USA: +1(669)289-8683 
Sign in
mos6625unit8peerreviewedarticles.zip
Unit VIII Assignment.
Operations Safety Management Plan
· Review the information in your textbook (Leveson, 2011, pp. 412-414) related to the Operations Safety Management Plan design.
· Review the Figure C.1 depiction of a sample Safety Control Structure in your textbook (Leveson, 2011, p. 500).
· Review and use a minimum of five (5) peer-reviewed journal articles from the CSU Online Library (Academic Search Complete) to support your design work.
Then, select one (1) of the following options:
Option 1: Design an Operations Safety Management Plan for your own organization or an organization that is familiar to you.
Option 2: Design an Operations Safety Management Plan for a bulk tank railcar off-loading facility for hydrocarbon products that has the following features:
a. one railcar switch located next to an interstate highway,
b. capacity to off-load both liquid hydrocarbon products and liquefied, compressed, natural gas products,
c. two 500,000-gallon bulk liquid storage tanks for liquid hydrocarbon products,
d. two 45,000-gallon bullet-styled, horizontal, liquid storage tanks for liquefied, compressed natural gas products,
e. one off-loading station (single-sided) that is elevated 12 feet from the ground,
f. one flare for over-pressurized gas tank railcars,
g. one 3,000-gallon condensation storage tank, and
h. one switch engine for staging railcars at the off-loading station and at railcar storage tracks.
Using the CSU APA-styled paper as a formatting template, design a minimum eight-page Operations Safety Management Plan with a minimum of five (5) scholarly sources (books and articles) using the following APA Level 1 headings:
1. General Considerations
2. Safety Organization
3. Procedures
4. Schedule
5. Safety Information System
6. Operations Hazard Analysis
7. Evaluation and Planned Use of Feedback for System Maintenance
8. Safety Control Structure Diagram (Design a Safety Control Structure Diagram for your work system and use as the content.)
Integration of safety risk data
References
Esmaeili, B., & Hallowell, M. (2013). Integration of safety risk data with highway construction schedules.Construction Management & Economics, 31(6), 528-541. doi:10.1080/01446193.2012.739288
Abstract:
The construction industry is characterized by a relatively high injury and illness rate compared to other industries. Within theconstruction industry, the highway construction and maintenance sector is one of the most dangerous. To improve safety in this sector, proactive methods of safety improvement and reliable risk data are needed. The safety risk quantification is the first step towards integrating safety data into design and planning. To enhance the current preconstruction safety practices,safety risks of highway construction and maintenance tasks were quantified and a decision support system was developed and tested that integrates safety risk data into the project schedules. Relative safety risks were quantified for 25 common highway construction tasks using the Delphi method. To ensure valid and reliable results, experts were selected according to rigorous requirements and multiple controls were employed to decrease cognitive biases. The data were incorporated into a decision support system called Scheduled-based Safety Risk Assessment and Management (SSRAM) that facilitates integration of safety risk data with project schedules. The resulting data-driven system produces predictive plots of safetyrisk over time based on the temporal and spatial interactions among concurrent activities. To test the utility of the decision support system and the validity of the underlying risk data, the system was tested on 11 active case study projects in the US. It was found that the database and associated decision support tool produce accurate and reliable risk forecasts that increase the viability of existing safety preconstruction activities. [ABSTRACT FROM AUTHOR
Intergration of safety risk data with highway construction schedules
Integration of safety risk data with highway
construction schedules
BEHZAD ESMAEILI* and MATTHEW HALLOWELL
Department of Civil, Environmental, and Architectural Engineering, University of Colorado at Boulder, 428 UCB,
1111 Engineering Drive, Boulder, CO 80303, USA
Received 29 March 2012; accepted 9 October 2012
The construction industry is characterized by a relatively high injury and illness rate compared to other
industries. Within the construction industry, the highway construction and maintenance sector is one of the
most dangerous. To improve safety in this sector, proactive methods of safety improvement and reliable risk
data are needed. The safety risk quantification is the first step towards integrating safety data into design
and planning. To enhance the current preconstruction safety practices, safety risks of highway construction
and maintenance tasks were quantified and a decision support system was developed and tested that inte-
grates safety risk data into the project schedules. Relative safety risks were quantified for 25 common
highway construction tasks using the Delphi method. To ensure valid and reliable results, experts were
selected according to rigorous requirements and multiple controls were employed to decrease cognitive
biases. The data were incorporated into a decision support system called Scheduled-based Safety Risk
Assessment and Management (SSRAM) that facilitates integration of safety risk data with project schedules.
The resulting data-driven system produces predictive plots of safety risk over time based on the temporal
and spatial interactions among concurrent activities. To test the utility of the decision support system and
the validity of the underlying risk data, the system was tested on 11 active case study projects in the US. It
was found that the database and associated decision support tool produce accurate and reliable risk forecasts
that increase the viability of existing safety preconstruction activities.
Keywords: Decision support systems, occupational health and safety, risk management, scheduling.
Introduction
Typically, safety management activities take place
during the construction phase (e.g. job hazard analy-
ses and site audits). In recent years, new safety man-
agement strategies have been introduced that help the
project team to identify and control hazards during
design and preconstruction. However, according to
Szymberski (1997), the potential to influence site
safety and health conditions decreases exponentially
as the project commences. Recent research has con-
firmed these findings and indicates that the most
effective safety programme elements occur during the
programming and preconstruction phases (Rajendran
and Gambatese, 2009). Unfortunately, the current
methods for considering safety and health in these
early phases are inconsistent, informal, and based
primarily on intuition and judgment (Hallowell,
2008). Thus, there is clearly a need to enhance pre-
construction safety management strategies, to create
user-friendly tools, and to increase their use in all
sectors of the industry.
One of the preconstruction methods that has been
shown to be highly effective is the integration of the
safety aspect into project schedules using risk data
(Yi and Langford, 2006). Unfortunately, integration
is limited because of a lack of data for specific con-
struction work tasks and the lack of reliable tools that
interface with existing scheduling software. The cur-
rent study aims to test the theory that loading safety
risk data into the project schedule is practical and will
improve predictions of high risk work periods. The
objectives are to (1) quantify relative safety risk values
for common highway construction activities; (2)
*Author for correspondence. E-mail: behzad.esmaeili@colorado.edu
Construction Management and Economics, 2013
Vol. 31, No. 6, 528–541, http://dx.doi.org/10.1080/01446193.2012.739288
� 2013 Taylor & Francis
integrate these risk data into project schedules using a
novel decision support system; and (3) validate the
analytical procedure on case study projects.
This study focuses on risk quantification and risk
modelling for highway construction because the high-
way construction sector is one of the most dangerous
in the industry (Bureau of Labor Statistics, 2012). In
2005, this sector accounted for approximately 469
vehicle- and mobile heavy equipment-related deaths,
279 of which (59%) occurred in traffic work zones
(Center for Construction Research and Training,
2008). Furthermore, the Federal Highway Adminis-
tration (2004) estimates that a work zone fatality
occurs once every 10 hours and a work zone injury
occurs every 13 minutes. The presence of high-speed
traffic near work zones, prevalence of night-time
work, use of heavy equipment, exposure to weather,
and highly repetitive work tasks contribute to this rel-
atively high number of injuries (Bryden and Andrew,
1999; Arditi et al., 2005). Thus research is needed to
help practitioners to identify, analyse and respond to
high risk periods on highway construction projects.
Literature review
This study was guided by a large body of literature.
In particular, literature that focused on the safety-
schedule integration and construction engineering and
management (CEM) decision support systems (DSSs)
proved to be most helpful. This body of literature was
used to guide the risk quantification process and the
development of a framework for integrating safety risk
into project schedules. A review of the salient findings
from relevant literature is provided below.
Safety schedule integration
Integrating safety planning and management in early
phases of construction projects is essential to effective
injury prevention and the development of a culture of
safety (Tarrants, 1980; Sawacha et al., 1999). Coble
and Elliott (2000) argued that integration of safety
aspects into planning starts with considering safety
during the scheduling of a construction project. There
have been a multitude of studies that attempt to inte-
grate various forms of safety information with project
schedules. These studies can be divided into two gen-
eral categories: those that attempted to cover safety
planning, injury prevention and regulatory informa-
tion, and those that integrated risk data. The majority
of studies focused on the former because these safety
data, such as regulatory information, are readily
available and not difficult to obtain.
Safety-schedule integration began with the work of
Kartam (1997) who designed a framework for
integrating extensive safety knowledge (e.g. Occupa-
tional Safety and Health Administration (OSHA)
regulations) into critical path method (CPM) sched-
ules using Microsoft Project, Primavera P6, Primavera
Suretrack and Timeline. According to Hinze et al.
(2005) the major weakness of this initial effort was
that there was never any success in making a link
between the safety elements and the electronic sche-
dule. In response to this shortcoming, Hinze et al.
(2005) built upon this research effort by developing
SalusLink, a tool that allows project managers to
access textual safety data contained in databases
managed by Primavera P6 and Suretrack. Though a
working prototype was produced, the software is not
commercially available. Saurin et al. (2004) and
Cagno et al. (2001) took a different approach by
developing safety planning and control models that
attached injury prevention strategies and methods of
safety planning to scheduled activities.
In the past five years, researchers have attempted
to integrate risk data into project schedules as a
means to identify high risk work periods and
leverage scheduling controls to prevent periods of
excessive risk. For example, Wang et al. (2006)
developed a simulation-based model (SimSAFE) that
integrates expected injury cost data for each activity
in a network schedule. This stand-alone software
system allows safety managers to identify work zones
that are associated with relatively high risk as mea-
sured by cumulative potential accident costs. Yi and
Langford (2006) took risk integration a step further
by developing a robust framework for ‘safety
resource scheduling’ using patterns which are similar
to resource levelling. Although Yi and Langford
(2006) offered a strong framework for the integration
of safety risk data with project schedules, there were
no robust risk data as the database only included
fatalities that occurred as a result of falls from
height. Furthermore, Navon and Kolton (2006,
2007) created an automated monitoring and control
model that is capable of identifying fall hazards and
their location. The major limitations of this body of
literature are that there is not a robust safety risk
database and the interactions (i.e. compatibility and
incompatibility) among tasks were ignored.
Researchers have begun to model the interactions
among risk factors and create frameworks that inte-
grate detailed user-provided data into preconstruction
planning tools. For example, a series of studies mod-
elled the spatial and temporal interactions of concur-
rent work tasks by using information available in 4D
geographic models and user-provided data for ‘loss-of-
control events’ (Rozenfeld et al., 2009, 2010; Sacks
et al., 2009). The major limitation of these models is
that hazards related to each task must be identified
Safety risk data 529
and quantified by the user, which can be time inten-
sive, laborious and unrealistic in practice (Rozenfeld
et al., 2009). In another study, Hallowell et al. (2011)
adapted Yi and Langford’s (2006) model and sug-
gested a new framework to integrate safety risk data
into project schedules. In addition to integrating base-
level risks for individual tasks, this framework also
considered robust task interactions obtained through
the Delphi process. The limitation of this work was
that they did not test the applicability of the framework
on actual projects and base-level task risks were not
quantified. Thus, the current research aims to address
the limitations of the previous studies by quantifying
highway construction safety risks for common work
tasks and testing the efficacy of the framework
presented by Hallowell et al. (2011) on active projects.
CEM decision support systems
As computing technologies have improved, increased
attention has been paid to the development of com-
puter applications that increase the speed and quality
of decision making. One category of these tools is
decision support systems (DSSs), which are defined
as, ‘an interactive IT-based system that helps decision
makers utilize data and models in making their deci-
sions’ (Carter et al., 1992, p. 3). Typically, the two
main objectives for using DSSs are: performing a
given task in the decision-making process more
quickly and with fewer resources (efficiency); and
improving the quality of the outcome of the decision-
making process (effectiveness). In addition, DSSs
help a manager to make more informed decisions,
consider a multitude of criteria and alternatives,
reduce the time needed to make an effective decision,
and focus attention on the most important elements
of a scenario. They also reduce complexity of the
problem to a manageable level and reduce uncertainty
(Carter et al., 1992). In CEM, DSSs have been uti-
lized in many areas such as: resource sharing (Perera,
1983); prequalifying subcontractors (Russell et al.,
1990); optimizing heavy lift planning (Lin and Hass,
1996); resource levelling (Leu et al., 2000); making
go/no-go decisions for international projects (Han and
Diekmann, 2001); selecting appropriate project deliv-
ery methods (Molenaar and Songer, 2001); schedul-
ing steel fabrication (Karumanasseri and AbouRizk,
2002); and providing guidance during dispute resolu-
tion (Palaneeswaran and Kumaraswamy, 2008).
In addition to the applications mentioned above,
some DSSs have been developed to enhance decision
making in the area of safety. For example, Kak et al.
(1995) developed a knowledge-based program to
facilitate access to the explicit safety knowledge on
construction sites. Their program searches applicable
safety regulations (e.g. OSHA) for a particular task
and provides suggestions to improve compliance.
Gambatese et al. (1997) presented a tool for incorpo-
rating safety-related issues in the design phase of a
project called ‘Design for Construction Safety Tool-
Box’, which has the ability to identify project-specific
hazards and provide design suggestions to mitigate
those hazards. Recently, Hadikusumo and Rowlinson
(2004) applied a visual reality concept to develop a
design for safety tool to capture tacit knowledge of
safety professionals. The results contribute to the
current arsenal of CEM and safety tools by providing
an applied DSS that integrates safety risk data into
project schedules.
Point of departure
To build upon safety risk management research, the
relative safety risk of common highway reconstruction
tasks was assessed and the efficacy of a DSS that inte-
grates safety risk data into project schedules was
tested. A thorough review of relevant literature
revealed no study that has directly quantified highway
reconstruction safety risks or attempted to assess tem-
poral models for safety risk integration using actual
data. It is expected that the findings presented will
aid project managers in their preconstruction safety
management activities and will be especially effective
for safety managers who are responsible for multiple
concurrent projects.
Research methods
The research objectives were achieved in three distinct
phases. In the first phase, the Delphi method was
employed to quantify relative safety risks. In the
second phase, a graphical user interface was developed
in MATLAB, called Scheduled-based Safety Risk
Assessment and Management (SSRAM), that is capa-
ble of creating temporal safety risk profiles for highway
construction projects. Finally, the output of the system
was validated by employing a Multi-Attribute Utility
Assessment (MAUA) technique and conducting 11
case studies. The following sections discuss the details
of the research methods employed in these three
phases.
Phase I method: risk quantification
In order to develop an appropriate scope for data col-
lection, clear definitions of common highway con-
struction work tasks were needed. Therefore, the 25
highway tasks identified and described by Pandey
530 Esmaeili and Hallowell
(2009) and refined by Hallowell et al. (2011) were
used as a foundation. To quantify the relative risk val-
ues for these tasks, the Delphi method was selected.
The traditional paradigm in risk quantification
adopted by Brauer (1994) and Hallowell and
Gambatese (2009) was followed where frequency and
severity ratings for each task are solicited from an
expert panel through multiple rounds of surveys and
controlled feedback.
The Delphi method was chosen for obtaining
safety risk values for six main reasons. First, there
were no objective highway repair and maintenance
safety risk data available from government databases.
The common national databases such as Occupa-
tional Safety and Health Administration Integrated
Management Information System (OSHA IMIS,
n.d.) and National Institute of Occupational Safety
and Health Fatality Assessment and Control Reports
(NIOSH FACE, n.d.) include only high severity
injuries and do not provide enough information
regarding the task performed when the injury
occurred. Second, according to Gyi et al. (1999), the
validity of statistical data obtained from accident
reports is significantly compromised by underreport-
ing, especially for minor injuries. Third, Snashall
(1990) stated that accident report processes are not
consistent between and within companies
(e.g. definition of construction activities) such that
empirical data cannot be easily interpreted and com-
pared. Fourth, accidents happen in a complex sys-
tem created by interrelated worksite characteristics
that cannot be separated from the project context
(Mitropoulos et al., 2005). Fifth, according to Dijk-
sterhuis et al. (2006), intuitive decision processes like
Delphi that use heuristic principles lead to accurate
risk estimates in complex scenarios. Finally, Delphi
is a rigorous process that allows researchers to obtain
unbiased data using the judgment of qualified
experts, which has been used successfully for risk
quantification in similar studies (e.g. Hallowell and
Gambatese, 2009; Hallowell et al., 2011).
The Delphi method was developed by Rand
Corporation for the US Air Force in late 1940s to
elicit reliable and unbiased judgments from a group of
experts by conducting an iterative process and
providing controlled feedback (Helmer, 1967; Lin-
stone and Turoff, 1975). The Delphi method involves
assembling qualified experts, developing appropriate
questionnaires, and conducting multiple rounds of
surveys with controlled feedback between rounds to
achieve consensus (Cabaniss, 2001; Hallowell and
Gambatese, 2010). This method is applied under the
assumption that the collective expertise of the panel is
superior to the judgment of individuals (Hogarth,
1978; Boje and Murnighan, 1982; Hill, 1982).
The Delphi process was conducted in two rounds
where expert panellists were asked to provide inde-
pendent frequency and severity ratings for each of the
25 highway construction tasks. In order to maintain
consistency, the authors have adopted an objective
risk scale created by Hallowell and Gambatese (2009)
that incorporates a complete spectrum of frequency
and severity scales (see Table 1). The severity scale
ranges from negligible injury to fatality and the fre-
quency scale ranges from one incident occurrence
every six minutes (0.1 w-h) to one incident occur-
rence every 100 million or more worker-hours (>100
million w-h). After the first round of surveys, the data
were aggregated and the level of consensus was mea-
sured and evaluated. In the second round, panellists
were asked to review the median responses from the
first round and provide final ratings. As will be dis-
cussed, a third round of data collection was not
needed because the target consensus was achieved in
the second round.
Selection of expert panellists
As the number of panellists in a Delphi study
increases, the accuracy of the results also tends to
increase (Murphy et al., 1998). In a review of past
Delphi studies, Rowe and Wright (1999) found that
the number of panellists has ranged from 3 to 80. As
noted by Linstone and Turoff (1975) factors such as
the expected volume of the data, time constraints,
and the number of experts available can affect the
appropriate number of panellists. A relatively large
panel was desired to quantify risks for the 25 highway
construction tasks that can be performed in a variety
of work environments.
Careful attention was paid to ensure that all
panellists were highly qualified. The expert panel was
Table 1 Frequency and severity scales (adapted from
Hallowell and Gambatese, 2009)
Frequency Severity
Worker hours per incident Subjective level Score
>100 million Temporary discomfort 2
10–100 million Persistent discomfort 4
1–10 million Temporary pain 8
100 000–1 million Persistent pain 16
10 000–100 000 Minor first aid 32
1000–10 000 Major first aid 64
100–1000 Medical case 128
10–100 Lost work time 256
1–10 Permanent
disablement
1024
0.1–1 Fatality 26 214
Safety risk data 531
assembled using the 165 contacts provided
by the National Work Zone Safety Information Clear-
inghouse (http://www.workzonesafety.org/expert_con-
tacts/browse/all_experts). Because this website provides
no information regarding the qualification of any of the
contacts as ‘experts’, the research team independently
validated expert status with an introductory survey
using guidance provided by Hallowell and Gambatese
(2010). Of the 165 individuals contacted, 75 (45%)
responded, and 27 (36%) were qualified as experts.
According to Moser and Kalton (1971), this response
rate is acceptable for Delphi studies.
