case) also affects the system inputs (i.e., safety management and
work characteristics). This circular analysis means it is difcult
to clearly differentiate which factors are absolute root causes but
instead they have a circular effect to each other.
The introduction of risk tolerance into accident causal system offers detail explanation of accidents. For example, deciency
safety management is known to have effects on workers safety

7.1. Critical inuencing factors and paths

Factors identied in this research play different roles in contributing toward safety risk tolerance. In terms of internal factors,
individuals subjective perception is negatively associated with
their safety risk tolerance. Specically, sensitivity to potential danger is highly related to workers safety awareness, which can help
workers avoid unnecessary injuries by reminding them the existences of safety risks. Basically, workers with more sensitivity to
potential danger are more likely to behave carefully and take more
active attitudes toward identifying and handling confronted safety
risks, rather than just do it without taking the potential outcomes into consideration. As for risk preference, risk averters may
feel more danger and care more about negative effects than risk
seekers, and as a result, less willingness to gain risky benets by
compromising or losing something (Brockhaus, 1980). Thus, lower
risk tolerance occurs.
The uniqueness and non-repeatability of each construction
project make site experiences play a great role in accident prevention. Professional knowledge and experiences not only can provide
workers quality professional skills, but also clear understanding of
the seriousness and possible losses of taking risks, and the limitation of their capability to handle such serious safety risks. Therefore,
workers with rich experiences and professional knowledge are
more rational and less likely to increase risk tolerance when confronted with safety risks.
From the perspective of external factors, construction working
characteristics showed insignicant effect on workers risk tolerance. In contrast, safety management shows the largest effect on
workers risk tolerance. Specically, safety regulation and supervision are compulsory ways, generally with harsh punishments for
unsafe behaviors, and thus, lower safety risk tolerance happens
due to unwillingness to be punished. On the contrary, availability of safety protection equipment and safety training are soft ways
to develop higher safety awareness. More safety protection equipment and targeted safety training bring safety awareness of that the
work they are doing. As a result, the assessment of self-capability
to respond to safety risks tends to be more rational and objective.

276

J. Wang et al. / Accident Analysis and Prevention 93 (2016) 267279

Fig. 9. A modied accident causal model.

Effects from co-workers behaviors are also of great importance.

For instance, when workers nd peers completed more works or
doing faster than them, willingness to take risks may increase due
to they want to doing better and getting more pay than others. As
for attitude of immediate supervisor, Huang et al. (2004) indicated
that managers with higher safety conscious tend to talk more and
supervise more carefully with workers. Since immediate supervisors are the ones that have the most contact with workers, to what
extent and how they emphasize on safety issues contribute a lot to
workers risk tolerance.
7.2. Implications
7.2.1. Theoretical implications
It is noticeable that among the identied critical inuencing factors, there are more external factors (such as safety regulation,
safety training, peers behavior, etc.) than internal factors (such
as professional knowledge, risk preference, sensitivity to potential
risks, etc.) making contribution to workers safety risk tolerance.
This nding is different from the studies of inuencing factors of
risk attitude by Wang and Yuan (2011), who claimed individuals
internal factors (personal traits, characteristic, motivation, etc.)
play a dominant role in their attitude toward risks. Thus, risk tolerance and risk attitude, both as the subjective part of an individual,
affected by factors belong to different aspects. The reasons behind
this difference may result from the meaning of these two conceptions. Risk tolerance focuses on the amount of potential losses a
person can bear, which means it depends on individuals resources
that can be used for overcoming risks. It reects more about the
capability to respond and bear losses. However, risk attitude is
more like personal preference, which emphasizes decision makers
willingness to tolerant potential losses.
Exploration of workers safety risk tolerance contributes to the
body of knowledge of systematic safety risk management in terms
of human characteristics. Little is currently known about the contributing factors of individuals internal factors. In order to deeper
understand the reasons behind why different decision makers
behave differently when confronting the same situation, the trigger
behind their emotions, feelings must be dig out and quantied. This
research analyzed the critical factors and paths inuencing workers safety risk tolerance, which made safety risk management more
complete by providing the causes of individuals making different
assessments of their risk tolerance.
7.2.2. Practical implications
There are several practical implications to ensure the effectiveness of accident prevention, as discussed below:
First, a specic safety risk management plan should be put in
place that pays very detail attention to risk tolerance. It should
emphasize that risk tolerance not only specic to workers, but also