The resulting pool of individuals averaged over 25
years of highway construction safety experience. Over
80% of respondents had a Professional Engineering
(PE) licence, were a Certified Safety Professional
(CSP), or had at least a bachelor’s degree in a related
field and all respondents were upper-level managers
or executives (e.g. corporate safety manager, director
of research, and senior project manager). It should be
noted that, despite the relative large publication lists
of some participants, the panel was largely profes-
sional in nature. This was preferred as accurately
quantifying relative risks relies upon a wealth of
professional experience.
Number of iterations and feedback
One of the objectives of the Delphi process is to reach
consensus, which can be achieved by conducting mul-
tiple iterations of questionnaires and providing anony-
mous feedback between rounds. Two to seven rounds
have been used in the previous large-scale Delphi stud-
ies (Dalkey et al., 1970). According to Jolson and
Rossow (1971), iterations can be terminated when the
changes in variance are no longer significant. The
research team administered two rounds of surveys
because the size of the expert panel (27) exceeded the
minimum size recommended (eight) for traditional
Delphi studies (Brockhoff, 1975; Boje and Murnighan,
1982) and there was a high degree of consensus
among the experts after the second round.
Cognitive biases
In order to decrease the complexity of probability
assessment, many individuals use a limited number of
heuristic controls (Tversky and Kahneman, 1974).
However, relying on these heuristics may produce sys-
tematic errors in judgment known as cognitive biases.
Despite their importance, cognitive biases have not
received adequate attention in previous Delphi studies
(Hallowell and Gambatese, 2010). The following
eight biases were identified and controlled: collec-
tive unconscious, contrast effect, neglect of probabil-
ity, Von Restorff effect, myside bias, recency effect,
primacy effect and dominance (Hallowell and
Gambatese, 2010).
To minimize the potential influence of cognitive
biases, several controls were implemented. First,
respondents were kept anonymous. Maintaining the
anonymity of respondents reduces the impact of
group dynamics, dominant personalities and the
bandwagon effect (Manoliadis et al., 2006). Second,
randomizing the question order of the surveys mini-
mizes the potential influence of primacy and contrast
effects (Hallowell and Gambatese, 2010). Third, the
median ratings from the previous rounds were
provided as feedback, which significantly reduces vari-
ability among panellists (Martino, 1970). Fourth,
experts were asked to rate frequency and severity lev-
els separately to avoid the neglect of probability bias.
Finally, to ensure internal validity and to enhance the
reliability of the results, all experts were provided with
consistent task names and descriptions.
Phase II method: decision support system
development
One of the structured design methods to develop a
decision support system (DSS) is prototyping (Andr-
iole, 1989). The prototyping principles established
by Boar (1984) were followed where the develop-
ment of a DSS involves input from perspective users
and is refined with professional feedback in an itera-
tive process. Following the guidance provided by
Andriole (1989), the first step in designing the DSS
involved identifying the tasks that the system must
perform and the requirements of the user. According
to Boar (1984), 20 to 40% of DSS’s problems can
be attributed to the design process. Well-defined
requirements will make a link between users, tasks
and organizational needs (Andriole, 1989). Here, a
quick prototype was made and its features were
modified by receiving feedback from the users in an
iterative process. In the second step of the DSS
development, the safety risk data were mathemati-
cally integrated with activity sequences. The data
from Hallowell et al. (2011) and those established
through the Delphi process were used to populate
the theoretical model shown in Equation 1. In the
subsequent phase, the research team tested this
model with active construction projects in the US.
½SFTask�1�n ¼ ½RIndividual�1�25 � ð½RInteraction�25�25
� ½XSchedule�25�nÞ ð1Þ
where:
[RIndividual] is a matrix that includes safety risk values
for individual tasks; [RInteraction] is a matrix that includes
the safety risk interactions among tasks from Hallowell
532 Esmaeili and Hallowell
http://www.workzonesafety.org/expert_contacts/browse/all_experts
http://www.workzonesafety.org/expert_contacts/browse/all_experts
et al. (2011); [XSchedule] is a matrix that includes 0’s and
1’s depending on whether or not particular activities
are scheduled for a given time period. If in time t,
activity i is being performed, then Xit=1, otherwise Xit
= 0. [SFTask] is the resulting safety risk matrix that
includes the resulting risk for each time period.
Phase III method: risk data and DSS validation
One of the methods that has been used extensively to
decompose the general measure of effectiveness of a
DSS is Multi-Attribute Utility Assessment (MAUA)
(Adelman and Donnell, 1986; Sage, 1991). MUAU is
a formal structure that maps different measures of
effectiveness against one another and is defined as,
‘scoring and weighting procedures to evaluate the
overall utility of a knowledge-based system to users
and sponsors’ (Adelman and Riedel, 1997, p. 37).
This method has been used to evaluate similar DSSs
in several studies in the past (e.g. Adelman and
Ulvila, 1991). The total measure of effectiveness is
the weighted sum of all the utility scores, shown as
Equation 2 (Adelman, 1992):
UðiÞ ¼ W1: uðxi1Þ þ W2: uðxi2Þ þ � � �
þ WJ: uðxijÞ ð2Þ
where: U(i) is the overall utility for alternative i; wj is
the cumulative relative weight on attribute j; u(xij) is
the utility scale value for alternative i on attribute j.
Figure 1 presents the hierarchy of effectiveness
criteria that was created from existing literature and
discussions with potential users. The three main
evaluation criteria were: usability, applicability and
reliability. Usability was defined as the system’s ease
of use, response time, ease of training and graphic
displays; applicability was defined as the extent that
the program and its output can be used by a
construction firm to enhance decision making and
resource allocation; and reliability was defined as the
predictive accuracy of the system. It is notable that
predictive accuracy of the framework and developed
DSS relies heavily on the reliability of the safety risk
database. In other words, the reliability scores
obtained from the MAUA process are an indicator of
the validity of quantified safety risks.
In order to determine the total utility of the system,
the research team used a case study approach where
relative weights of the criteria and scores for the sys-
tem were obtained through interviews with prospec-
tive users. Case studies were chosen because the
sample size and randomization requirements of a true
experiment were not feasible (Adelman, 1992) and
case studies are appropriate for studying new strate-
gies in context (Yin, 2003). The main units of analy-
sis were active or recently completed projects.
To obtain a representative sample of US highway
construction projects, highway construction firms that
were members of the Associated General Contractors
(AGC) or the Colorado Asphalt Pavement Associa-
tion were asked to participate. Of the 39 contractors
that were contacted, a total of five firms agreed to
provide project data and participate in a series of
interviews. The revenue of the companies ranged
from $50 million to $2.5 billion with the average of
$583 million. The companies, on average, had more
than 700 workers and had been in the highway con-
struction business for over 50 years.
Overall utility
Applicability
Extent of use
Reliability
Usefulness of output
Impact on the current
procedures
Performance
Usability
General ease of use
Ease of training
Ease of data entry
Workload
Graphical features
Figure 1 Hierarchy of measures of effectiveness
Safety risk data 533
According to Yin (2003), the number of cases com-
pleted and the quality of pattern matching have signifi-
cant impact on the validity and reliability of the
results. Literature suggests that four to 10 cases will
provide valid and reliable data as long as pattern
matching is strong and data are collected consistently
among cases (Eisenhardt, 1989; Yin, 2003). To ensure
adequate data, a total of 11 case studies were con-
ducted. The demographics of these cases are summa-
rized in Table 2. As shown in Table 2, a diverse set of
projects is included ranging from large scope and long
duration to small scope and short duration. Also, a
higher number of projects were located in Colorado,
which limits the external validity of the results.
In order to increase the reliability and internal
validity of the study, a specific case study protocol
was implemented. The following four steps were
conducted for every case study:
(1) Interviews were conducted with the construc-
tion project manager or safety managers to
quantify the relative weights of the attributes by
conducting pairwise comparisons between crite-
ria. The interviewees were asked to use a pro-
vided comparison scale that was based on
previously successful studies described by Saaty
(1980). A consistency ratio was then used to
ensure that each respondent’s ratings were inter-
nally consistent. As suggested by Shapira and
Goldenberg (2005), participants were asked to
repeat the rating process if their internal consis-
tency ratio exceeded 0.1. In other words, if an
individual’s pairwise comparisons among crite-
ria resulted in 10% or greater internal inconsis-
tency, they were asked to repeat the process
until their ratings were in agreement. An
acceptable internal consistency ratio indicates
that there is no intolerable conflict in the com-
parisons of a participant’s response (Shapira and
Goldenberg, 2005).
(2) After the weights had been found, interviews
were conducted to determine the DSS’s scores
for different criteria. In order to gather opin-
ions about the usability and applicability of the
DSS, the operation of the system was demon-
strated to the participants. Immediately follow-
ing the demonstration, the users were asked to
complete an 18-question survey (two questions
for each attribute) that addressed all criteria
shown in Figure 1. The participants were asked
to rate the system’s performance on a scale
from 0% (very poor performance) to 100%
(very strongly performance), with 50% being
neutral.
(3) The project schedule was then obtained and
the project manager was interviewed to ensure
that the research team had an accurate under-
standing of the actual activities that were per-
formed on the project. With the project
manager, the tasks and durations were matched
with the tasks described in Hallowell et al.
(2011). This mapping process was required
because the DSS was built around the data
from previous research and consistency of task
names was required for the system to operate
effectively. Once the tasks were mapped, the
schedule integration function of the DSS was
used to produce a safety risk profile. The con-
struction project manager or safety manager
was then asked to compare the risk profile
Table 2 Projects demographic information
Project
number
Scope
(million $)
Duration
(month)
Percent of
completion (%)
Delivery
method
Method of
payment
Recordable
injuries Location
1 150 50 88 Bid/Build Unit price 2 California
2 110 30 65 Design/
Build
Lump sum 0 Utah
3 48 36 94 Bid/Build Monthly
progress
2 California
4 5.5 4 100 – Pay
estimates
0 Colorado
5 4.5 6 100 Bid/Build Unit price 0 Washington
6 0.32 1.5 99 Bid/Build Unit price 0 Colorado
7 0.38 3 90 Bid/Build – 0 Colorado
8 0.66 5 100 – – 0 Colorado
9 0.37 10 100 Bid/Build Unit price 0 Colorado
10 0.49 1.5 100 Bid/Build – 0 Colorado
11 1.5 3 100 – Pay
estimates
0 Colorado
534 Esmaeili and Hallowell
created by the DSS with the actual level of risk
and provide an approximate percentage agree-
ment with the system. The interviewees aimed
to compare the pattern of the risk profile with
near misses and the actual hazards that existed
during the work. In fact, the current study did
not aim to predict injuries in the jobsite.
Rather, the focus was on predicting high risk
work periods where the potential for injury is
relatively high. It is important to distinguish
the difference between hazards and accidents.
For example, if a worker was exposed to adja-
cent traffic, there were significant hazards even
though no injury was realized.
(4) The final step of the case study involved a fol-
low-up questionnaire that included open-ended
questions that gave the participants an oppor-
tunity to share their thoughts on the perceived
strengths and weaknesses of the system.
Results
Phase I results: risk quantification
All 27 expert panellists provided complete responses
to the Delphi surveys in the first round and the abso-
lute variance of responses for frequency and severity
were 0.733 and 0.838, respectively. Once the data
were aggregated and summarized for the panel, sec-
ond rounds of surveys were administered. In the sec-
ond round, the median responses from the previous
round were provided to the panellists and they were
given the option to agree with the other group’s col-
lective assessment or provide a new rating. Of the 27
surveys that were sent in the second round, 24 sur-
veys were returned resulting in 89% response rate.
Absolute variances of responses in the second round
were 0.198 and 0.191 for frequency and severity,
respectively. Because the established consensus was
achieved in the second round, there was no need for
a third round. Additionally, the median ratings did
not change between rounds, which is evidence of
strong internal validity.
To facilitate calculations, the frequency ratings
were converted from a range of values with units of
worker-hours per incident to a single point value with
units of incidents per worker-hour. The mean value
was selected as a point value and inverted to obtain a
number with appropriate units. For example, if the
Delphi panel rated the average frequency as 10–100
w-h/incident, the mean value, 55 w-h/ incident, was
inverted (0.018 incidents/w-h) to determine the fre-
quency value for that particular risk and activity.
Severity values were not changed from the severity
scale in Table 1.
The frequency ratings ranged from 1.8E-8 to 0.018
incidents per worker-hour and the severity ratings
ranged from 4 to 256 units on the severity scale. Unit
risk scores were calculated by multiplying the average
frequency scores by the average severity scores. The
resulting data for the 25 work tasks are provided in
descending order of relative risk in Table 3. In this
table, risk is described in terms of units of severity per
worker-hour (S/w-h). The task ‘construction zone
traffic control’ has the highest unit risk (0.047 S/w-h)
while ‘watering and dust palliatives’ (1.8 � 10-8
S/w-h) and ‘install field facilities’ (1.8 � 10-8 S/w-h)
have the lowest unit risk.
Phase II results: DSS development
In order to provide a user friendly environment to
integrate safety risk data and project schedules, a
graphical user interface (GUI) was developed in
MATLAB called Scheduled-based Safety Risk Assess-
ment and Management (SSRAM). MATLAB was
chosen for two main reasons: it is a strong program-
ming language to develop graphical user interfaces
and it allows the research team to make an active con-
nection between standard project management soft-
ware and safety risk databases. An applied DSS for
integrating safety risk data into project schedules
should include two main capabilities: (1) receiving the
schedule from the user; and (2) creating the safety
risk profile. These capabilities were considered during
the development. Although the DSS can be used to
manually add tasks, start dates and end dates to build
the schedule, the researchers built a bridge between
Primavera 6, MS Project and the DSS using MS
Excel as medium to increase efficiency. After entering
projects to the program, the user can save the sched-
ules as M-file (
⁄
.m or
⁄
.matt). The resulting DSS
(SSRAM) is a knowledge-based system with a
schedule integration engine.
The conceptual formulation and computational
process of the SSRAM is shown in Figure 2. The
safety risk database includes the base-level safety risk
(obtained from Delphi panel in the first phase) and
the safety risk interactions (a 25 � 25 matrix from
Hallowell et al., 2011). The user can insert the
schedule manually or import it from scheduling
software (e.g. Primavera 6). Once the schedule is
entered, the SSRAM loads safety risk data from the
database to the imported schedule using Equation 1
and subsequently plots the risk profile. There are sev-
eral practical applications of the safety risk profiles.
For example, the risk profiles can be used to identify
high risk periods during the project, the safety risk
can be levelled utilizing float of activities, or the pro-
ject manager can allocate safety resources according
Safety risk data 535
to the risk profile. In addition, the program is able to
create safety risk profiles for multiple projects or port-
folios of a company. This is important because safety
managers for highway construction companies must
often manage multiple concurrent projects. Using the
SSRAM helps them to strategically allocate their time
and safety resources.
Phase III results: risk data and DSS validation
The results of pairwise comparisons made by the
user group are shown in Table 4. The users
believed that the reliability, the general ease of use
of the program, and the usefulness of the output
are the most important attributes. Because the
usability and applicability have a subset of
Table 3 Safety risk data for common highway reconstruction work tasks
Task name Frequency score (incident/w-h
⁄
E-5) Severity score Unit risk scores (S/w-h
⁄
E-5)
Construction zone traffic control 1800 256 4700
Install traffic control devices 18 64 4700
Installing flexible pavement/patching 18 64 12
Pavement marking 18 64 12
Seal joints and cracks 18 64 12
Excavation 18 64 12
Install culverts, drains, sewers 18 64 12
Install culvert pipe and water lines 18 64 12
Reset structures 18 64 12
Heat and scarifying 18 64 12
Survey 1.8 32 5.8
Clear and grub 0.18 16 2.9
Recycle cold bituminous pavement 0.18 16 2.9
Install curb and gutters 0.18 16 2.9
Install rigid pavement (concrete) 0.18 16 2.9
Install cribbing 0.18 16 2.9
Recondition bases (compaction) 0.18 16 2.9
Install water control devices 0.18 16 2.9
Lay aggregate base course 0.18 16 2.9
Mobilization/demobilization 0.18 16 2.9
Prime, coat, rejuvenate pavement 0.18 16 2.9
Demolition of existing pavement 0.18 16 2.9
Landscape 0.18 16 0.029
Install field facilities 0.0018 4 0.0073
Watering and dust palliatives 0.0018 4 0.0073
Integration procedure
Report
[SF
Task
]
1×n =[RIndividual]1×25 ×([RInteraction]25×25 ×[XSchedule]25×n)
1. Safety risk profiles
Project’s schedule
1. Inserting manually
2. Importing from
scheduling software
2. Risk threshold
Safety risk database
1. Base level safety risks
(Table 2)
2. Safety risk interactions
(Hallowell et al., 2011)
Figure 2 SSRAM’S framework
536 Esmaeili and Hallowell
attributes, the relative weights for the higher tier
attributes were computed by finding the products of
the subsets (see Table 4). For example, the relative
weight of workload (0.17) was multiplied by the rel-
ative weight assigned to usability (0.15) to reach
the total ‘workload’ weight of (0.03). Once the
weights of the criteria were found, they were multi-
plied against the scores and the resulting products
were summed to compute the global utility factor
(0.67). This number can be interpreted as the value
that the SSRAM adds to the current safety manage-
ment practice, which ranges from a score of 1 that
corresponds to a revolutionary product that
completely changes current industry practice and is
perfectly executed to a score of 0 where no value is
added. One should note that the interviewees, who
rated the output of the program, were not informed
of the analytical procedure that resulted in the out-
put. Therefore, any inconsistencies among the inter-
viewees’ opinions and the Delphi experts’ judgment
decrease the actual utility of the system as a whole.
In fact, the proposed validation methodology tested
the ability of the SSRAM to forecast hazardous
conditions. Considering the complex and dynamic
nature of the construction projects, reaching 100%
accuracy was not realistic.
Although the research team was satisfied with the
results, there are no similar DSS validation studies
to compare against. Fortunately, the follow-up inter-
view questions validated the SSRAM global utility
score because respondents indicated that the pro-
gram is easy to use and greatly improves precon-
struction safety management at the project and
program levels.
Limitations
Though the results have the potential to impact posi-
tively on preconstruction safety management, there
are several limitations of the research.
(1) The risk quantification portion of the study
required that the Delphi panel assume typical
conditions in their ratings. Consequently, the
data are limited by the fact that actual conditions
such as weather, crew safety culture and fatigue
affect the true risk values (Manu et al., 2010).
Because there are large numbers of external risk
factors that can affect the base-level risk of activ-
ities, considering their effects was unrealistic.
The influence of external factors may explain
some of the variation between the predicted
values and actual values during the case studies.
(2) The risk values were estimated as a single point
estimate for each task. The average risk values
may not capture all characteristics of risk or, as
Kaplan and Garrick (1981) stated, a single
number cannot communicate risk effectively
due to the great loss of information.
(3) Although several measures were employed to
decrease cognitive biases in the Delphi process,
there are still several limitations to the frequency
and severity values provided by experts. One of
the common limitations is related to accidents
with low probability of occurrence and high
impacts. Taleb (2007), one of the prominent
researchers in this area, called these extreme
events, ‘Black Swans’. He stated that it is almost
impossible to predict extreme events because
they do not have predecessor events (Taleb,
Table 4 Weights of measures of effectiveness and their consistency ratio
Measures of effectiveness Relative weights Total weights Scores Utility
Usability 0.15 – – –
Applicability 0.25 – – –
Reliability 0.60 0.60 0.66 0.40
Usability General ease of use 0.31 0.05 0.69 0.03
Ease of training 0.26 0.04 0.75 0.03
Ease of data entry 0.17 0.03 0.66 0.02
Workload 0.17 0.03 0.60 0.02
Graphical features 0.07 0.01 0.68 0.01
Applicability Extent of use 0.23 0.06 0.58 0.03
Usefulness of output 0.36 0.09 0.73 0.06
Impact on the current procedures 0.29 0.07 0.61 0.04
Performance 0.13 0.03 0.70 0.02
Total 0.67
Safety risk data 537
2007). In fact, predicting these low-probability,
high-impact events is extremely difficult and
more attention should be paid to reduce the vul-
nerability of the system towards their conse-
quences than anticipating them (Taleb, 2004).
(4) The safety risks were quantified for only 25
tasks. In order to add a new task to the schedule,
its base-level safety risk and the interactions with
other tasks must be quantified separately.
(5) The external validity is limited because the
data and DSS were validated on projects in
Colorado, Oregon, California, Utah and Wash-
ington, with a higher number in Colorado.
Although it is expected that projects are repre-
sentative of the US, the scope of inference is
theoretically limited only to these states.
Despite these limitations, the resulting data and
SSRAM significantly furthered knowledge and were
accurate and useful enough to gain favourable
responses from industry users.
Conclusions and recommendations
According to Esmaeili and Hallowell (2012a), the
construction industry is saturated with respect to
traditional injury prevention strategies and new safety
innovations are needed. Previous research has estab-
lished that the potential to prevent construction inju-
ries is at its highest during the preconstruction phase
and decreases exponentially as a project progresses
(Gambatese et al., 1997; Szymberski, 1997). Tradi-
tionally, preconstruction safety improvement tech-
niques such as designing for safety have faced
significant barriers that stem from the fact that they
are largely designer-controlled (Hinze and Wiegand,
1992). One of the contractor-controlled practices that
can be used to overcome this shortcoming is safety-
schedule integration.
Although there have been attempts to integrate
safety risk data into project schedules (e.g. Yi and
Langford, 2006; Sacks et al., 2009), these attempts
were not successful because of the absence of a valid
and reliable safety risk database. This limitation was
addressed by quantifying the relative risk of 25 com-
mon highway construction tasks and combining this
information with risk interaction data obtained by
Hallowell et al. (2011) to populate a new risk integra-
tion framework. To facilitate implementation, validate
the risk data, and test the utility of the underlying
framework, the research team created a safety DSS
(SSRAM) using MATLAB. The resulting tool inter-
faced with Primavera P6 to produce plots of safety risk
over time for single or multiple concurrent projects.
The tool was then tested on 11 case study projects.
This validation effort revealed that the data are valid
and reliable despite recognized limitations and the
SSRAM has the potential to improve safety resource
optimization and preconstruction safety management.
To summarize, the authors tested the validity of the
risk database as an input and reliability of the risk pro-
files generated by integrating base-level risk data and
risk interaction data with highway project schedules.
The traditional safety management approach
involves investing safety resources such as time and
money at a uniform rate throughout the lifespan of a
project (Rozenfeld et al., 2009). However, because
physical conditions change rapidly, safety risk levels
may also fluctuate. Consequently, uniform resource
allocation to safety within a project and among
projects may not be the optimum strategy. One of the
viable solutions for this problem is to apply lean
thinking to the construction process (Womack and
Jones, 2003) so that injury prevention practices can
be treated as production control activities (Rozenfeld
et al., 2010). According to Rozenfeld et al. (2010),
the ability to predict fluctuating safety risk levels is
essential to practical lean-based safety management.
The findings can be used to predict safety risk levels
and use schedule float to distribute or concentrate
risk. In addition, risk profiles enable thoughtful safety
planning and effective allocation of safety resources in
a single project or multiple projects.
In addition, the data and framework presented can
be used by project managers to enhance preconstruc-
tion safety management by identifying high risk peri-
ods. In response, safety managers can plan for extra
precautionary measures during these high risk periods
(e.g. lane closure), develop customized injury preven-
tion strategies, or at a minimum, inform workers of
the tasks and interactions known to cause high risk
periods. In addition to using the schedule-based tech-
nique described, the writers also recommend that
practitioners focus attention on high risk work tasks
(e.g. construction zone traffic control, installing traffic
control devices and excavation).
To address the aforementioned study limitations
the writers suggest three complementary research
efforts. First, new safety risk quantification methods
should be explored to produce robust and reliable risk
data independently from specific tasks, trades and
construction objects. For example, Esmaeili and
Hallowell (2011, 2012b) utilized genome concept to
quantify safety risks at attribute level independently
from tasks and objects. Population of this method can
be a major step towards a universal safety risk
assessment in the construction industry. Second,
researchers should consider modelling probability dis-
tributions of accident occurrence for individual tasks.
538 Esmaeili and Hallowell
Using probability distributions instead of the average
estimated points allows an individual to consider
uncertainty in the data and investigate its propagation
thorough the model (Fischhoff et al., 1984). Finally,
the relative impacts of environmental risk factors on
the base safety risk values must be better understood
to create robust models. The external risk factors may
include the intensification of risk due to night-time
work, exposure to weather, and adjacent traffic.
Acknowledgements
The writers would like to thank Bentley Systems for
the resources and high quality feedback during the
project and all of the Delphi panellists for their
enthusiastic participation.
References
Adelman, L. (1992) Evaluating Decision Support and Export
Systems, John Wiley & Sons, New York.
Adelman, L. and Donnell, M.L. (1986) Evaluating decision
support systems: a general framework and case study, in
Andriole, S.J. (ed.) Microcomputer Decision Support Systems,
QED Information Systems, Wellesley, MA, pp. 285–309.
Adelman, L. and Riedel, S. (1997) Handbook for Evaluating
Knowledge-Based Systems, Kluwer Academic, Boston.
Adelman, L. and Ulvila, J.W. (1991) Evaluating expert sys-
tem technology, in Andriole, S.J. and Halpin, S.M. (eds)
Information Technology for Command and Control, IEEE
Press, New York, pp. 537–47.
Andriole, S.J. (1989) Handbook of Decision Support Systems,
TAB Books Inc, Blue Ridge Summit, PA.
Arditi, D., Ayrancioglu, M. and Shi, J. (2005) Worker
safety issues in night-time highway construction. Engineer-
ing Construction and Architectural Management, 12(5),
487–501.
Boar, B. (1984) Application Prototyping: A Requirements Defi-
nition Strategy for the 80s, Wiley, New York.
Boje, D.M. and Murnighan, J.K. (1982) Group confidence
pressures in iterative decisions. Management Science, 28
(10), 1187–96.
Brauer, R.L. (1994) Risk management and assessment, in
Brauer, R.L. Safety and Health for Engineers, Van Nos-
trand Reinhold, New York, pp. 645–64.
Brockhoff, K. (1975) The performance of forecasting
groups in computer dialogue and face-to-face discussion,
in Linstone, H.A. and Turloff, M. (eds) The Delphi
Method: Techniques and Applications, Addison-Wesley,
Reading, MA, pp. 291–321.
Bryden, J.E. and Andrew, L. (1999) Serious and fatal inju-
ries to workers on highway construction projects. Trans-
portation Research Record, 1657, 42–7.
Bureau of Labor Statistics (2012) Occupational Injuries/Ill-
nesses and Fatal Injuries Profiles, available at http://goo.gl/
CNJb3 (accessed January 2012).
Cabaniss, K. (2001) Counseling and computer technology in
the new millennium: an Internet Delphi study, PhD thesis,
Virginia Polytechnic Institute and State University, Virginia.
Cagno, E., Giulio, A.D. and Trucco, A. (2001) Algorithm
for the implementation of safety improvement programs.
Safety Science, 37, 59–75.
Carter, G.M., Murray, M.P., Walker, R.G. and Walker, W.
E. (1992) Building Organizational Decision Support Sys-
tems, Academic Press, Book News, Inc., Portland, OR.
Center for Construction Research and Training (2008) Con-
struction Chartbook, Center for Construction Research and
Training, Silver Spring, MD.
Coble, R.J. and Elliott, R.B. (2000) Scheduling for
construction safety, in Coble, R.J., Hinze, J. and Haupt,
T.C. (eds) Construction Safety and Health Management,
Prentice Hall, Englewood Cliffs, NJ, pp. 43–57.
Dalkey, N., Brown, B. and Cochran, S. (1970) Use of self-
ratings to improve group estimates. Technological Forecasting,
1, 283–91.
Dijksterhuis, A., Bos, M.W., Nordgren, L.F. and von Baa-
ren, R.B. (2006) On making the right choice. the deliber-
ation-without-attention effect. Science, 311, 1005–7.
Eisenhardt, K.M. (1989) Building theories from case study
research. Academy of Management Review, 14(4), 532–50.
Esmaeili, B. and Hallowell, M.R. (2011) Using network
analysis to model fall hazards on construction projects, in
Tjandra, K. (ed.) Proceedings of Safety and Health in Con-
struction, CIB W099, Washington, DC, 24–26 August.
Esmaeili, B. and Hallowell, M. (2012a) Diffusion of safety
innovations in the construction industry. Journal of Con-
struction Engineering and Management, 138(8), 955–63.
Esmaeili, B. and Hallowell, M.R. (2012b) Attribute safety
risk model for measuring safety risk of struck-by acci-
dents, in Cai, H., Kandil, A., Hastak, M. and Dunston,
P.S. (eds) Proceedings of The 2012 Construction Research
Congress (CRC), West Lafayette, 21–23 May.
Federal Highway Administration (2004) Work zone safety
facts and statistics, updated on 24 December 2004, avail-
able at http://goo.gl/X7crE (accessed February 2012).
Fischhoff, B., Watson, S. and Hope, C. (1984) Defining
risk. Policy Sciences, 17, 123–39.
Gambatese, J.A., Hinze, J.W. and Haas, C.T. (1997) Tool
to design for construction worker safety. Journal of Archi-
tectural Engineering, 3(1), 32–41.
Gyi, D.E., Gibb, A.G.F. and Haslam, R.A. (1999) The
quality of accident and health data in the construction
industry: interviews with senior managers. Construction
Management and Economics, 17, 197–204.
Hadikusumo, B.H.W. and Rowlinson, S. (2004) Capturing
safety knowledge using Design-for-Safety-Process Tool.
ASCE Journal of Construction Engineering and Management,
130(2), 281–9.
Hallowell, M.R. (2008) A formal model of construction
safety and health risk management, unpublished PhD dis-
sertation, Oregon State University, Corvallis, OR.
Hallowell, M.R. and Gambatese, J.A. (2009) Activity-based
safety risk quantification for concrete formwork construction.
ASCE Journal of Construction Engineering and Management,
135(10), 990–8.
Safety risk data 539
http://goo.gl/CNJb3
http://goo.gl/CNJb3
http://goo.gl/X7crE
Hallowell, M.R. and Gambatese, J.A. (2010) Qualitative
research: application of the Delphi method to CEM
research. ASCE Journal of Construction Engineering and
Management, 136(1), 99–107.
Hallowell, M.R., Esmaeili, B. and Chinowsky, P. (2011)
Safety risk interactions among highway construction work
tasks. Construction Management and Economics, 29(4),
417–29.
Han, S.H. and Diekmann, J.E. (2001) Approaches for mak-
ing risk-based go/no-go decision for international projects.
ASCE Journal of Construction Engineering and Management,
127(4), 300–8.
Helmer, O. (1967) Analysis of the Future: The Delphi Method,
The Rand Corporation, Santa Monica, CA.
Hill, G.W. (1982) Group versus individual performance.
Psychological Bulletin, 91, 517–39.
Hinze, J., Nelson, J. and Evans, R. (2005) Software integra-
tion of safety in construction schedules, in Haupt, T.C.
and Smallwood, J. (eds) Proceedings of the 4th Triennial
International Conference, Rethinking and Revitalizing Con-
struction Safety, Health, Environment and Quality, Port
Elizabeth, South, Africa, 16–17 May, pp. 212–22.
Hinze, J. and Wiegand, F. (1992) Role of designers in con-
struction worker safety. ASCE Journal of Construction
Engineering and Management, 118(4), 677–84.
Hogarth, R.M. (1978) A note on aggregating opinions.
Organizational Behavior and Human Performance, 21,
40–6.
Jolson, M.A. and Rossow, G. (1971) The Delphi process in
marketing decision making. Journal of Marketing Research,
8, 443–8.
Kak, A., Zacharias, K., Minkarah, I. and Pant, P. (1995) A
knowledge based software for construction safety, in
Mohsen, J.P. (ed.) Proceedings of the Second Congress on
Computing in Civil Engineering, Atlanta, GA, 5–8 June,
pp. 997–1002.
Kaplan, S. and Garrick, B.J. (1981) On the quantitative
definition of risk. Risk Analysis, 1(1), 11–28.
Kartam, N. (1997) Integrating safety and health perfor-
mance into construction CPM. ASCE Journal of Construc-
tion Engineering and Management, 123(2), 121–6.
Karumanasseri, G. and AbouRizk, S. (2002) Decision sup-
port system for scheduling steel fabrication projects.
ASCE Journal of Construction Engineering and Management,
128(5), 392–9.
Leu, S.S., Yang, C.H. and Huang, J.C. (2000) Resource
leveling in construction by genetic algorithm-based opti-
mization and its decision support system application.
Automation in Construction, 10(1), 27–41.
Lin, K. and Haas, C. (1996) Multiple heavy lifts optimiza-
tion. ASCE Journal of Construction Engineering and Man-
agement, 122(4), 354–62.
Linstone, H. and Turoff, M. (1975) The Delphi Method:
Techniques and Applications, Addison-Wesley, Reading,
MA.
Manoliadis, O., Tsolas, O. and Nakou, A. (2006) Sustain-
able construction and drivers of change in Greece. a Del-
phi study. Construction Management and Economics, 24(2),
113–20.
Manu, P., Ankrah, N., Proverbs, D. and Suresh, S. (2010)
An approach for determining the extent of contribution
of construction project features to accident causation.
Safety Science, 48(6), 687–92.
Martino, J. (1970) The precision of Delphi estimates. Tech-
nological Forecasting, 1(3), 293–9.
Mitropoulos, P., Abdelhamid, T. and Howell, G. (2005)
Systems model of construction accident causation. ASCE
Journal of Construction Engineering and Management, 131
(7), 816–25.
Molenaar, K.R. and Songer, A.D. (2001) Web-based deci-
sion support systems: case study in project delivery. ASCE
Journal of Computing in Civil Engineering, 15(4), 259–67.
Moser, C.A. and Kalton, G. (1971) Survey Methods in
Social Investigation, 2nd edn, Heinemann Educational,
London.
Murphy, M.K., Black, N.A., Lamping, D.L., McKee, C.
M., Sanderson, C.F.B., Askham, J. and Marteau, T.
(1998) Consensus development methods and their use in
clinical guideline development. Health Technology Assess-
ment, 2(3), 1–88.
National Institute of Occupational Safety and Health, Fatal-
ity Assessment and Control Evaluation (FACE) program
(n.d.) NIOSH FACE reports indexed by industry or
cause of fatality, available at http://goo.gl/CYHuw
(accessed February 2012).
Navon, R. and Kolton, O. (2006) Model for automated
monitoring of fall hazards in building construction. Jour-
nal of Construction Engineering and Management, 132(7),
733–40.
Navon, R. and Kolton, O. (2007) Algorithms for automated
monitoring and control of fall hazards. ASCE Journal of
Computing in Civil Engineering, 21(1), 21–8.
Occupational Safety and Health Administration Integrated
Management Information System (OSHA IMIS) (n.d.)
Statistics and Data, Standard Industry Classification
(SIC) system search, available at http://goo.gl/iC2dV
(accessed February 2012).
Palaneeswaran, E. and Kumaraswamy, M.M. (2008) An
integrated decision support system for dealing with time
extension claims in construction projects. Automation in
Construction, 17(4), 425–38.
Pandey, S. (2009) Risk quantification for highway recon-
struction activities, unpublished Masters thesis, University
of Colorado, Boulder, CO.
Perera, S. (1983) Resource sharing in linear construction.
ASCE Journal of Construction Engineering and Management,
109(1), 102–11.
Rajendran, S. and Gambatese, J.A. (2009) Development
and initial validation of sustainable construction safety
and health rating system. ASCE Journal of Construction
Engineering and Management, 135(10), 1067–75.
Rowe, G. and Wright, G. (1999) The Delphi technique as
a forecasting tool: issues and analysis. International Journal
of Forecasting, 15, 353–75.
Rozenfeld, O., Sacks, R. and Rosenfeld, Y. (2009)
‘CHASTE’: construction hazard assessment with spatial
and temporal exposure. Construction Management and Eco-
nomics, 27(7), 625–38.
540 Esmaeili and Hallowell
http://goo.gl/CYHuw
http://goo.gl/iC2dV
Rozenfeld, O., Sacks, R., Rosenfeld, Y. and Baum, H.
(2010) Construction job safety analysis. Safety Science,
48, 491–8.
Russell, J.S., Skibniewski, M.J. and Cozier, D.R. (1990)
QUALIFIER-2: knowledge based system for contractor
prequalification. ASCE Journal of Construction Engineering
and Management, 116(1), 157–71.
Saaty, T.L. (1980) The Analytic Hierarchy Process, McGraw-
Hill, London.
Sacks, R., Rozenfeld, O. and Rosenfeld, Y. (2009) Spatial
and temporal exposure to safety hazards in construction.
ASCE Journal of Construction Engineering and Management,
135(8), 726–36.
Sage, A.P. (1991) Decision Support Systems Engineering,
Wiley, New York.
Saurin, T.A., Formoso, C.T. and Guimaraes, L.B.M.
(2004) Safety and production: an integrated planning and
control model. Construction Management and Economics,
22, 159–69.
Sawacha, E., Naoum, S. and Fong, D. (1999) Factors
affecting safety performance on construction sites. Inter-
national Journal of Project Management, 17(5), 309–15.
Shapira, A. and Goldenberg, M. (2005) AHP-based equip-
ment selection model for construction projects. Journal of
Construction Engineering and Management, 131(12),
1263–73.
Snashall, D. (1990) Safety and health in the construction
industry. British Medical Journal, 301, 563–4.
Szymberski, R. (1997) Construction project safety planning.
TAPPI Journal, 80(11), 69–74.
Taleb, N.N. (2004) Fooled by Randomness: The Hidden Role
of Chance in Life and in the Markets, 2nd edn, Random
House, New York.
Taleb, N.N. (2007) The Black Swan: The Impact of the
Highly Improbable, Random House, New York.
Tarrants, W.E. (1980) The Measurement of Safety Perfor-
mance, Garland STPM Press, New York.
Tversky, A. and Kahneman, D. (1974) Judgment under
uncertainty: heuristics and biases. Science, New Series,
185(4157), 1124–31.
Wang, W.C., Liu, J.J. and Chou, S.C. (2006) Simulation-
based safety evaluation model integrated with network
schedule. Automation in Construction, 15, 341–54.
Womack, J.P. and Jones, D.T. (2003) Lean Thinking:
Banish Waste and Create Wealth in Your Corporation, Free
Press, New York.
Yi, K. and Langford, D. (2006) Scheduling-based risk esti-
mation and safety planning for construction projects.
ASCE Journal of Construction Engineering and Management,
132(6), 626–35.
Yin, R.K. (2003) Case Study Research: Design and Methods,
3rd edn, Sage, Thousand Oaks, CA.