to project managers and stakeholders. The ndings of this research

revealed that individuals risk tolerance are changeable within
internal and external factors. Thus, though early and comprehensive assessment of risk tolerance is complicated and burdensome, it
should be conducted carefully due to dividends would be paid later
as the projects issues become larger and more complex (Kwak and
Laplace, 2005).
Second, it is vital for managers to reduce workers safety risk
tolerance by adopting targeted strategies. From the macro point
of view, more attention should be paid on external factors (safety
management and working characteristics) due to their effects on
risk tolerance are larger than that of internal factors (subjective
perception and knowledge & experiences). From the micro point of
view, the items with higher factor loading need to be focused on.
For example, Items of SS01 (I dont like to get higher pay by nishing
working in a hurry.) and SSO2 (I dont want to do more work even
I could get higher pay when I feel fatigue.), which represent individuals risk preference, have higher factor loadings on subjective
factors than other factors, therefore, assessment of workers risk
preference is one of the key tasks for project managers. Based on
the meaning of the Items with higher factor loading in each group,
the strategies for regulating workers safety behaviors should be:
(1) assessment of workers risk preference; (2) assessment of workers working experiences; (3) providing complete information of
specic working task; (4) providing fully access to safety protection equipment; and (5) implementation of safety regulation and
training.
While practitioners safety risk tolerance is at the heart of risk
decision making, it is too often overlooked in the process of risk
management, because it is ever changing with different inuential
factors. This concept is coming to grips and this paper provides
practical ndings for construction safety risk management.

7.3. Future research

The research proposed a useful method to investigate the critical inuencing factors and paths of workers risk tolerance. What
should be noted is that the empirical data were collected from
construction sites in China; whether workers risk tolerance still
shows the same characteristics under different national culture
background is a question that needs to be answered in future
research.
It is notable that risk tolerance is just one of the internal factors, and this research is a beginning in improving systematic safety
risk management. In the next step, more internal factors need to
be explored; then, full understanding of the formation of safety
behavior by integrating risk tolerance and other important subjective factors such as risk perception and risk attitude can be obtained
by quantifying the effects of these factors have on safe behaviors,

J. Wang et al. / Accident Analysis and Prevention 93 (2016) 267279

and develop system dynamics model to simulate workers safety

behavior.
8. Conclusion
Construction workers risk tolerance are inuenced and determined by many factors belong to different domains. It is neither
practical nor necessary to identify and understand all related factors for minimizing the bias in risk decision making due to limited
management resources such as money, time, manpower, etc. In this
study, through the statistical ranking analysis, 14 factors are identied as the critical ones. Among them, there are nine related to
external environment, while ve depend on workers internal characteristics. Further, factor analysis provided a grouping of these 14
factors into: (1) personal subjective perception; (2) knowledge and
experiences; (3) work characteristics; and (4) safety management.
Afterwards, the critical inuencing paths are identied by quantifying the effects on the four groups, which contain the direct and

under 30
less than 5 years
Junior High school
Formwork

Age:
Working experiences
Highest Education
Type of work:

277

the indirect effects. Safety management is found to be a dominator.

After introducing the critical inuencing factors and paths of risk
tolerance into the accident causal model, it revealed the root causes
of unsafe behaviors.
In conclusion, workers safety risk tolerances are inuenced by
many factors, some are internal, others external, and these factors
are interrelated with each other. Among all, implementing a rigorous safety management system is most important while attention
should also be paid to personal subjective perception.
Acknowledgement
This research was supported by National Natural Science Foundation of China (project ID 71272088).
Appendix I. Questionnaires
Part I Background Information (please tick
applies to you)

3040
510 years
Senior High School
Earth work

4050
1020 years
College
Steel xing

Part II Factors (please tick

Factor

Risk tolerance

in the that applies to you)

RT1
RT2

A worker steps to an open-sided oor.