Safety risk data 541
Copyright of Construction Management & Economics is the property of Routledge and its
content may not be copied or emailed to multiple sites or posted to a listserv without the
copyright holder’s express written permission. However, users may print, download, or email
articles for individual use.
Safety risks interactions among highway construction work tasks x
References
HALLOWELL, M., ESMAEILI, B., & CHINOWSKY, P. (2011). Safety risk interactions among highway construction work tasks. Construction Management & Economics, 29(4), 417-429. doi:10.1080/01446193.2011.552512
Abstract:
Recent research has produced frameworks for integrating safety risk data into project schedules, visual models and otherconstruction planning tools. Unfortunately, only a few studies have attempted to quantify base-level safety risk forconstruction tasks and no study has attempted to quantify the degree to which spatial and temporal interactions among tasks contribute to the potential for injury. A research study was performed to quantify the impact that pair-wise spatial and temporal interactions have on the base-level risk of 25 common highway construction work tasks in the United States. Six hundred risk interactions were quantified by obtaining and aggregating over 23 500 individual ratings from certified experts using the Delphi method. The results indicate that incompatible tasks may increase the base-level risk up to 60%. The most incompatible highway construction tasks are: (1) installing curbs and gutters and installing rigid pavement; and (2)construction zone traffic control and installing rigid pavement. Additionally, watering and dust palliatives and pavement marking is the one compatible task pair and there are 45 neutral task pairs. The resulting database and analysis have the potential to increase the efficacy of existing frameworks for integration of safety risk data with project planning tools. [ABSTRACT FROM AUTHOR]
Safety risk interactions among highway construction work tasks
Construction Management and Economics
(
April 2011)
29
, 417–429
Construction Management and Economics
ISSN 0144-6193 print/ISSN 1466-433X online © 2011 Taylor & Francis
http://www.informaworld.com
DOI: 10.1080/01446193.2011.552512
Safety risk interactions among highway construction
work tasks
MATTHEW HALLOWELL
*
,
BEHZAD ESMAEILI and PAUL CHINOWSKY
Department of Civil, Environmental and Architectural Engineering, University of Colorado, 428 UCB, 1111
Engineering Drive, Boulder, 80303 USA
Taylor and Francis
Received 10 August 2010; accepted 21 December 2010
10.1080/01446193.2011.552512
Recent research has produced frameworks for integrating safety risk data into project schedules, visual models
and other construction planning tools. Unfortunately, only a few studies have attempted to quantify base-level
safety risk for construction tasks and no study has attempted to quantify the degree to which spatial and
temporal interactions among tasks contribute to the potential for injury. A research study was performed to
quantify the impact that pair-wise spatial and temporal interactions have on the base-level risk of 25 common
highway construction work tasks in the United States. Six hundred risk interactions were quantified by
obtaining and aggregating over 23 500 individual ratings from certified experts using the Delphi method. The
results indicate that incompatible tasks may increase the base-level risk up to 60%. The most incompatible
highway construction tasks are: (1) installing curbs and gutters and installing rigid pavement; and (2)
construction zone traffic control and installing rigid pavement. Additionally, watering and dust palliatives and
pavement marking is the one compatible task pair and there are 45 neutral task pairs. The resulting database
and analysis have the potential to increase the efficacy of existing frameworks for integration of safety risk data
with project planning tools.
Keywords:
Project management, risk analysis, safety.
Introduction
Over the last 40 years the construction industry has
accounted for an injury and fatality rate that is nearly
five times greater than the all-industry average (Bureau
of Labor Statistics, 2010). Although injury rates have
declined dramatically in this time, in each of the past
15 years the construction industry has accounted for
over 1200 deaths and 460 000 disabling injuries in the
United States (National Safety Council, 2009). In
addition to physical pain and emotional suffering expe-
rienced by the victims and their families, these inci-
dents have substantial societal costs totalling an
estimated $15.64 billion annually (National Safety
Council, 2009). Furthermore, it has also been shown
that injuries alone account for 7.9% to 15% of the costs
of new construction (Everett and Frank, 1996). These
costs cripple entrant firms and have a strong, negative
impact on the gross domestic product (GDP).
Following the Occupational Safety and Health Act of
1970, numerous attempts have been made to improve
understanding of construction safety. For example,
Bernold and Guler (1993) identified common activities
and physical motions that contribute to back injuries;
Hinze
et al.
(1998) suggested a new classification
method for identifying root causes of injuries; Chi
et al.
(2005) identified key contributing factors to fall inci-
dents; Hinze
et al.
(2005a) studied the root causes of
struck-by accidents; Sobeih
et al.
(2009) identified
causes of musculoskeletal disorders; Lombardi
et al.
(2009) evaluated factors affecting workers’ perception
of risk; and Mitropoulos and Guillama (2010)
suggested a protocol to evaluate the potential for injury
when constructing residential framing. Though the
contributions of these previous studies are considerable,
they are limited in application because they evaluate
injuries, activities and preventive measures as individual
issues and isolated subjects (Sacks
et al.
, 2009).
*
Author for correspondence. E-mail: matthew.hallowell@colorado.edu
418
Hallowell
et al
.
Construction projects are characterized by
complexity and uncertainty which stems from an ever-
changing environment. The dynamic nature of
construction projects requires safety measures to be
adapted to new situations. Consequently, many
experts believe that injury prevention activities should
be conducted early in the project life cycle (Hinze,
1997). One emerging proactive safety management
strategy is to integrate safety information into project
schedules (Kartam, 1997; Chantawit
et al.
, 2005;
Hinze
et al.
, 2005a). Recently, Yi and Langford
(2006) and Sacks
et al.
(2009) developed techniques
for ‘safety loading’ safety risk data into critical path
method (CPM) schedules. According to Yi and Lang-
ford (2006), the quantity of safety risk varies during
the project schedule and limited resources should be
allocated to projects in proportion to their safety risk
at any given time. To analyse temporal safety risk,
both studies concluded that safety risk data should be
numerically integrated into the project schedule. Prior
to these efforts, resource allocation for safety manage-
ment was inefficient because resources (e.g. safety
personnel) were assigned to projects for longer peri-
ods than they were actually required for (Sacks
et al.
,
2009).
In order to effectively integrate safety risk data with
project schedules, managers must identify and quan-
tify safety risk for all scheduled tasks. Though the
framework for schedule integration established by Yi
and Langford (2006) only requires base-level risk data
for the performance of individual tasks, in isolation,
under typical circumstances, several authors have
postulated that the actual risk of construction opera-
tions also depends on the interactions that occur
among tasks throughout space and time (Lee and
Halpin, 2003; Sacks
et al.
, 2009; Rozenfeld
et al.
,
2010). These studies argue that interactions among
incompatible tasks may contribute to a greater risk
than the sum of the base-level task risks alone.
Unfortunately, no study has quantified these potential
interactions.
The objective of the present study was to quantify
the impact that the interactions of common highway
construction tasks have on base-level safety risk levels.
Risk interactions are defined as the pair-wise impacts
that tasks have on each other due to task compatibility
or incompatibility. Interactions were measured as the
percentage increase or decrease in safety risk resulting from
the concurrent performance of the tasks in the same physical
workspace
. The research focused on the highway
construction sector because this is one of the most
dangerous in the construction industry (Bai, 2002;
Bureau of Labor Statistics, 2010) and highway
construction tasks are limited in number and well
defined (Pandey, 2009).
Literature review
Spatial and temporal interactions
Traditionally, safety has not received the attention that
it deserves in comparison with other objectives in
jobsite planning (Anumba and Bishop, 1997).
Recently, however, researchers have begun to study the
impact of site layout schemes on safety performance.
For example, Shapira and Lyachin (2009) showed that
crowded jobsites, resources constraints and overlap of
activities may increase safety risks. In an effort to inte-
grate safety into site layout planning, Elbeltagi
et al.
(2004) presented a method of modelling safety zones
around temporary facilities. They used genetic algo-
rithms to optimize the distances between facilities in
order to minimize their negative interactions. Similarly,
El-Rayes and Khalafallah (2005) suggested a model to
consider the influence of crane operations, hazardous
materials and travel routes on safety. Navon and
Kolton (2006) took a different approach and showed
how interactions among site layouts and planned tasks
can produce fall hazards. This body of literature
confirms the importance of studying risk interactions
but has two main limitations: (1) the models are
conceptual and are not based upon an underlying data-
base; and (2) the interactions among tasks were ignored
in the quantitative analyses.
While spatial safety management typically occurs
during site layout planning, temporal safety manage-
ment typically involves safety–schedule integration.
Kartam (1997) made the first attempt to integrate
safety data into schedules; however, as Hinze
et al.
(2005b) recognized, there was not an actual relation
between schedule and safety resources in Kartam’s
model. Consequently, Hinze
et al.
(2005b) developed
software called SalusLink which allows safety person-
nel to load safety components into the schedule of a
project. Taking schedule integration a step further, Yi
and Langford (2006) suggested a framework to inte-
grate safety risk into schedules using a similar method
to resource loading (i.e. assigning a safety risk quan-
tity to each scheduled activity). This framework can
be used to identify periods with a relatively high level
of safety risk and allows managers to use resource
levelling techniques to level the safety risk in a sched-
ule. Similar to the spatial modelling of safety, these
schedule-based techniques are not based on robust
underlying data nor do they consider the interactions
among tasks.
Sacks
et al.
(2009) recently proposed CHASTE, a
model that simultaneously considers spatial and tempo-
ral interactions of work tasks. By using information
available in 4D geographic models and user-provided
data for ‘loss-of-control events’, the method can be used
Safety risk management
419
to produce a 4D view of the regions of the worksite with
high levels of safety risk (Sacks
et al.
, 2009). The most
significant limitation of this framework is that, in order
to quantify the risk for ‘loss-of-control events’, the
hazards related to each task must be identified and
quantified by the user, which can be time intensive and
laborious. As discussed by Jannadi and Almishari
(2003), quantifying these risk values is not practical for
most firms. To address the limitations in the current
body of literature and to enhance the efficacy of the
aforementioned safety integration models and frame-
works, a database of task interactions was created.
Safety risk quantification
The prevailing methods of injury risk assessment typi-
cally involve qualitative risk ratings on either linguistic
or numerical scales (e.g. Hallowell and Gambatese,
2009). Typically, injury risks are evaluated using a
combination of frequency ratings, severity ratings and
exposure durations. When sufficient historical data are
available, safety risk can be calculated by finding the
product of likelihood of occurrence and magnitude of
impact (Baradan and Usmen, 2006; Navon and
Kolton, 2006). To date, no research has evaluated the
impact of risk interactions in risk assessment. Rather,
base-level task risks are evaluated individually and are
rarely aggregated.
The methods used to obtain risk data and the units
of analysis are diverse in existing literature. For exam-
ple, Jannadi and Almishari (2003) considered risks
posed by construction activities, equipment, hazardous
substances and external stimuli to estimate total safety
risk on worksites; Baradan and Usmen (2006) used
American Bureau of Labor Statistics (BLS) data to
analyse safety risk in 16 different construction trades;
Hallowell and Gambatese (2009) quantified risk at the
activity level using the Delphi method; Gürcanli and
Müngen (2009) proposed a fuzzy rule based system to
analyse safety risk with linguistic variables; and Rozen-
feld
et al.
(2010) used a technique similar to job hazard
analysis to identify loss of control events for 14
construction activities. As previously indicated, one of
the major limitations of this previous work is that risk
analyses consider activities, tasks and processes to be
independent. That is, safety risks are quantified for
individual tasks in isolation without considering the
impacts of other concurrent tasks.
Research method
In order to develop an appropriate scope for data
collection, clear definitions of common highway
construction work tasks were needed. Recently, Pandey
(2009) used data from literature, project schedules and
interviews to identify and describe 25 common high-
way construction tasks (see Table 1). As will be
described in detail, the interactions among these 25
highway construction tasks were quantified using the
Delphi method.
The Delphi method is a systematic and interactive
research strategy for achieving consensus among a
panel of experts. With this technique, panellists are
selected according to specific guidelines and are invited
to participate in two or more rounds of structured
surveys. After each round, an anonymous summary of
the experts’ input from the previous survey is provided
as feedback to the panel. In each subsequent round,
participants are encouraged to review the feedback
provided by the other panellists and consider revising
their previous response. The process is concluded after
a pre-defined criterion (e.g. number of rounds or the
achievement of consensus) is achieved.
The Delphi method was selected over alternative
research methods because archival data are incomplete
(Bureau of Labor Statistics, 2010; Shapira and
Lyachin, 2009; Rozenfeld
et al.
, 2010), empirical data
could not be obtained during a realistic timeframe, and
because the Delphi method is preferred when attempt-
ing to obtain complex data that cannot be separated
from project context due to confounding factors
(Linstone and Turoff, 1975). Furthermore, the Delphi
method has seen increased use over the past decade for
construction engineering and management research
(Hallowell and Gambatese, 2010). In fact, this method
has been successfully employed to enhance bridge
condition assessments and predict remaining service
life (Saito and Sinha, 1991), select procurement
systems for construction projects (Chan
et al.
, 2001),
identify and evaluate factors affecting international
construction (Gunhan and Arditi, 2005), identify
components and characteristics of supply change flexi-
bility (Lummus
et al.
, 2005), quantify indicators for
measuring partnering performance (Yeung
et al.
,
2008), and to select contractors using qualitative
measures (Manoliadis
et al.
, 2009).
Expertise requirements
The careful selection of expert panellists is one of the
most important aspects of the Delphi method. A well-
qualified, well-rounded and diverse panel of experts is
essential to ensure minimal bias and maximum inter-
nal and external validity. A review of literature reveals
various methods to qualify an individual as an ‘expert’
using objective criteria. Though Rogers and Lopez
(2002) and Linstone and Turoff (1975) are the two
most commonly cited references when selecting
expertise requirements, these publications offer very
420
Hallowell
et al
.
different sets of requirements. To address these
inconsistencies, Hallowell and Gambatese (2010)
created a new set of objective and flexible require-
ments that can be used when certifying potential
panellists as ‘experts’ in the field of construction engi-
neering and management. According to this study
every panellist must score at least 12 total points in
the related field of research using the point system
shown in Table 2 to qualify.
To ensure that the panel is well rounded and profes-
sionally oriented every panellist was required to have at
least eight years of professional experience in the archi-
tecture, engineering and construction (AEC) industry.
It was expected that safety managers, project managers,
safety officers, Occupational Safety and Health Admin-
istration (OSHA) representatives, construction safety
and health researchers and representatives from work-
ers compensation insurance providers would be the
most highly qualified panellists.
More than 500 potential experts in the field of high-
way safety risk management were identified. Contact
information for potential experts was gathered mainly
from The National Work Zone Safety Information
Clearinghouse website (www.workzonesafety.org),
Table 1
Highway reconstruction work tasks and descriptions (after Pandey, 2009)
Work tasks in highway
reconstruction Description
Clear and grub Clearing vegetation, debris and existing structures (e.g. abandoned utility services)
Excavation Excavating and constructing embankments and the construction of erosion control
devices
Demolition of existing pavement Removing existing pavement
Landscape Preparing soil, mulching and constructing irrigation systems
Watering and dust palliatives Applying water for density and moisture control of soil, applying palliatives for dust
control, and soil stabilization
Reset structures Installing guardrails, fencing, cattle guards, delineators and lighting
Lay aggregate base course Furnishing and placing one or more courses of additives on a prepared sub grade
Recondition bases (compaction) Blading, shaping, wetting and compacting the existing sub grade
Installing flexible pavement/
patching
Laying hot mix asphalt and installing geosynthetics beneath pavements
Install rigid pavement (concrete) Forming, pouring, floating and finishing rigid pavement
Heat and scarifying Recycling the top portion of existing bituminous pavement by cleaning, heating,
scarifying, re-levelling, compacting and rejuvenating existing pavement
Recycle cold bituminous
pavement
Pulverizing the existing bituminous pavement, surfacing to the required depth and mixing
a recycling agent with water
Prime, coat, rejuvenate
pavement
Preparing and treating an existing pavement surface with bituminous and blotter materials
Seal joints and cracks Furnishing and placing hot-poured joint and crack sealant in properly prepared cracks in
asphalt pavements
Install cribbing Installing concrete cribbing, rip rap and paving slopes/ditches
Install culverts, subsurface
drains and maintain sewers
Constructing culverts, sewers, storm drains, under drains, edge drains, geocomposite
drains and French drains
Install curb and gutters Installing curb and gutters, constructing sidewalks and bikeways and installing median
cover material
Install traffic control devices Constructing signs, signals, street markings and other restriction systems that regulate and
guide traffic
Install water control devices Constructing water and erosion control devices
Install culvert pipe and water
lines
Constructing culvert pipe and installing of water lines
Install field facilities Installing field offices, laboratories and sanitary facilities on the worksite
Survey Surveying the worksite during planning, construction and operation
Mobilization/demobilization Mobilizing and demobilizing personnel and equipment
Pavement marking Furnishing and applying pavement markings and removing existing markings
Construction zone traffic
control
Preparing or removing lane closures, flagging, traffic diversions, cones, delineators,
barricades, sign stands, flashing beacons, flashing arrow trailers and changeable message
signs
Safety risk management
421
OSHA, Departments of Transportation (DOTs), the
Federal Highway Administration (FHWA), Associated
General Contractors (AGC) and university websites.
Invitation e-mails that included basic information for
the project and estimated time commitments were sent
to all potential experts. Of the initial pool of 500 poten-
tial experts, 57 individuals agreed to participate. All 57
potential experts were asked to fill out an introductory
survey that solicited information that was later used to
assess each individual’s level of expertise. Of the 51
introductory surveys that were received, 37 individuals
were certified as experts using the aforementioned
criteria and were randomly assigned to one of three
panels.
The 37 experts had an average of over 21 years of
professional experience with highway work zone safety
management. Approximately 80% of the respondents
are professional engineers (PE), certified safety profes-
sionals (CSP), or have at least a bachelor’s degree in a
related field. In addition to the professional experience
of respondents, the panel has collectively authored 457
conference papers and 45 peer-reviewed journal arti-
cles on safety or risk-related topics. Moreover, the
panel was geographically dispersed including all major
regions of the United States except for Alaska and
Hawaii (i.e. the contiguous United States).
Number of panellists
The number of panellists has varied in previous studies
from 3 to 80 (Rowe and Wright, 1999). In fact, the
number of panellists is affected by the volume of data
targeted, timeframe of the research, number of
accessible experts in the field and the capability of the
facilitator to handle the panellists (Linstone and Turoff,
1975). The relationship between the number of
panellists and accuracy of the results was investigated
by Brockhoff (1975) and Boje and Murnighan (1982).
These studies found that optimum number of panellists
ranges from 8 to 15. A range between 10 and 13 was
targeted for this study because this number ensures an
adequate population if a member defaults during
the process, is easily manageable and ensures a high
level of internal and external validity (Rajendran and
Gambatese, 2009).
Owing to the great volume of data required for this
study (i.e. aggregated ratings of 600 interactions), the
authors elected to conduct the study using three inde-
pendent panels with 12 or more panellists each. Two
panels were responsible for quantifying the pair-wise
interaction among eight tasks (i.e. 192 ratings) each
while the third panel quantified the pair-wise interac-
tions among nine tasks (i.e. 216 ratings). The task
interactions were randomly assigned to each panel
using a pseudo random number generator in Microsoft
Excel.
Number of iterations and feedback process
There are two prominent reasons to conduct multiple
iterations of surveys during the Delphi process: reach-
ing consensus by reducing variance and improving
precision (Hallowell and Gambatese, 2010). The
number of rounds and methods used to measure
consensus has been seen as an indicator for accuracy
of the Delphi method. The number of iterations in
previous large-scale studies ranged from two to six
(Dalkey, 1972; Linstone and Turoff, 1975; Gupta
and Clarke, 1996). Over half of these studies found
acceptable convergence after three or fewer iterations.
Hallowell and Gambatese (2010) suggested that a
study with three iterations is ideal because expert
panellists may review reasons for outlying responses
in the third and final round thereby minimizing
several forms of cognitive bias. Thus, this Delphi
study was designed to include three initial rounds of
data collection and a fourth round to cross-validate
the results. A description of each round is provided in
Table 3.
A feature of the Delphi method that distinguishes it
from other similar methods is providing anonymous
feedback to decrease the potential impacts of cognitive
bias. Providing anonymous feedback facilitates indirect
communication among panellists in an effort to reach a
high level of consensus (Linstone and Turoff, 1975;
Chan
et al.
, 2001). Research has been conducted to
evaluate the effects of different forms of feedback on
accuracy of final results (Best, 1974; Rowe and Wright,
1999). These studies found that Delphi studies lead to
more accurate results when reasons and simple statisti-
cal summaries are included in feedback. For the
present study, medians and reasons for outlying
Table 2
Flexible point system for the qualification of expert
panellists (after Hallowell and Gambatese, 2010)
Achievement or experience Points (each)
Professional registration 3
Year of professional experience 1
Conference presentation 0.5
Member of a committee 1
Chair of a committee 3
Peer-reviewed journal article 2
Faculty member at an accredited university 3
Author/editor of a book 4
Author of a book chapter 2
Advanced degrees:
BS 4
MS 2
PhD 4
422
Hallowell
et al
.
responses have been chosen as feedback because
median responses are impacted on very little by biased
responses and reviewing and providing reasons for
outlying responses requires deeper thinking about more
complex interactions. The specific feedback provided
in each round is provided in Table 3.
Methods to minimize bias
The research team held the minimization of cogni-
tive bias paramount because the validity and reliabil-
ity of the Delphi process depends on the unbiased
judgment of its experts. Various sources of bias may
exist despite the panellists’ status as certified experts.
Identifying potential cognitive biases that affect one’s
ability to accurately rate risk values is essential
because it allows the research team to strategically
design the Delphi process in such a way that potential
biases are minimized. Any panellist is likely to be
susceptible to one or more of the following eight
forms of judgment-based bias during the Delphi
process: collective unconscious, contrast effect,
neglect of probability, Von Restorff effect, myside
bias, recency effect, primacy effect and dominance
(Hallowell and Gambatese, 2010). Literature suggests
several different methods to avoid the cognitive biases
listed above. Specific controls that apply to this study
include: (1) maintaining the anonymity of the respon-
dents; (2) providing reasons as a part of the
controlled feedback; (3) reporting results as medians
rather than means; (4) randomizing the question
order of the surveys. It was expected that these
controls would reduce the potential effects of cogni-
tive bias thereby enhancing the reliability and validity
of the results.
Results
In each round of the Delphi process, experts were
asked to provide 192 or 216 ratings, depending on their
panel assignments. Of the 37 experts who agreed to
participate in this research effort, 28 completed all
survey rounds resulting in an ultimate Delphi response
rate of 76%. In total, over 5900 ratings were obtained
per round resulting in a total of 17 776 ratings after the
three rounds of initial data collection. The validation
effort conducted in the fourth round required an addi-
tional 5900 ratings.
One of the goals of the Delphi process is to reach
consensus; however, measures of consensus are not
consistent in previous studies. Lummus
et al.
(2005)
compared changes in standard deviations between
rounds and conducted t-tests to measure level of signif-
icance. Another test that has been used to assess level
of agreement between panellists in Delphi research is
Kendall’s coefficient of concordance (W) (Chan
et al.
,
2001; Yeung
et al.
, 2007, 2008; Hon
et al.
, 2010).
Using Kendall’s coefficient to measure consensus is not
appropriate for this study because the test is designed
to measure the level of concordance among rankings
with few ties within the resulting database. The data
targeted, however, are ratings of pair-wise influence.
Ties among ratings were welcomed for interactions of
the same magnitude. Thus, the absolute deviation (i.e.
average deviation from the median) alone was used as
a measure of consensus, which is consistent with
Delphi studies with similar data profiles (e.g. Hallowell
and Gambatese, 2009).
Prior to initiating the Delphi process, the research
team set the goal to reach an absolute variance of less
than 5% for all three Delphi panels after the third
Table 3
Iterations of the Delphi process
Duration
(days)
Description
Introductory
survey
15 Individuals were asked to fill out introductory surveys that solicited information used to
objectively qualify potential panellists as experts.
Round 1 30 Panellists were asked to rate the pair-wise interactions among randomly assigned tasks (i.e. the
increase or decrease in base-level safety risk resulting from compatibility or incompatibility of
work tasks).
Round 2 30 Medians responses and personal ratings from round 1 were provided as feedback for round 2.
Panellists were asked to provide written reasons for round 2 ratings that they believe were
≥
10% greater or less than the median response from round 1.
Round 3 30 In addition to medians and personal ratings from round 2, reasons supplied by experts for
outlying responses were included for consideration. Panellists were given the opportunity to
review medians and reasons for outlying responses. The median ratings from this round
represented the final aggregated rating.
Round 4
(validation)
30 Panellists were asked to evaluate the median responses provided by one of the other panels.
Panellists had the opportunity to accept the medians provided by other panellists or choose a
new rating.
Safety risk management
423
round with a 95% agreement in the validation ratings.
The absolute variance for each panel after each round
is shown in Table 4. It can be seen that the target
consensus of <5% was achieved for all panels after the
third round of Delphi surveys. Notably, medians did
not change from round 2 to round 3.
The resulting dataset (after round 3) is shown in
Table 5. Each median rating in Table 5 represents the
aggregate of at least eight expert panellists’ ratings.
These ratings are the percentage increase or decrease in
effectiveness that result from the concurrent perfor-
mance of two tasks in a proximate physical space. One
should note that for each interaction, two different
ratings exist in the database. Two ratings are provided
for each interaction because the effect of activity A on
activity B is not necessarily equal to the effect of activity
B on A. For example, when laying aggregate base
course and installing rigid pavement are performed
simultaneously in overlapping physical work spaces, the
base-level safety risk of laying aggregate base course
increases by 40% while the base-level safety risk of
installing rigid pavement increases by only 20%. The
range of the interactions is from
−
5% up to 60%. The
only compatible interaction is the effect of pavement
marking on watering and dust palliatives (
−
5%). This
shows that performing different activities at the same
time will usually increase safety risk. For some activi-
ties, the interaction is zero, which means that there is
no risk interaction when the task pairs are concurrently
implemented.
As indicated, experts were asked to provide reasons
for outlying responses during the Delphi process.
Though there was a high degree of consensus after the
three rounds of surveys, several respondents provided
compelling reasons for outlying responses. For exam-
ple, a few experts believed that safety risk interactions
among specific tasks (e.g. mobilization and demolition
of existing pavement) should be rated higher because of
the concurrence of equipment intensive tasks, using
heavy and noisy machinery and changing of traffic
patterns. Overlap between such attributes was thought
to increase the chance of spatial interference and,
consequently, safety risk. Another example involved
the interaction between tasks with heavy materials and
noisy machinery. Such tasks were thought to impact on
other tasks more than the median rating because
communications among workers becomes more
difficult. Finally, construction zone traffic control was
expected to increase the risk of other tasks because of
changing of traffic patterns.
In addition to the increases in risk interactions, some
experts provided reasons why some interactions should
have lower ratings. For example, one of the experts stated
that resetting structures takes place primarily beyond the
shoulder of the road while prime, coating and rejuve-
nating pavement would occur on the roadway. Conse-
quently, it is unlikely that these tasks would have a spatial
interaction. Similarly, watering and dust palliatives and
pavement marking are very unlikely to be performed
concurrently in the same location on a project. Other
experts noted that it would be very unrealistic for some
tasks to be performed concurrently due to typical
construction sequencing. However, the research team
purposefully did not remove any task interactions from
the analysis to minimize the potential for bias from the
research team and to preserve a comprehensive dataset.
Analysis
By summing the rows and columns of this matrix
(Table 5), one can evaluate the impacts that each task
has on the others and the extent to which each task is
affected by the presence of others. The results of this
analysis are provided in Table 6. It should be noted
that the measures in this table are a unit-less relative
measure of influence. Three activities: construction
zone traffic control (7.10), installing flexible pavement/
patching (6.20) and excavation (6.10) have the greatest
impact on the other activities. Thus, when performing
these tasks simultaneously with other tasks, there is a
great increase in the base-level risk of the other activi-
ties. Additionally, the base-level safety risk installing
rigid pavement (concrete) (6.20), demolition of exist-
ing pavement (6.10) and recondition bases (compac-
tion) (5.90) are affected most by the presence of other
activities. Another interesting finding is that construc-
tion zone traffic control has the most significant impact
on base-level risk of all tasks and yields the most unsta-
ble work environment. Alternatively, installing field
and facilities is the most stable task because it is
affected the least by presence of other tasks and has the
lowest effect on other tasks.
An analysis of the distribution of interaction ratings
produced interesting results. The average, median and
standard deviation of all ratings were 0.17, 0.2 and
0.14 respectively. Twenty-nine per cent of the ratings
were between 0.0 and 0.1, 40% were between 0.2 and
0.3 and 11% were between 0.4 and 0.5.
Another analysis was performed to identify the most
significant two-way interactions. The magnitude of
Table 4
Absolute variance of responses for panels in
different rounds (percent deviation)
Round 1 Round 2 Round 3
Panel 1 14.67% 5.71% 4.95%
Panel 2 30.39 9.94 4.22
Panel 3 34.31 6.42 1.35
424
Hallowell
et al
.
T
a
b
le
5
M
ed
ia
n
s
af
et
y
ri
sk
i
n
te
ra
ct
io
n
a
ft
er
r
o
u
n
d
3
(
p
er
ce
n
t
in
cr
ea
se
i
n
b
as
e-
le
ve
l
ri
sk
)
T
as
ks
E
ff
ec
t
o
n
Clear and grub
Excavation
Demolition of existing pavement
Landscape
Watering and dust palliatives
Reset structures
Lay aggregate base course
Recondition bases (compaction)
Installing pavement/patching
Install rigid pavement (concrete)
Heat and scarifying
Recycle cold bituminous pavement
Prime, coat, rejuvenate pavement
Seal joints and cracks
Install cribbing
Install culverts, drains and sewers
Install curb and gutters
Install traffic control devices
Install water control devices
Install culvert pipe and water lines
Install field facilities
Survey
Mobilization/demobilization
Pavement marking
Construction zone traffic control
Effect of
C
le
ar
a
n
d
g
ru
b
2
0
2
0
1
0
0
1
0
2
0
2
0
1
0
0
0
0
0
0
4
0
1
0
0
0
2
0
2
0
0
2
0
0
1
0
0
E
xc
av
at
io
n
4
0
4
0
4
0
4
0
3
0
2
0
2
0
2
0
2
0
2
0
2
0
2
0
2
0
4
0
4
0
2
0
3
0
2
0
4
0
1
0
2
0
0
2
0
2
0
D
em
o
li
ti
o
n
o
f
ex
is
ti
n
g
p
av
em
en
t
2
0
2
0
2
0
2
0
4
0
2
0
2
0
2
0
4
0
4
0
2
0
2
0
0
0
2
0
3
0
3
0
2
0
2
0
0
2
0
0
1
0
4
0
L
an
d
sc
ap
e
4
0
4
0
1
0
2
0
0
2
0
0
2
0
0
0
0
0
0
2
0
2
0
0
0
2
0
2
0
0
1
0
0
2
0
0
W
at
er
in
g
an
d
d
u
st
p
al
li
at
iv
es
0
2
0
2
0
0
2
0
2
0
2
0
0
0
2
0
0
0
0
0
1
0
0
1
0
0
0
0
1
0
0
0
0
R
es
et
s
tr
u
ct
u
re
s
2
0
2
0
4
0
2
0
2
0
2
0
2
0
2
0
6
0
4
0
3
0
6
0
2
0
0
2
0
3
0
2
0
2
0
2
0
0
2
0
0
2
0
4
0
L
ay
a
gg
re
ga
te
b
as
e
co
u
rs
e
0
2
0
3
0
1
0
0
2
0
3
0
2
0
4
0
3
0
5
0
2
0
0
0
2
0
3
0
3
0
2
0
4
0
0
2
0
0
1
0
4
0
R
ec
o
n
d
it
io
n
b
as
es
(
co
m
p
ac
ti
o
n
)
2
0
2
0
3
0
2
0
2
0
2
0
2
0
2
0
4
0
2
0
3
0
3
0
0
2
0
2
0
3
0
4
0
2
0
2
0
0
2
0
0
2
0
2
0
In
st
al
li
n
g
p
av
em
en
t/
p
at
ch
in
g
0
2
0
2
0
2
0
2
0
4
0
2
0
5
0
5
0
4
0
5
0
5
0
4
0
2
0
2
0
2
0
4
0
1
0
2
0
0
1
0
0
2
0
4
0
In
st
al
l
ri
gi
d
p
av
em
en
t
(c
o
n
cr
et
e)
0
2
0
2
0
1
0
2
0
2
0
2
0
4
0
2
0
4
0
4
0
1
5
0
0
2
0
6
0
3
0
2
0
2
0
0
1
0
0
2
0
6
0
H
ea
t
an
d
s
ca
ri
fy
in
g
0
2
0
2
0
1
0
2
0
2
0
1
0
3
0
2
0
2
0
3
0
2
0
2
0
0
1
0
3
0
3
0
2
0
2
0
0
1
0
0
2
0
2
0
R
ec
yc
le
c
o
ld
b
it
u
m
in
o
u
s
p
av
em
en
t
0
2
0
3
0
2
0
1
0
1
0
2
0
4
0
2
0
2
0
2
0
4
0
2
0
2
0
2
0
3
0
2
0
2
0
4
0
0
1
0
0
1
0
3
5
P
ri
m
e,
c
o
at
,
re
ju
ve
n
at
e
p
av
em
en
t
0
0
2
0
1
0
0
2
0
2
0
4
0
2
0
3
0
4
0
3
0
2
0
0
2
0
3
0
2
0
1
0
2
0
0
2
0
0
2
0
4
0
S
ea
l
jo
in
ts
a
n
d
c
ra
ck
s
0
0
3
0
0
0
3
0
2
0
2
0
2
0
4
0
2
0
4
0
4
0
0
2
0
1
0
2
0
1
0
2
0
0
0
2
0
2
0
4
0
In
st
al
l
cr
ib
b
in
g
2
0
2
0
2
0
2
0
2
0
2
0
1
0
0
2
0
2
0
0
2
0
0
0
2
0
1
0
1
0
2
0
2
0
0
1
0
0
2
0
2
0
In
st
al
l
cu
lv
er
ts
,
d
ra
in
s
an
d
s
ew
er
s
2
0
4
0
4
0
2
0
2
0
3
0
2
0
4
0
2
0
3
0
0
0
2
0
2
0
2
0
4
0
1
0
2
0
2
0
0
2
0
0
2
0
2
0
In
st
al
l
cu
rb
a
n
d
g
u
tt
er
s
0
2
0
4
0
2
0
2
0
3
0
2
0
4
0
2
0
6
0
4
0
4
0
2
0
2
0
0
3
0
3
0
2
0
2
0
0
2
0
0
2
0
2
0
In
st
al
l
tr
af
fi
c
co
n
tr
o
l
d
ev
ic
es
0
2
0
2
0
2
0
2
0
0
2
0
4
0
3
0
3
0
6
0
1
0
2
0
2
0
0
3
0
4
0
2
0
2
0
0
1
0
2
0
3
0
4
0
In
st
al
l
w
at
er
c
o
n
tr
o
l
d
ev
ic
es
2
0
4
0
2
0
2
0
2
0
1
0
1
0
1
0
0
2
0
0
1
0
0
0
2
0
2
0
3
0
1
0
2
0
0
0
0
1
0
2
0
In
st
al
l c
u
lv
er
t
p
ip
e
an
d
w
at
er
li
n
es
4
0
4
0
4
0
2
0
2
0
2
0
2
0
3
0
2
0
3
0
1
0
2
0
2
0
2
0
2
0
3
0
3
0
0
2
0
0
2
0
2
0
2
0
2
0
In
st
al
l
fi
el
d
f
ac
il
it
ie
s
0
0
1
0
0
0
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
2
0
0
0
S
u
rv
ey
0
2
0
2
0
1
0
0
0
2
0
0
1
0
0
2
0
0
2
0
0
0
0
0
0
2
0
0
0
1
0
0
0
M
o
b
il
iz
at
io
n
/d
em
o
b
il
iz
at
io
n
0
0
1
0
1
0
0
0
2
0
1
0
1
0
0
0
0
0
0
0
2
0
0
0
1
0
0
0
0
2
0
0
P
av
em
en
t
m
ar
ki
n
g
0
0
2
0
1
0
–5
2
0
1
0
3
0
2
0
2
0
4
0
4
0
3
0
4
5
0
1
0
3
0
3
0
0
2
0
0
2
0
0
2
0
C
o
n
st
ru
ct
io
n
z
o
n
e
tr
af
fi
c
co
n
tr
o
l
0
2
0
4
0
0
3
0
2
0
4
0
4
0
5
0
5
0
5
0
5
0
2
0
6
0
0
2
0
3
0
6
0
2
0
2
0
0
3
0
2
0
4
0
Safety risk management
425
these two-way interactions was calculated by summing
both interactions for each pair. For example, if demoli-
tion increases the base-level risk of excavation by 20%
and excavation increases the base-level risk of demoli-
tion by 40%, the magnitude of the two-way interaction
for this pair would be 60. This two-way interaction
value is a relative, unit-less measure that quantifies the
relative magnitude of a two-way interaction between
two tasks. Of these two-way interactions, the most
significant are installing rigid pavement and installing
curb and gutters (120); installing rigid pavement and
construction zone traffic control (110); construction
zone traffic control and sealing joints and cracks (100);
construction zone traffic control and installing traffic
control devices (100); construction zone traffic control
and installing pavement and patching (90); and install-
ing traffic control devices and heating and scarifying
(90). Interestingly, there were 45 two-way interactions
with a magnitude of zero indicating that there are a
significant number of neutral interactions. Finally,
there was one two-way interaction, pavement marking
and watering and dust palliatives (–5), that is compati-
ble indicating that overlapping these two tasks in the
project schedule
decreases
the base-level safety risk. It
should be noted that the actual impact that these two-
way interactions have on site safety depends on the
magnitude of the base-level risks.