A worker does not wear safety helmet when he is working in the
second oor.
There is no safety sign or safety net around the reserved elevator hole.
There is inadequate personal protective equipment.
Safety nets do not cover the building when construction is in progress.
When excavation work is ongoing, the soils are placed at the edge of
the foundation pit.
Improper guards in place on power tool management resulted in easy
body contact with power cables.
The crane coordination course does not include real trainingit was
just heres the form, ll it out and sign it.

1
1

2
2

3
3

4
4

5
5

1
1
1
1

2
2
2
2

3
3
3
3

4
4
4
4

5
5
5
5

RT3
RT4
RT5
RT6

Codes

Statement

Strongly disagree

Disagree

Neutral

Agree

Strongly agree

SS01

I dont like to get more pay by nishing

working in a hurry.
I dont want to do more work even if I could
get more pay when I feel fatigue and tired.
I always care about the unsafe behaviors of
other workers.
Before an accident happens, I always notice
something related to the danger.

I know how to perform my job in a safe

manner.
I know how to maintain or improve the health
and safety of workplace.
When I confronted with an accident or injury, I
can usually nd some solutions.
Professional knowledge is helpful in protecting
me from danger.
I am very clear about all types of construction
accidents.
Thanks to my rich experiences, I know how to
handle unforeseen accident situations.

High temperature in summer and low

temperature in winter will affect normal work.
Site supervisors and technicians always tell me
all specic instructions about the work I am
going to do.

SS02

SS04
SK01
SK02
SK03
SK04
SK05
SK06
SW01
Work nature and
characteristics

The amount of risk you can accept is

increasing from 1 to 5, 1 means minimum,
5 means maximum.

SS03

Knowledge and
experiences

51 above
21 years and above
University and above
Interior Decoration

Construction safety scenario

RT8

Personal Subjective
perception

in the that

Codes

RT7

Factor

SW02

278
Factor

J. Wang et al. / Accident Analysis and Prevention 93 (2016) 267279

Codes

Statement

Strongly disagree

Disagree

Neutral

Agree

Strongly agree

SW03

It is necessary to attend safety meeting with

managers and supervisors/technicians.
I can keep calm and behave safely even when
asked to nish the work as soon as possible.
It would cause deciency in quality if I am
asked to crash the work schedule.

SW04
SW05
SM01

Safety
management

SM02

SM03
SM04
SM05
SM06
SM07
SM08
SM09

When I saw unsafe behavior of my co-workers,

the safety regulations would remind me to
stop it and educate them.
When I face safety technical problem, there is
always relative safety operation guides for
reference.
Project managers place a high priority on
safety training.
The safety training covers all risks associated
with the work for which I am responsible.
There are specic managerial staff who are
responsible for safety work checking.
There are sufcient resources available for
safety in my company.
The safety protection tools are available at any
time I need.
If I ask for advice on safety matters, my
immediate supervisor would be very helpful
I will tell and correct unsafe behaviors of my
colleagues.