Validation
As previously stated, the members of three distinct
panels of experts and each panel were asked to provide
safety risk interaction ratings for 196 or 216 interac-
tions. The first three rounds focused on obtaining
initial interaction ratings while the fourth or final round
was used to cross-validate the resulting matrix. To
perform this validation, surveys similar in structure to
the initial Delphi surveys were distributed. In the vali-
dation round, experts were asked to review the round 3
responses from a different panel that rated a completely
different set of interactions. Panellists were given the
option to agree with the other group’s collective assess-
ment or provide a new rating. In order to decrease bias,
surveys were randomly assigned to the panellists. The
only limitation was that no panellist was allowed to rate
the same interactions that they were assigned during
the initial Delphi process. Of the 28 surveys that were
Table 6
The cumulative effect of activities on each other (relative unit-less measure)
Tasks Total effects of the activity on
other activities
Total effects of other activities on
the activity
Clear and grub 2.30 2.40
Excavation 6.10 4.60
Demolition of existing pavement 4.90 6.10
Landscape 2.60 3.40
Watering and dust palliatives 1.50 3.35
Reset structures 5.80 4.50
Lay aggregate base course 4.80 4.40
Recondition bases (compaction) 5.00 5.90
Installing flexible pavement/patching 6.20 4.30
Install rigid pavement (concrete) 5.05 6.20
Heat and scarifying 4.00 5.50
Recycle cold bituminous pavement 4.75 5.30
Prime, coat, rejuvenate pavement 4.30 4.65
Seal joints and cracks 4.20 3.25
Install cribbing 3.20 2.20
Install culverts, drains and sewers 4.90 4.50
Install curb and gutters 5.50 5.30
Install traffic control devices 5.20 4.70
Install water control devices 3.10 3.80
Install culvert pipe and water lines 5.30 4.60
Install field facilities 0.50 0.10
Survey 1.50 3.30
Mobilization/demobilization 1.10 1.10
Pavement marking 4.10 4.00
Construction zone traffic control 7.10 5.55
426
Hallowell
et al
.
sent, 27 surveys were returned resulting in a 96%
response rate for the validation. One month was allo-
cated for this validation process and a total of 5276
ratings were obtained. Absolute variance of responses
for each panels have been calculated 0.6, 1.13 and
0.9% for panel 1, 2 and 3, respectively. Additionally,
medians from validation were the same as medians of
each panel, which is evidence of strong validation.
Application of results
There are several potential applications of this database
to safety management and planning, which served as the
impetus for this research effort. One of the most impor-
tant aspects of the interaction database is its application
to project schedule integration. Previously, researchers
have developed a model for integrating safety risk data
into project schedules (Yi and Langford, 2006; Sacks
et
al.
, 2009). This technique involves safety-loading the
risk data with the project schedule using the same strat-
egy as resource loading a schedule. This framework is
mathematically summarized in Equation 1.
where:
[
SF
] is an ultimate safety risk matrix and its members
are total safety risk for each time unit (day, week and
month);
[
R
Individual
] is a matrix which includes safety risk
values related to performing each task individually;
[
X
Schedule & Time
] is a matrix which includes just 0 and
1. If in time t, activity i is performing, then
X
it
= 1,
otherwise
X
it
= 0.
Unfortunately, this model and available safety risk
data only allow one to model the independent, base-
level risks associated with various work tasks and does
not account for the influence that multiple concurrent
work tasks can have on one another. With the new
dataset in Table 5, each task risk can be adjusted by
multiplying the base-level risk by all interaction values
for all concurrent tasks. The new data from Table 5 can
be incorporated into a schedule analysis using a modi-
fication of Yi and Langford’s (2006) framework shown
in Equation 2.
where:
[
SF
Task
] is a safety risk matrix resulting from
performing tasks by considering interaction among
them and its members are total safety risk for each time
unit (day, week and month);
[
R
Individual
] is a matrix which includes safety risk
values related to performing each task individually;
[
R
Interaction
] is a matrix (Table 5) which includes
impact of performing each task simultaneously with
other tasks on safety risk values of other tasks;
[
X
Schedule
] is a matrix which includes just 0 and 1. If
in time t, activity i is performing, then
X
it
= 1, other-
wise
X
it
= 0.
In this new framework, the safety risk data, which
include spatial and temporal interactions of work tasks,
can be simply integrated with the schedule and the
safety risk can be plotted over time. This method can
be used to identify high risk periods that may not be
identified intuitively. In response, contractors can
attempt to consume float to level risk, take extra
precautionary measures during these high risk periods
(e.g. lane closure), inform workers of high risk periods
and strategically design injury prevention strategies to
focus on high risk tasks. When the risk profiles for
multiple concurrent projects are overlaid in the same
plot a manager can identify when and where safety
resources should be deployed and could evaluate the
risk profile for the company’s portfolio simply by
computing the cumulative risks for all projects in the
company’s programme and plotting the risk over time.
In addition to integrating these safety data into
schedules, risk interaction values can be applied to
information models. For example, safety risk data for
specific construction tasks and the task interaction data
can be assigned to temporal and spatial elements of the
model in the same way as cost, duration, quality, mate-
rial and other data. These data are essential to identify
high risk locations and time periods based on the
planned sequence and location of tasks. The collection
and dissemination of the risk data presented takes a
major step towards the creation of a safety information
model. Though the dataset presented does not include
tasks associated with building construction tasks as
would be necessary to integrate with building informa-
tion models, the research methods and framework
could be applied to a future study on the topic.
Several limitations to the application of the results
should be noted. First, the interaction data may only
be applied to the highway work tasks as they are
described in Table 1. Though these task descriptions
are representative of typical work scenarios, as
described by Pandey (2009), the data presented are
not representative of any deviations from these stan-
dard procedures. For example, if a crane were used to
install field facilities or if excavations were unusually
deep, the magnitude of any task interactions associated
with these deviations may no longer be accurate. This
limitation was essential because adding new criteria to
SF R X
n Individual Schedule Time xn
[ ] = [ ] × [ ]× ×1 1 25 25& (1)
SF R
R X
Task n Individual
Interaction Schedule n
[ ] = [ ]
× [ ] ×[ ]
× ×
× ×
1 1 25
25 25 25
(2)
Safety risk management
427
the Delphi survey would have resulted in an over-
whelming burden to the panellists, each of whom had
already been asked to provide 1800 ratings over the
course of three rounds of surveys. Because of this limi-
tation, the writers suggest future research on the
impacts of relevant subtasks, alternative means and
methods, and specialty equipment. The second major
limitation is that these data should only be applied to
daytime construction on projects in the contiguous
United States. The limitation must be imposed
because the Delphi panellists only had significant
experience in the contiguous United States and
construction deviates significantly from standard
means and methods when work is performed at night.
Finally, the data must be applied with the understand-
ing that the task descriptions are general in nature and
do not reference specific design features, environmen-
tal conditions, crew capabilities and competencies, or
any other project-specific characteristics.
Conclusions and study limitations
The research objective was to quantify the pair-wise
safety interactions among 25 highway construction
work tasks that result from task compatibility or incom-
patibility using the Delphi method. After three itera-
tions of Delphi surveys with three separate panels,
consensus was achieved. In a fourth and final round,
the results were successfully cross-validated.
The results of this research indicate that construction
zone traffic control, installing flexible pavement/patch-
ing and excavation have the greatest impact on the
base-level risk of other construction activities. In
contrast, installing rigid pavement (concrete), demoli-
tion of existing pavement and reconditioning bases
(compaction) are affected most by other concurrent
activities. Though the pair-wise data are interesting and
valuable on their own, the most significant contribution
is that these data can be effectively integrated with cost,
schedule and quality planning. As discussed, the data-
base produced can be attached, along with base-level
safety risk data to common highway construction work
tasks in a project schedule thereby allowing a manager
to ‘safety-risk-load’ a project schedule in the same way
one would resource load a schedule. The risk interac-
tion data can be used to more accurately quantify
temporal safety risk on projects with many concurrent
tasks. The resulting temporal plot includes the base-
level safety risk and the influence that multiple concur-
rent work activities have on each other’s risk level.
Though it may be unrealistic to separate concurrent
construction tasks, such an analysis may yield more
accurate and reliable temporal risk analyses. Being able
to proactively identify high risk periods and communi-
cate risks with construction crews is very important for
successful safety management.
There are several limitations of this research. First,
though several controls were implemented to enhance
the rigour of the study and to promote the validity and
reliability of the results, there are inherent limitations
associated with quantifying risk-related information
using expert ratings. Second, the pair-wise interaction
database is limited to only 25 tasks (600 interactions).
The creation of a sufficiently representative and robust
database would require the quantification of many
more task interactions, including building construction
tasks. Thus, additional research in this area is
suggested. Third, these task interactions apply to the
construction environment at the time that the study
was conducted. Therefore, if common construction
tasks were to change due to the implementation of
technological innovations or new means and methods
the pair-wise interactions and base-level risk must be
re-evaluated. Fourth, the assumption made in this
research is that the tasks are performed as described in
Table 1 and that this performance is consistent
throughout the industry. Satisfying this assumption
requires competent and capable crews with sufficient
leadership and management control. The authors
recognize, however, that construction sites are
composed of a spectrum of crews with different levels
of safety experience, competencies and capabilities.
Therefore, the safety interaction risks presented here
are average for the industry and can be varied for differ-
ent projects and crews. Finally, this research does not
consider the impacts of environmental risk factors such
as weather or light conditions, the safety programme of
the contractor or productivity pressure from top
managers. Despite these limitations, the resulting data-
base makes a significant contribution to the body of
knowledge which can be used to enhance project
management capabilities through the integration of
safety with other project management functions.
References
Anumba, C. and Bishop, G. (1997) Importance of safety
considerations in site layout and organization.
Canadian
Journal of Civil Engineering,
24
(2), 229–36.
Bai, Y. (2002) Improving highway work zone safety, in
Proceedings, ASCE (Texas section),
Waco, Texas, 2–5
October, pp. 1–10.
Baradan, S. and Usmen, M.A. (2006) Comparative injury
and fatality risk analysis of building trades.
Journal of
Construction Engineering and Management,
132
(5), 533–9.
Bernold, L.E. and Guler, N. (1993) Analysis of back injuries
in construction.
Journal of Construction Engineering and
Management,
119
(3), 607–21.
428
Hallowell
et al
.
Best, R.J. (1974) An experiment in Delphi estimation in
marketing decision making.
Journal of Marketing Research,
11
(1), 448–52.
Boje, D.M. and Murnighan, J.K. (1982) Group confidence
pressures decisions.
Management Science,
28
(1), 1187–96.
Brockhoff, K. (1975) The performance of forecasting groups
in computer dialogue and face-to-face discussion, in The
Delphi Method: Techniques and Applications, Addison
Wesley, Reading, MA, pp. 291–321.
Bureau of Labor Statistics (2010) National census of fatal
occupational injuries in 2009, available at http://
www.bls.gov/news.release/pdf/cfoi (accessed 12
January 2010).
Chan, A.P.C., Yung, E.H.K., Lam, P.T.I., Tam, C.M. and
Cheung, S.O. (2001) Application of Delphi method in
selection of procurement systems for construction
projects. Construction Management and Economics, 19,
699–718.
Chantawit, D., Hadikusumo, B.H.W., Charoenngam, C.
and Rowlinson, S. (2005) 4DCAD-safety: visualizing
project scheduling and safety planning. Construction Inno-
vation, 5(2), 99–114.
Chi, C., Changa, T. and Ting, H. (2005) Accident
patterns and prevention measures for fatal occupational
falls in the construction industry. Applied Ergonomics, 36,
391–400.
Dalkey, N. (1972) An elementary cross-impact model. Tech-
nological Forecasting and Social Change, 3, 341–51.
Elbeltagi, E., Hegazy, A.H. and Eldosoky, A. (2004)
Dynamic layout of construction temporary facilities
considering safety. Journal of Construction Engineering and
Management, 130(4), 1186–95.
El-Rayes, K. and Khalafallah, A. (2005) Trade-off between
safety and cost in planning construction site layouts. Jour-
nal of Construction Engineering and Management, 131(11),
534–41.
Everett, J.G. and Frank, P.B. (1996) Costs of accidents and
injuries due to the construction industry. Journal of
Construction Engineering and Management, 122(2), 158–64.
Gunhan, S. and Arditi, D. (2005) Factors affecting interna-
tional construction. Journal of Construction Engineering and
Management, 131(3), 273–82.
Gupta, U.G. and Clarke, R.E. (1996) Theory and applica-
tions of the Delphi technique: a bibliography (1975–
1994). Technological Forecasting and Social Change, 53(2),
185–211.
Gürcanli, G.E. and Müngen, U. (2009) An occupational
safety risk analysis method at construction sites using
fuzzy sets. International Journal of Industrial Ergonomics,
39(2), 371–87.
Hallowell, M.R. and Gambatese, J.A. (2009) Activity-based
safety and health risk quantification for formwork
construction. ASCE Journal of Construction Engineering and
Management, 135(10), 990–8.
Hallowell, M.R. and Gambatese, J.A. (2010) Qualitative
research: application of the Delphi method to CEM
research. ASCE Journal of Construction Engineering and
Management, 136(1), 99–107.
Hinze, J. (1997) Construction Safety, Prentice Hall, Upper
Saddle River, NJ.
Hinze, J., Pedersen, C. and Fredley, J. (1998) Identifying
root causes of construction injuries. Journal of Construction
Engineering and Management, 124(1), 67–71.
Hinze, J., Huang, X. and Terry, L. (2005a) The nature of
struck-by accidents. Journal of Construction Engineering and
Management, 131(2), 262–8.
Hinze, J., Nelson, J. and Evans, R. (2005b) Software inte-
gration of safety in construction schedules. Paper
presented at the 4th Triennial International Conference,
Rethinking and Revitalizing Construction Safety, Health,
Environment and Quality, Port Elizabeth, South Africa,
16–17 May.
Hon, C.K.H., Chan, A.P.C. and Wong, F.K.W. (2010) An
analysis for the causes of accidents of repair, maintenance,
alteration and addition works in Hong Kong. Safety
Science, 48, 894–901.
Jannadi, O. and Almishari, S. (2003) Risk assessment in
construction. Journal of Construction Engineering and
Management, 129(5), 492–500.
Kartam, N. (1997) Integrating safety and health perfor-
mance into construction CPM. ASCE Journal of Construc-
tion Engineering and Management, 123(2), 121–6.
Lee, S. and Halpin, D. (2003) Predictive tool for estimating
accident risk. Journal of Construction Engineering and
Management, 129(4), 431–6.
Linstone, H. and Turoff, M. (1975) The Delphi Method: Tech-
niques and Applications, Addison Wesley, Reading, MA.
Lombardi, D.A., Verma, S.K., Brennana, M.J. and Perry, M.J.
(2009) Factors influencing worker use of personal protec-
tive eyewear. Accident Analysis and Prevention, 41, 755–62.
Lummus, R.R., Vokurkaz, R.J. and Duclos, L.K. (2005)
Delphi study on supply chain flexibility. International Jour-
nal of Production Research, 43(13), 2687–708.
Manoliadis, O.G., Pantouvakis, J. and Christodoulou, S.E.
(2009) Improving qualifications-based selection by use of
the fuzzy Delphi method. Construction Management and
Economics, 27, 373–84.
Mitropoulos, P. and Guillama, V. (2010) Analysis of resi-
dential framing accidents, activities, and task demands.
Journal of Construction Engineering and Management,
136(2), 260–9.
National Safety Council (2009) Accident Facts, IV, National
Safety Council, Itasca, IL.
Navon, R. and Kolton, O. (2006) Model for automated
monitoring of fall hazards in building construction.
Journal of Construction Engineering and Management,
132(7), 733–40.
Pandey, S. (2009) Risk quantification for highway recon-
struction activities, Masters thesis, University of Colorado,
Boulder, CO.
Rajendran, S. and Gambatese, J.A. (2009) Development and
initial validation of sustainable construction safety and
health rating system. Journal of Construction Engineering
and Management, 135(10), 1067–75.
Rogers, M.R. and Lopez, E.C. (2002) Identifying critical
cross-cultural school psychology competencies. Journal of
School Psychology, 40(2), 115–41.
Rowe, G. and Wright, G. (1999) The Delphi technique as a
forecasting tool: issues and analysis. International Journal
of Forecasting, 5, 353–75.
Safety risk management 429
Rozenfeld, O., Sacks, R., Rosenfeld, Y. and Baum, H. (2010)
Construction job safety analysis. Safety Science, 48, 491–8.
Sacks, R., Rozenfeld, O. and Rosenfeld, Y. (2009) Spatial
and temporal exposure to safety hazards in construction.
Journal of Construction Engineering and Management,
135(8), 726–36.
Saito, M. and Sinha, K. (1991) Delphi study on bridge
condition rating and effects of improvements. Journal of
Transportation Engineering, 117(3), 320–34.
Shapira, A. and Lyachin, B. (2009) Identification and analy-
sis of factors affecting safety on construction sites with
tower cranes. Journal of Construction Engineering and
Management, 135(1), 24–33.
Sobeih, T., Salem, O., Genaidy, A., Abdelhamid, T. and
Shell, R. (2009) Psychosocial factors and musculoskeletal
disorders in the construction industry. Journal of Construc-
tion Engineering and Management, 135(4), 267–77.
Yeung, J.F.Y., Chan, A.P.C., Chan, D.W.M. and Li, L.K.
(2007) Development of a partnering performance index
(PPI) for construction projects in Hong Kong: a Delphi
study. Construction Management and Economics, 25,
1219–37.
Yeung, J.F.Y., Chan, A.P.C. and Chan, D.W.M. (2008)
Establishing quantitative indicators for measuring the part-
nering performance of construction projects in Hong Kong.
Construction Management and Economics, 26, 277–301.
Yi, K. and Langford, D. (2006) Scheduling-based risk esti-
mation and safety planning for construction projects. Jour-
nal of Construction Engineering and Management, 132(6),
626–35.
Copyright of Construction Management & Economics is the property of Routledge and its content may not be
copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written
permission. However, users may print, download, or email articles for individual use.
Major Accident Factors for Effective Safety Management of Highway Projects
References
Kim, Y. A., Ryoo, B. Y., Kim, Y., & Huh, W. C. (2013). Major Accident Factors for Effective Safety Management of Highway Construction Projects. Journal Of Construction Engineering & Management, 139(6), 628-640. doi:10.1061/(ASCE)CO.1943-7862.0000640
Abstract:
Because of the variety of construction works and jobsite environments, research on highway construction safety is lacking. The purpose of this study is to identify accident factors by work type for highway construction projects through analysis of accidents occurring in Korea from 1997-2008. Analysis of variance and cross-tabulation analysis were used to identify statistically significant differences among nine factors and 248 variables in six work types. This analysis shows that contributing factors and variables to injuries vary depending on time of accident, work, and accident environment. The results outline trends of construction injuries by highway work. The results can be used to identify high-risk types of accidents, measure potential risks, improve safety training, and develop prevention plans for each work type under various work environments. The study shows that most accidents in highway construction fell into six construction work types. Because the relationships among contributing environment classification factors to accidents are identified, the factors can be considered to reduce the possibility of accidents by controlling closely related factors. This research can be used to develop an effective safety plan for highway construction projects. [ABSTRACT FROM AUTHOR]
Major Accident Factors
Major Accident Factors for Effective Safety Management
of Highway Construction Projects
Young Ai Kim1; Boong Yeol Ryoo2; Yong-Su Kim3; and Woon Chan Huh4
Abstract: Because of the variety of construction works and jobsite environments, research on highway construction safety is lacking. The
purpose of this study is to identify accident factors by work type for highway construction projects through analysis of accidents occurring in
Korea from 1997–2008. Analysis of variance and cross-tabulation analysis were used to identify statistically significant differences among
nine factors and 248 variables in six work types. This analysis shows that contributing factors and variables to injuries vary depending on time
of accident, work, and accident environment. The results outline trends of construction injuries by highway work. The results can be used to
identify high-risk types of accidents, measure potential risks, improve safety training, and develop prevention plans for each work type under
various work environments. The study shows that most accidents in highway construction fell into six construction work types. Because the
relationships among contributing environment classification factors to accidents are identified, the factors can be considered to reduce the
possibility of accidents by controlling closely related factors. This research can be used to develop an effective safety plan for highway
construction projects. DOI: 10.1061/(ASCE)CO.1943-7862.0000640. © 2013 American Society of Civil Engineers.