References
Abdelhamid, T.S., Everett, J.G., 2000. Identifying root causes of construction
accidents. J. Constr. Eng. Manage. 126, 5260.
Acar, E., Gc, Y., 2011. Prediction of risk perception by owners psychological traits
in small building contractors. Constr. Manage. Econ. 29, 841852.
Balaz, V., Williams, A.M., 2011. Risk attitudes and migration experience. J. Risk Res.
14, 583596.
Brockhaus, R.H., 1980. Risk taking propensity of entrepreneurs. Acad. Manage. J.
23, 509520.
Burton, M., 1996. A Random Walk Down Wall Street. W. W. Norton, New York.
Cordell, D.M., 2002. Risk tolerance in two dimensions. J. Finan. Plann. 15,
3033.
De Camprieu, R., Desbiens, J., Feixue, Y., 2007. Culturaldifferences in project risk
perception: an empirical comparison of China and Canada. Int. J. Proj. Manage.
25, 683693.
Dixit, V., Harrison, G.W., Rutstr, M,E.E., 2014. Estimating the subjective risks of
driving simulator accidents. Accid. Anal. Prev. 62, 6378.
Fang, D., Wu, C., Wu, H., 2015. Impact of the supervisor on worker safety behavior
in construction projects. J. Manage. Eng.
Fang, M.Z.D., 2012. Cognitive causes of construction workers unsafe behaviors and
management measures. China Civil Eng. J. 45, 4549.
Flynn, B.B., Schroeder, R.G., Sakakibara, S., 1994. A framework for quality
management research and an associated measurement instrument. J. Oper.
Manage. 11, 339366.
Frank, W., 2011. Challenges in developing and implementing safety risk tolerance
criteria. Process Saf. Prog. 30, 232239.
Fung, I.W., Lo, T.Y., Tung, K.C., 2012. Towards a better reliability of risk assessment:
Development of a qualitative & quantitative risk evaluation model (Q 2 REM)
for different trades of construction works in Hong Kong. Accid. Anal. Prev. 48,
167184.
Garrett, J., Teizer, J., 2009. Human factors analysis classication system relating to
human error awareness taxonomy in construction safety. J. Constr. Eng.
Manage. 135, 754763.
Goh, Y.M., Brown, H., Spickett, J., 2010. Applying systems thinking concepts in the
analysis of major incidents and safety culture. Saf. Sci. 48, 302309.
Gould, F.E., Joyce, N.E., 2003. Construction Project Management. Prentice Hall,
Upper Saddle River, NJ.
Grable, J.E., 2000. Financial risk tolerance and additional factors that affect risk
taking in everyday money matters. J. Business Psychol. 14, 625630.
Hair, J.F., 2010. Multivariate Data Analysis.
Hallahan, T.A., Faff, R.W., Mckenzie, M.D., 2004. An empirical investigation of
personal nancial risk tolerance. Finan. Serv. Rev.-Greenwich 13, 5778.
Hallowell, M., 2010. Safety risk perception in construction companies in the Pacic
Northwest of the USA. Constr. Manage. Econ. 28, 403413.
Haslam, R., Hide, S., Gibb, A.G., Gyi, D.E., Pavitt, T., Atkinson, S., Duff, A., 2005.
Contributing factors in construction accidents. Appl. Ergon. 36, 401415.
Heinrich, H.W., Petersen, D., Roos, N., 1950. Industrial Accident Prevention.
McGraw-Hill, New York.
Hopkinson, M.M., 2012. The Project Risk Maturity Model: Measuring And
Improving Risk Management Capability. Gower Publishing, Ltd.
Huang, Y.-H., Chen, P.Y., Krauss, A.D., Rogers, D.A., 2004. Quality of the execution of
corporate safety policies and employee safety outcomes: assessing the