CE Database subject headings: Highways and roads; Construction sites; Safety; Accidents; Earthmoving; Installation; Drainage;
Tunnels; Construction management.
Author keywords: Highway construction site; Safety management; Accident factors and variables; Empirical analysis; Earth moving;
Installation; Paving; Drainage; Tunnel; Structure.
Introduction
Significant accidents involving construction workers frequently
occur because workers are exposed to various environmental haz-
ards and poor safety management [Korea Occupational Safety and
Health Agency (KOSHA) 2009; Lehtola and Van Der Molen
2008]. In Korea, 2,181 workers died in 2009 because of industrial
accidents, and 606 (27.7%) of them were from construction indus-
try accidents. Such accidents show the highest number of fatalities,
with an industrial accident rate of 18.9 per 100,000 people [Korea
Ministry of Employment and Labor (KMOEL) 2009]. In the United
States, the construction industry also accounts for the highest death
toll, with 975 persons (22.5%) of 4,340 total industrial accident
deaths and an industrial accident rate of 11.1 per 100,000 people
in 2009 [Lehtola and Van Der Molen 2008; Occupational Health
and Safety Administration (OSHA) 2009]. A study that has com-
pared such industrial accident victims of these two countries reports
that the Korean construction industry has over twice the U.S. con-
struction industry’s industrial accident rate (Ahn et al. 2004).
The safety management of worksite accidents in highway
construction is very important owing to the nature of the outdoor
work environment, which is affected by topography, geography,
and climate. The death toll attributable to highway construction
accidents compared with that attributable to overall construction
accidents in Korea is reported as 57=585 workers (9.47%) in 2006,
46=572 workers (8.04%) in 2005, and 42=612 workers (8.50%) in
2004 (KOSHA 2004, 2005, 2006). In the United States, highway
construction deaths compared with overall construction accident
deaths are reported as 154=1,297 workers (11.87%) in 2006,
160=1,243 workers (12.87%) in 2005, and 142=1,278 workers
(11.03%) in 2004 (U.S. Bureau of Labor Statistics 2004,
2005, 2006).
Literature Review
Studies have been carried out by those who are involved in safety
management to effectively reduce accidents in the construction in-
dustry. Lim et al. (2008) and Sin and Lee (2010) indicate that, even
though the safety-management organizations of owners, contrac-
tors, and architects are putting more effort and research into safety
management and accident prevention, accidents in construction are
still at a high level compared with those in other industries. Hinze
and Raboud (1988), in a study measuring and evaluating worker
safety and injury at large building-construction projects in Canada,
concluded that implementation of safety polices at the corporate
level influences safety performance at the project level. Meanwhile,
Son et al. (2002) report that worksite safety management is at a low
standard, involving mere observation of safety laws rather than im-
plementation according to careful analysis and business strategies
of construction companies, and that the business focus of chief
1Researcher, Texas A&M Univ., College Station, TX 77843-3137;
formerly, Ph.D. Candidate, Dept. of Architecture and Building Science,
Chung-Ang Univ., Seoul 156-756, South Korea. E-mail: happywomen21@
hanmail.net
2Assistant Professor, Dept. of Construction Science, Texas A&M Univ.,
3137 TAMU, Langford A433, College Station, TX 77843-3137
(corresponding author). E-mail: bryoo@tamu.edu
3Professor, Dept. of Architecture and Building Science, Chung-Ang
Univ., Seoul 156-756, South Korea. E-mail: yongsu@cau.ac.kr
4Graduate School of Construction Engineering, Chung-Ang Univ.,
Seoul 156-756, South Korea. E-mail: huchan@ex.co.kr
Note. This manuscript was submitted on June 22, 2011; approved on
September 12, 2012; published online on September 14, 2012. Discussion
period open until November 1, 2013; separate discussions must besub-
mitted for individual papers. This paper is part of the Journal of Con-
struction Engineering and Management, Vol. 139, No. 6, June 1, 2013.
© ASCE, ISSN 0733-9364/2013/6-628-640/$25.00.
628 / JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / JUNE 2013
http://dx.doi.org/10.1061/(ASCE)CO.1943-7862.0000640
http://dx.doi.org/10.1061/(ASCE)CO.1943-7862.0000640
http://dx.doi.org/10.1061/(ASCE)CO.1943-7862.0000640
http://dx.doi.org/10.1061/(ASCE)CO.1943-7862.0000640
executive officers (CEOs) on worker safety is nearly absent and
thus has no influence on worksite safety management. Lim et al.
(2008) point out that safety-management tasks in construction
cannot be handled by one agent and divide the roles of owners into
preconstruction, construction, and postconstruction stages, empha-
sizing the role of owners, with architects and contractors as the
key players in establishing a comprehensive safety-management
system.
Hinze (1992), through surveys of U.S. companies (namely,
constructors, design firms, and design-build firms) dealing with
construction worker safety at the design stage, has found that only
a few architectural-design firms consider construction worker
safety within the scope of their responsibilities. To reduce acci-
dents, Hinze has proposed consideration of safety management
from the beginning stage of design. Furthermore, Hong (2004) pro-
poses a construction safety information model in the design stage.
Fang et al. (2004a), through surveys on correlations in safety-
management performances at construction projects in China,
present organization, economic, and daily task-management factors
as those that have close relationships with safety-management
performance in construction, and they have performed safety-
management evaluations on the projects with these three factors.
Currently, research on safety management in highway projects
has been limited to identification of accident types for improving
safety-management plans for construction.
The construction report of the Korean Expressway Corporation
(2008) categorizes construction accidents as death or injury and
classifies them into six to eleven accident types depending on six
construction work types. In its construction report, the New York
State Department of Transportation (NYSDOT) categorizes con-
struction accidents into fatal, hospital-level, minor, and none/
unknown according to the severity of injury, and classifies the ac-
cidents into eight accident types (NYSDOT 2002). Mohan and
Zech (2005), in their analysis of 1990–2001 construction accident
data as presented in this 2002 NYSDOT report, classify the acci-
dents into five construction accident types and five traffic accident
types involving construction workers. The OSHA classification
system categorizes accident types into five kinds: falls, electrical
shock, struck by, caught between, and other (OSHA 2009). Bryden
(1999), using such OSHA classification, and Huang and Hinze
(2003), using OSHA’s Integrated Management Information System
(IMIS), have proposed improvement plans for managing accident
causes by injury type to prevent injuries caused by safety accidents
in highway construction projects. Table 1 presents similar studies
on safety management in construction.
In addition, previous studies have focused on how construction
safety laws and regulations have been followed by owners or con-
tractors from the design phase to the construction phase. Also, even
though highway construction accidents or accidents according to
roadwork types occur because of many factors, including construc-
tion environment, problems of insufficient research have been
revealed as follows:
• The majority of the previous studies are focused on the con-
struction industry as a whole, making research on accidents
in highway construction insufficient; and
• The research on relations among practical worksite-related
accident factors is lacking—the question of whether there are
correlations among accident cause factors is overlooked because
previous studies focus only on management responsibilities
while ignoring accident environment and factors through acci-
dent data.
It is judged that the implementation policy for each management
responsibility will be clearer when assessment on correlations
among accident factors is performed.
In pursuit of such study, Huh et al. (2010) have analyzed acci-
dents limited to structure construction, presenting time of accident,
accident cause, and accident type as factors with number of acci-
dent victims and meaningful difference (p < 0.05) in structure, and
accident cause and accident height as factors with relationships
with accident type. Huh et al. (2010) also point out the necessity
for expanded research on whether the accident environment factors
can be equally applied in other types of highway construction work
and which other factors appear by each work type. Because the
work types of highway construction work are diverse, the effective
safety management of a work site is demanded in each work type
for the same or similar kinds of accidents.
Thus, this study intends to investigate correlations among the
accident factors of accidents occurring at highway construction
projects in Korea from 1997–2008 by additional analysis. The re-
sults of this study can be applied in effective hazard recognition and
management of critical factors for accident prevention and safety
precautions in each work type.
Research Objective
The purpose of this study is to present major accident factors and
variables by six work types to be applied foremost in the effective
design of safety precautions and management of highway construc-
tion projects. In detail, it is to identify statistically significant differ-
ences among nine factors and 248 variables in six work types
through carrying out the following tasks:
• Identify factors and variables among time of accident, work en-
vironment, and accident environment factors for each highway
construction work type, showing number of accident victims
and significant results;
• Analyze relationships between accident type and environment
(e.g., cause, location, height) for six work types; and
• Identify significant factors and variables among accident envir-
onment factors according to time, and work environment factors
by each highway construction type.
By synthesizing research results as described previously, acci-
dent factors and variables related with accident occurrence in
appurtenant, paving, drainage, earth, tunnel, and structure works
of highway construction sites will be presented. The tasks can be
practically utilized in accident prevention to effectively reduce
highway construction accidents.
Research Scope
Although this study is limited to highway construction in Korea,
the accident study may produce results with wider applicability.
This study will have five different analyses of six types of work
commonly performed in highway construction. Fig. 1 maps out
the five analyses that were used in this study and demonstrates their
applicability. The scope and contents to be assessed are summa-
rized in Table 2.
Injury Ratio Assessment and Work Classification
Since 1992, KOSHA (2007) has developed an injury-ratio assess-
ment (IRA), which documents each accident occurrence with the
goal of preventing future accidents. An IRA is a statistical indicator
showing the frequency of industrial accidents and accident inten-
sity. The unit of measurement is a convertible number of accident
victims. When an accident is fatal, it is given 10 times the weight of
an injury; this, added to the number of injuries in nonfatal acci-
dents, is then compared with the number of workers at a worksite
for the year. The following formulas are used to calculate the IRA
(KOSHA 2007):
JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / JUNE 2013 / 629
Convertible number of accident victims
¼ fatality × 10 þ number injured
Convertible accident rate ðpercentÞ ¼ convertible accident
− ðnumber of victims=permanent number of workersÞ × 100
The Korea Expressway Corporation’s (KEC) Highway Con-
struction Guide Specification: Civil Engineering divides the entire
road-construction process into 13 work types: surveying and soil
survey, soil improvement, earth moving, erosion control, drainage,
bridge, tunnel, antifrost layer subbase course and subbase, asphalt
concrete paving, cement concrete paving, safety facilities, environ-
ment management and installation, and material (KEC 2009).
Highway construction work in the United States is broken down
into the following categories: clearing; earth moving; base con-
struction; paving; installations (e.g., utilities); drainage; services
(e.g., surveying); and finishing (e.g., sign installation or lane mark-
ing) (Hassanein and Moselhi 2004). Highway tunnel construction
is separately classified.
Data Collection
The data are from the KEC’s lists of accidents occurring during
highway construction in South Korea from 1997–2008 (KEC
2008). The accidents are classified into seven work types: drainage;
installations; tunnel; earth moving; paving; structure; and miscella-
neous (site preparation and clearing, aggregate hauling, receiving
transmitted electricity, material delivery, painting, and traffic con-
trol). In this study, miscellaneous work types are excluded because
Table 1. Construction Safety-Management Research with Findings
Author(s) Research aims Research scope Finding factors
Li (2010) To investigate perceptions of
experts on highway work zone
safety—a framework for
developing guidelines for
implementing work zone safety
Work zone safety, problems and effective
countermeasures, types of projects suitable
for auditing, audit frequencies, audit team
composition and funding sources, audit
tasks and tools, project staff safety-training
audits
Cause of work zone safety problems and effective
countermeasures, types of projects suitable for
auditing and audit frequencies, audit team
composition and funding sources, audit tasks and
tools, project staff safety-training audits
Arditi et al.
(2007)
To investigate accident
characteristics of nighttime and
daytime highway construction
activities
Lighting and weather conditions Condition of vehicle operator, accidents caused by
through-traffic and construction equipment operating
inside the work area
Teo et al.
(2005)
To provide framework for
project managers to manage
construction site safety
Policy, process, personnel, incentive Inadequate company policies, unsafe practices, poor
attitudes of construction personnel, poor management
commitment, and insufficient safety knowledge and
training of workers
Mohan and
Zech (2005)
To provide cost-effective safety
measures to protect construction
workers in highway work zones
Construction work–area accidents, traffic
accidents involving construction workers
Construction work–area accidents compose nearly
91% of total costs: struck/pinned by large equipment,
trip or fall (elevated), contact with electrical or gas
utility, struck by moving/falling load, and crane/lift
Arditi et al.
(2005)
To investigate highway
construction maintenance
professionals’ perceptions
Nighttime construction conditions on
worker visibility and issues associated with
safety vests in nighttime activities
Important roles in accidents: condition of vehicle
operator, construction equipment operating inside
work area, and poor worker visibility
Gambatese
et al. (2005)
To include safety of
construction workers in
construction-project design
Change in designer mindset toward safety,
increase in designer knowledge of concept,
incorporation of construction safety
knowledge in design phase, mitigation of
designer liability exposure
Establishment of motivational force to promote
designing for safety, increase in designer knowledge
of concept, incorporation of construction safety
knowledge in design phase, mitigation of designer
liability exposure
Fanga et al.
(2004b)
To perform factor analysis–
based studies on construction
workplace safety management
Foreman-related factor, worker-related
factor, crew-related factor, manager-related
factor, and safety training–related factor
Five major factors (foreman-related, worker-related,
crew-related, manager-related, safety training–related
factors) and 31 variables
Huang and
Hinze (2003)
To identify additional
information that might be
helpful in reducing future
incidences of construction
worker falls
Fall accidents, accident location, seasonal
construction projects
Fall accidents taking place at heights of less than
9.15 m (30 ft), misjudged by worker and occurring
primarily on new commercial building–construction
projects and relatively low–construction cost
residential projects
Toole (2002) To clarify roles of design and
construction professionals in
site safety
Design engineers, general contractors,
subcontractors
Ability to influence root causes: high for
subcontractors, moderate for general contractors,
mixed for architect/engineers, and low for owners
Sawacha
et al. (1999)
To identify factors influencing
safety on construction sites
Historical, economical, psychological,
technical, procedural, organizational,
environmental
Group of factors influencing safety construction:
organization policy; top five influencing site safety:
management talk on safety, providing safety booklets,
providing safety equipment, providing safety
environment, appointing trained on-site safety
representative
Dedobbeleer
and Beland
(1991)
To provide safety climate model
on construction workers
Management concerns, management
safety activities, employee risk perception,
management commitment, worker
involvement
Two factors: management’ commitment to safety and
workers’ involvement in safety
630 / JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / JUNE 2013
they are fragmented into many tasks and the significance of work
type analysis cannot be expressed. For this study, the relationships
between work environment and accident environment of construc-
tion zones are analyzed using the other six work types (Table 3).
Because of its geographical location, Korea experiences all four
seasons. The 30-year average temperature from November to
February ranges from −11.5 to 16.3°C, and the average precipita-
tion of the period ranges between 14.2 and 80.3 mm. In summer,
the 30-year average temperature from May to August ranges from
−0.3 to 30.7°C, and the average precipitation of the period in-
creases from 75.6 to 364.0 mm. The 30-year average temperature
from March to October ranges from −2.8 to 27.4°C, and the
30-year average precipitation of the period increases from 38 to
243.8 mm (Korea Meteorological Administration 2011). The aver-
age working hours of full-time construction workers range from
158.6–190.8 h per month, and the number of working days per
month is between 19.7 and 23.8 days (KMOEL 2011).
Research Methods
In this study, construction accidents are empirically analyzed to
find significance in number of accident victims related to time
(hour, day, or month), work environment (temperature, weather),
Fig. 1. Research model
Table 2. Scope of Research
Breakdown Research contents
Work classification Structure, installations, paving, drainage, earth moving, and tunnel
Analysis scope Analysis 1–5
Significance level (α < 0.1): broaden scope because construction accidents are mostly serious accidents
Accident factor classification: time, work environment, and accident environment by six work types
Analysis 4 and 5
Significant difference analysis of time/work environment and accident environment (if time/work environment factors
show significant result in accident environment, research design on causal relationships among factors can be provided)
Analysis factors Nine factors: e.g., month, day, hour, weather, temperature, accident cause, location, type, height (year and work characteristic
are excluded after analysis)
Analysis goal To empirically analyze major accident factors and variables in each work type that can be applied to effective safety
management of accidents that occur in structure- and other highway construction work types
Table 3. Highway Construction Accidents by Work Type from 1997–2008
Work type
Number of
occurrences
Convertible
number of
accident victims
Number
of fatalities
Number
of injured
Drainage 50 88.0 4 48
Paving 16 24.0 1 14
Installations 39 67.0 3 37
Earth moving 52 207.0 17 37
Tunnel 70 300.0 22 80
Structure 368 1,188.0 86 328
Total 595 1,874 133 544
Note: In this paper, number of accident victims indicates convertible
number of accident victims.
JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / JUNE 2013 / 631
and accident environment (accident type, elevation, cause, and
location), and meaningful differences among the nine factors.
The main analysis methods used in this research are as follows:
• Analysis of variance (ANOVA), significance probability (p),
and significance level (α < 0.1) to analyze statistically mea-
ningful differences in number of accident victims and time
(Analysis 1), work environment (Analysis 2), and accident
environment (Analysis 3); and
• Analysis of cross-tabulation, chi-square test, significance level
(α < 0.1), and p-value (p) for difference analyses of accident
type and accident environment (Analysis 4), and accident envir-
onment according to time/work environment (Analysis 5).
Research Model
Fig. 1 shows the research model of empirical analysis of factors of
highway construction accidents by work type. If analyses of the
relations between victims and time (Analysis 1), work environment
(Analysis 2), and accident environment and work type (Analysis 3)
show a statistically meaningful difference, the accident factors can
be discerned. Factors shown by work types are control factors that
have to be responded to in safety precautions and management.
Analysis 4 is a difference analysis determining whether any rela-
tionship between accident environment and type factors exists in
the work types. Analysis 5, the difference analysis of time/work
environment and accident environment, can be viewed as being
related to accident environment. If relationships are found, quali-
tative research designed to investigate cause and result can be
proposed, and effective worksite safety-management items can be
determined.
Six Work Classification Measurement Factors and
Variables
Nine factors and 248 variables of accidents that occur at highway
construction zones have been selected. These were divided into two
groups: time/work environment, with five factors and 54 variables
such as month (12), day (7), hour (24), weather (6), and temper-
ature (5); and accident environment, with four factors and 194 var-
iables such as accident type (47), accident location (92), accident
cause (36), and accident height (19). The significance level of
α < 0.1 has been used because of the severity of the accidents
measured.
Empirical Analysis
Variance analysis has been carried out for the difference analysis of
number of accidents by accident factors. Time of accident (hour,
day, month), work environment (weather, temperature), and acci-
dent environment (accident type, accident cause, accident location,
accident height) factors have been analyzed for statistical difference
by six work types with a confidence level of p < 0.1.