moderating role of supervisor safety support and the mediating role of

employee safety control. J. Business Psychol. 18, 483506.
Hunter, D.R., 2002. Risk Perception and Risk Tolerance in Aircraft Pilots. DTIC
Document.
Hunter, D.R., 2006. Risk perception among general aviation pilots. Int. J. Aviat.
Psychol. 16, 135144.
ILO, 2006. C187 Promotional Framework for Occupational Safety and Health.
Irwin Jr., C.E., 1993. Adolescence and Risk Taking: How Are They Related.
ISO, I., 2009a. 31000: 2009 Risk Management Principles and Guidelines.
International Organization for Standardization, Geneva, Switzerland.
ISO, I., 2009b. IEC 31010Risk Management: Risk Assessment Techniques.
International Organization for Standardization.
Kahneman, D., Lovallo, D., 1993. Timid choices and bold forecasts: a cognitive
perspective on risk taking. Manage. Sci. 39, 1731.
Karimiazari, A., Mousavi, N., Mousavi, S.F., Hosseini, S., 2011. Risk assessment
model selection in construction industry. Expert Syst. Appl. 38, 91059111.
Kwak, Y.H., Laplace, K.S., 2005. Examining risk tolerance in project-driven
organization. Technovation 25, 691695.
Leung, M.-Y., Chan, I.Y.S., Yu, J., 2012. Preventing construction worker injury
incidents through the management of personal stress and organizational
stressors. Accid. Anal. Prev. 48, 156166.
Leveson, N., 2004. A new accident model for engineering safer systems. Saf. Sci. 42,
237270.
Lichtenstein, S., Slovic, P., Fischhoff, B., Layman, M., Combs, B., 1978. Judged
frequency of lethal events. J. Exp. Psychol. [Hum. Learn]. 4, 551.
Liu, J., Meng, F., Fellows, R., 2014. An exploratory study of understanding project
risk management from the perspective of national culture. Int. J. Proj. Manage.
Moreschi, R., 2004. Incorporating investor risk tolerance into the nancial planning
process. J. Pers. Finance 3, 8998.
Mu, S., Cheng, H., Chohr, M., Peng, W., 2014. Assessing risk management capability
of contractors in subway projects in mainland China. Int. J. Proj. Manage. 32,
452460.
Neal, A., Grifn, M.A., Hart, P.M., 2000. The impact of organizational climate on
safety climate and individual behavior. Saf. Sci. 34, 99109.
Olson, R., Austin, J., 2001. ABCs for lone workers: a behavior-based study of bus
drivers. Prof. Saf. 46, 20.
Pohjola, V.J., 2003. Fundamentals of safety conscious process design. Saf. Sci. 41,
181218.
Raheem, A.A., Hinze, J.W., 2014. Disparity between construction safety standards:
a global analysis. Saf. Sci. 70, 276287.
Rawlinson, F., Farrell, P., 2009. The vision of zero risk tolerance in craft workers
and operatives: an unattainable goal. In: Proceedings of the 25th Annual
ARCOM Conference, Association of Researchers in Construction Management,
Reading, pp. 12031212.
Roszkowski, M.J., Davey, G., 2010. Risk perception and risk tolerance changes
attributable to the 2008 economic crisis: a subtle but critical difference. J.
Finan. Serv. Prof. 64, 4253.
Rowlinson, S., Yunyanjia, A., Li, B., Chuanjingju, C., 2014. Management of climatic
heat stress risk in construction: a review of practices, methodologies, and
future research. Accid. Anal. Prev. 66, 187198.
Sec, S.A.E.C., 2010. Beginners Guide to Asset Allocation, Diversication, and
Rebalancing. SEC Pub. No.062(2/10), Available at https://www.my-benetinfo.com/pdf/Beginners Guide.pdf.

J. Wang et al. / Accident Analysis and Prevention 93 (2016) 267279

Segal, G., Borgia, D., Schoenfeld, J., 2005. The motivation to become an
entrepreneur. Int. J. Entrepren. Behav. Res. 11, 4257.
Shin, M., Lee, H.-S., Park, M., Moon, M., Han, S., 2014. A system dynamics approach
for modeling construction workers safety attitudes and behaviors. Accid. Anal.
Prev. 68, 95105.
Wang, 2014. A critical review on risk decision-making and its attributes. J. Eng.
Manage. 28, 2731.

279

Wang, J., Yuan, H., 2011. Factors affecting contractors risk attitudes in construction
projects: case study from China. Int. J. Proj. Manage. 29, 209219.
Zou, P.X., Sunindijo, R.Y., 2013. Skills for managing safety risk, implementing safety
task, and developing positive safety climate in construction project. Autom.
Constr. 34, 92100.