Difference Analysis of Time of Accident and Number of
Accident Victims
Table 4 summarizes the result of magnitudes of specific month, day,
and hour factors and variables from the difference analysis of num-
ber of accident victims according to time by six work types. Work
types that show the differences in number of accident victims in
month factors are installations, earth moving, and structure. Instal-
lations have difference in number of accident victims according to
month at the confidence level of 0.019 < 0.05 and show the mag-
nitude order of October, December, June, and so on. In day factors,
earth moving, tunnel, and structure have statistically meaningful
differences. Tunnel displays difference in number of accident vic-
tims according to day at the confidence level of 0.035 < 0.05, with
the magnitude order of Wednesday, Thursday, Saturday, and so on.
In hour factors, meaningful difference is shown in drainage, tunnel,
and structure. Because such results can be interpreted as showing
that time has a relationship with accident-victim occurrence, there
should be safety precautions and management according to the fac-
tors and variables in each work type that shows meaningful results.
Difference Analysis of Work Environment and Number
of Accident Victims
Determined from analysis regarding whether numbers of accident
victims show statistically meaningful differences in accidents ac-
cording to work environment by six work types, the magnitudes
of weather variables are shown in Table 5. Structure has a differ-
ence in number of accident victims according to weather, with the
Table 4. Difference Analysis of Accident Time and Number of Accident Victims
Work time Drainage Tunnel Installations Earth moving Paving Structure
Month F ¼ 1.012 F ¼ 1.294 F ¼ 2.643 F ¼ 2.450 F ¼ 0.958 F ¼ 1.743
p ¼ 0.454 p ¼ 0.255 p ¼ 0.019a p ¼ 0.019a p ¼ 0.542 p ¼ 0.063b
— — October > December >
June > September >
March > August > May
February > December >
June > April > March >
July > November
— May > February >
September > June >
April > August
Day F ¼ 0.875 F ¼ 2.431 F ¼ 1.409 F ¼ 2.196 F ¼ 0.438 F ¼ 1.972
p ¼ 0.521 p ¼ 0.035a p ¼ 0.242 p ¼ 0.061b p ¼ 0.813 p ¼ 0.069a
— Wednesday >
Thursday > Saturday >
Sunday > Tuesday >
Monday > Friday
— Friday > Tuesday >
Thursday > Saturday >
Sunday > Monday>
Wednesday
— Thursday > Sunday >
Monday > Wednesday >
Saturday > Friday >
Tuesday
Hour F ¼ 8.560 F ¼ 4.979 F ¼ 0.683 F ¼ 0.955 F ¼ 0.725 F ¼ 1.778
p ¼ 0.000c p ¼ 0.000c p ¼ 0.743 p ¼ 0.501 p ¼ 0.671 p ¼ 0.036a
7 > 15 > 5 2 > 16 > 5 > 14 >
8 > 17 > 15
— — — 15 > 16 > 12 > 11 >
18 > 17
Note: Dashes indicate work type and factors that do not show statistically meaningful results.
ap < 0.05.
bp < 0.1.
cp < 0.01.
632 / JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / JUNE 2013
confidence level of 0.062 < 0.1, and appears in the magnitude order
of clear, mostly cloudy, and so on. Temperature factor is shown not
to have any difference in number of accident victims in all work
types. Such results indicate that work performance among all work
types is affected by weather and temperature owing to highway
construction’s taking place outdoors. Therefore, although such fac-
tors do not give meaningful difference in accident occurrence, it is
still necessary to manage them as safety factors considering their
magnitudes of variables. This is because weather and temperature
influence workers psychologically and physically (Sin 2010).
Difference Analysis of Accident Environment and
Number of Accident Victims
According to the difference analysis in Table 6, earth moving, pav-
ing, and structure are shown to have statistically meaningful results
according to accident cause factors and number of accident victims.
Earth moving has a difference in number of accident victims ac-
cording to cause at the confidence level of 0.070 < 0.1, with ma-
chine defect being the greatest cause. Accident type factors show
meaningful difference in drainage, installation, and structure.
Drainage has a relationship with accident type and number of
accident victims at the confidence level of 0.000 < 0.01, with
breakdown/collapse as its greatest accident type. Accident location
factors show meaningful differences in drainage, and accident
height factors in show meaningful differences in earth moving
and structure. Table 6 shows the specific orders of magnitude of
factors and variables of the six work types.
Difference Analysis of Accident Type Factor according
to Accident Environment
Cross-tabulation and chi-square tests have been performed for
difference analysis of accident type factor according to accident
Table 5. Difference Analysis of Work Environment and Number of Accident Victims
Working environment Drainage Tunnel Installations Earth moving Paving Structure
Weather F ¼ 0.197 F ¼ 0.352 F ¼ 0.542 F ¼ 0.525 F ¼ 1.850 F ¼ 2.124
p ¼ 0.939 p ¼ 0.842 p ¼ 0.706 p ¼ 0.718 p ¼ 0.192 p ¼ 0.062a
— — — — — Clear > mostly cloudy > rain > cloudy >
snow > partly cloudy
Temperature — — — — — —
Note: Dashes indicate work type and factors that do not show statistically meaningful results.
ap < 0.1.
Table 6. Difference Analysis of Accident Environment and Number of Accident Victims
Accident
environment Drainage Tunnel Installations Earth moving Paving Structure
Accident
cause
F ¼ 0.569 F ¼ 1.385 F ¼ 1.336 F ¼ 2.205 F ¼ 3.555 F ¼ 3.032
p ¼ 0.723 p ¼ 0.242 p ¼ 0.274 p ¼ 0.070a p ¼ 0.059a p ¼ 0.003b
— — — Machine defect > worker
carelessness > faulty
work method > faulty
safety facilities > faulty
safety equipment >
faulty scaffolding
Faulty safety
equipment > worker
carelessness > faulty
work method
Defective design > faulty
machine > faulty safety
facilities > faulty safety
equipment > faulty work
method > worker
carelessness > physical
defect > faulty scaffolding
Accident
type
F ¼ 19.923 F ¼ 0.956 F ¼ 4.275 F ¼ 0.801 F ¼ 0.762 F ¼ 3.818
p ¼ 0.000b p ¼ 0.462 p ¼ 0.002b p ¼ 0.574 p ¼ 0.487 p ¼ 0.000b
Breakdown/collapse >
electric shock >
construction
equipment > fall >
overturning >
disease > accidental
injuries
— Falling/flying >
breakdown/collapse >
fall > construction
equipment >
overturning
— — Breakdown/collapse > fall >
electric shock > disease >
struck by > construction
equipment, machine
defect > overturning
Accident
location
F ¼ 2.033 F ¼ 0.172 F ¼ 0.495 F ¼ 1.371 F ¼ 10.355 F ¼ 1.158
p ¼ 0.056a p ¼ 0.688 p ¼ 0.933 p ¼ 0.212 p ¼ 0.487 p ¼ 0.262
Drainage channel >
lateral drain >
culvert > channel
box > V-type
gutter > mountainside
ditch, temporary
— — — — —
Accident
height
F ¼ 0.806 F ¼ 1.011 F ¼ 0.079 F ¼ 5.808 F ¼ 0.477 F ¼ 26.960
p ¼ 0.374 p ¼ 0.318 p ¼ 0.924 p ¼ 0.002b p ¼ 0.501 p ¼ 0.000b
— — — Over 10 m > surface
level > 1–5 m > 5–10 m
— Over 10 m > 5–10 m >
surface level > 1–5 m
Note: Dashes indicate work type and factors that do not show statistically meaningful results.
ap < 0.1.
bp < 0.01.
JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / JUNE 2013 / 633
environment by six work types. As described subsequently, differ-
ences of accident cause and accident type (Table 7), accident
location and accident type (Table 8), and accident height and
accident type (Table 9) have inferred major factors and variables
by six work types at the confidence levels of p < 0.1, α < 0.1.
Difference Analysis of Accident Cause and Accident Type
In accident type according to accident cause by six work types,
statistically meaningful difference is shown in installation, drain-
age, earth moving, and structure. The accident type variables are
shown differently by work type according to accident cause vari-
ables, although worker carelessness is the majority in all accident
cause variables of the four work types. In structure, fall is shown as
the greatest among all accident type variables according to the mag-
nitude of accident cause variables. This is shown in Table 7.
Difference Analysis of Accident Location and Accident Type
Accident type according to accident location by work type
shows statistically meaningful results only in drainage, with the
highest accident location variable being culvert, and with the acci-
dent cause variables being fall. This is shown in Table 8.
Difference Analysis of Accident Height and Accident Type
Difference in accident height and accident type by work type is
shown in all work types, with surface level being highest among
accident height variables, and with various accident type variables.
This means that the factors of accident height and accident type
in all work types have a relationship to number of accident
victims. The orders of magnitude of accident type variables and
accident height variables according to work type are shown in
Table 9.
Difference Analysis of Accident Environment
according to Time of Accident and Work Environment
To analyze whether there is statistically meaningful difference in
accident environment factors of accident type, location, cause,
and height according to month (Table 10), day (Table 11), hour
(Table 12), weather (Table 13), and temperature (Table 14) factors,
major factors and variables have been inferred through cross-
tabulation and chi-square tests at the confidence levels of
p < 0.1, α < 0.1.
Table 7. Difference Analysis of Accident Cause and Accident Type
Work type χ2, p Accident cause Accident type
Installations χ2 ¼ 77.749,
p ¼ 0.000a
Worker carelessness Fall, struck by > construction equipment > accidental injuries > overturning
Faulty safety equipment Fall
Faulty work method Struck by > construction equipment > fall, falling/flying, breakdown/collapse
Faulty safety facilities
Physical defect
Fall, breakdown/collapse Disease
Faulty temporary Overturning
Drainage χ2 ¼ 77.749,
p ¼ 0.000a
Worker carelessness Fall > construction equipment, falling/flying, struck by > overturning > electric shock
Faulty work method Overturning, falling/flying > fall, construction equipment, breakdown/collapse
Faulty safety equipment Falling/flying > fall
Faulty safety facilities Overturning > fall, falling/flying
Faulty temporary Overturning
Physical defect Disease
Earth moving χ2 ¼ 52.433,
p ¼ 0.007a
Worker carelessness Construction equipment > fall > struck by > overturning > accidental injuries
Faulty work method Construction equipment > struck by > breakdown/collapse > fall, falling/flying
Machine defect Struck by
Faulty safety facilities Fall, struck by
Faulty temporary Falling/flying, breakdown/collapse
Faulty safety equipment Falling/flying
Structure χ2 ¼ 465.03,
p ¼ 0.000a
Worker carelessness Fall > overturning > falling/flying > electric shock > traffic accident
Faulty safety equipment Fall > falling/flying
Faulty work method Fall > struck by > falling/flying, overturning > construction equipment
Faulty safety facilities Fall > overturning > struck by, traffic accident > breakdown/collapse, falling/flying
Machine defect Fall > construction equipment > falling/flying > struck by
Faulty temporary Fall, struck by
Defective design Breakdown/collapse, struck by
Note: Only work types, factors, and variables with significant results are shown.
ap < 0.01.
Table 8. Difference Analysis of Accident Location and Accident Type
Work type χ2, p Accident location Accident type
Drainage χ2 ¼ 120.805,
p ¼ 0.002a
Culvert Fall > overturning > falling/flying > accidental injuries > construction equipment, struck by
Lateral drain Struck by > accidental injuries > breakdown/collapse
V-type gutter Construction equipment
Channel box Struck by > construction equipment
Mountainside ditch Overturning, accidental injuries
Temporary Falling/flying, disease
Slab Falling/flying
U-type gutter Falling/flying
Note: Only work types, factors, and variables with significant results are shown.
ap < 0.01.
634 / JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / JUNE 2013
Difference Analysis of Accident Environment according to
Month Factor
Differences among three factors are shown from analysis on
whether accident type, location, cause, and height have meaningful
difference according to month factor by work types (Table 10). In
accident type, installations and earth moving have to deal with the
month factor; installations have difference in month and accident
type at the confidence level of 0.055 < 0.1, with June as the greatest
Table 9. Difference Analysis of Accident Height and Accident Type
Work types Accident height Accident type χ2, p
Installations Surface level Struck by > construction equipment > accidental injuries > overturning >
breakdown/collapse > disease
χ2 ¼ 315.03, p ¼ 0.000a
1–5 m Fall > construction equipment
5–10 m Fall
Tunnel Surface level Falling/flying > construction equipment > breakdown/collapse >
struck by > overturning
χ2 ¼ 65.747, p ¼ 0.000a
1–5 m Fall > overturning
Drainage Surface level Falling/flying > overturning > construction equipment > struck by χ2 ¼ 38.095, p ¼ 0.000a
Earth moving Surface level Construction equipment > struck by > falling/flying > fall, overturning χ2 ¼ 34.151, p ¼ 0.012b
1–5 m Fall > breakdown/collapse
5–10 m Fall
Over 10 m Fall
Paving Surface level Struck by > overturning > construction equipment > traffic accident >
falling/flying
χ2 ¼ 9.143, p ¼ 0.013b
1–5 m Fall > overturning
Structure Surface level Struck by > falling/flying > overturning > construction equipment >
electric shock > disease > breakdown/collapse
χ2 ¼ 315.03, p ¼ 0.000a
5–10 m Fall > overturning > falling/flying, struck by
1–5 m Fall > overturning > falling/flying
Over 10 m Fall > falling/flying
ap < 0.01.
bp < 0.05.
Table 10. Difference Analysis of Month Factor and Accident Environment
Work type Month Accident environment χ2, p
Accident type
Installations June Struck by > construction equipment > fall χ2 ¼ 99.19,
p ¼ 0.055aMarch Struck by > fall
September Overturning > struck by
August Fall > struck by
May Fall > construction equipment
October Construction equipment, falling/flying
Earth moving May Construction equipment > fall χ2 ¼ 93.634,
p ¼ 0.014bApril Construction equipment, fall > struck by
August Construction equipment > fall, overturning, struck by
October Fall, breakdown/collapse, struck by
Accident cause
Earth moving May Worker carelessness > faulty work method > faulty temporary χ2 ¼ 69.13,
p ¼ 0.095aApril Worker carelessness > faulty work method > faulty safety facilities
August Faulty work method > worker carelessness > faulty safety facilities
September,
October
Worker carelessness > faulty work method
Structure June Worker carelessness > faulty safety equipment > faulty work method >
faulty safety facilities, machine defect
χ2 ¼ 108.38,
p ¼ 0.069a
October Worker carelessness > faulty work method > faulty safety equipment > faulty safety facilities
May Worker carelessness > faulty work method > machine defect > faulty safety equipment
September Worker carelessness > faulty work method > faulty safety equipment > faulty safety facilities
July Worker carelessness > faulty work method > faulty safety equipment
April Faulty work method > worker carelessness > faulty safety equipment
Accident height
Earth moving May Surface level > 1–5 m χ2 ¼ 44.97,
p ¼ 0.080aApril Surface level > 1–5, 5–10 m
August, October >
September >
July > June,
March, November
Surface level, 1–5 m
Note: Only work types, factors, and variables with significant results are shown.
ap < 0.1.
bp < 0.05.
JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / JUNE 2013 / 635
month variable and stricture as the accident type. Accident cause,
earth moving and structure, and accident height, earth moving,
show difference. Accident location factor does not show any mean-
ingful result in all work types.
Difference Analysis of Accident Environment according to
Day Factor
From day analysis of whether there is a difference in accident
type, location, cause, and height by work type, a meaningful
result has been obtained only from accident type. That is,
from day and accident type, significant result is shown in struc-
ture, whereas no meaningful result is shown for accident part,
cause, and location factors in all work types. Among days,
Thursday is shown to be highest, and among accident types,
fall is shown to be highest in all days. This shows that the
variables of accident type and day should be controlled in struc-
ture (Table 11).
Table 11. Difference Analysis of Day Factor and Accident Environment
Work type Day Accident type χ2, p
Structure Thursday Fall > falling/flying > struck by > overturning χ2 ¼ 71.583, p ¼ 0.055, p < 0.1
Sunday Fall > struck by > falling/flying > construction equipment > electric shock
Wednesday Fall > struck by > overturning > falling/flying > construction equipment
Monday Fall > construction equipment > falling/flying > overturning
Friday Fall > struck by > overturning > breakdown/collapse > falling/flying, electric shock,
construction equipment
Saturday Fall > struck by > overturning > falling/flying, electric shock, construction equipment
Tuesday Fall > struck by > accidental injuries, overturning > falling/flying > breakdown/collapse,
construction equipment
Note: Accident part, accident cause, and accident location factors do not have meaningful result in all six work types. Only work types, factors, and variables
with significant results are shown.
Table 12. Difference Analysis of Hour Factor and Accident Environment
Work type Hour (h) Accident environment χ2, p
Accident type
Structure 1500–1800 Fall > accidental injuries > struck by > overturning > falling/flying > disease χ2 ¼ 69.712,
p ¼ 0.011a0900–1200 Fall > struck by > falling/flying > overturning > accidental injuries > construction equipment >
breakdown/collapse
1200–1500 Fall > struck by > accidental injuries > overturning > falling/flying > construction equipment
0600–0900 Fall > struck by > falling/flying > electric shock > accidental injuries
0000–0600 Fall > falling/flying
1800–2400 Fall
Accident height
Structure 1500–1800 Surface level > 5–10 m > 1–5 m > over 10 m χ2 ¼ 27.915,
p ¼ 0.022a0900–1200 5–10 m > surface level > 1–5 m > over 10 m
1200–1500 Surface level > 5–10 m > 1–5 m
0600–0900 Surface level > 5–10 m > 1–5 m > over 10 m
0000–0600 5–10 m, surface level > 1–5 m > over 10 m
1800–2400 5–10 m > over 10 m
Note: Accident part and accident cause factors do not have significant result in all six work types. Only work types, factors, and variables with significant
results are shown.
ap < 0.05.
Table 13. Difference Analysis of Weather Factor and Accident Environment
Work type Weather Accident environment χ2, p
Accident cause
Tunnel Clear Worker carelessness > faulty work method > faulty safety facilities χ2 ¼ 49.26, p ¼ 0.000a
Rain Faulty work method > worker carelessness > faulty safety facilities,
faulty safety equipment
Mostly cloudy Faulty work method > worker carelessness > faulty safety facilities
Partly cloudy Worker carelessness > faulty safety equipment > machine defect
Snow Faulty work method
Accident height
Earth moving Clear Surface level > 1–5 m > over 10 m χ2 ¼ 34.543, p ¼ 0.001a
Mostly cloudy Surface level
Rain Surface level > over 10 m
Partly cloudy Surface level > 1–5 m
Snow Surface level > 5–10 m
Note: Only work types, factors, and variables with significant results are shown.
ap < 0.01.
636 / JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / JUNE 2013
Difference Analysis of Accident Environment according to
Hour Factor
From analysis on whether there is meaningful difference in accident
type, location, cause, and height according to hour factor by work
type, only structure shows a difference in accident type and acci-
dent height. Conversely, accident location and accident cause ac-
cording to hour show no significant result in all work types. This
does show that, in structure, hour variables have to be dealt with in
accident type and accident height (Table 12).
Difference Analysis of Accident Environment according to
Weather Factor
From analysis on whether there is a difference in accident type,
location, cause, and height according to weather factor by work
type, weather and accident cause factors in tunnel and weather
and accident height factors in earth moving show a significant
difference. Accident type and accident location do not have a sig-
nificant result with weather in all work types. Meanwhile, in tunnel,
accident cause shows a relationship with weather at the confidence
level of 0.000 < 0.01, and it appears in the order of magnitude of
worker carelessness, faulty work method, accidental injuries, in
clear weather. In earth moving, accident location shows a relation-
ship with weather (Table 13).
Difference Analysis of Accident Environment according to
Temperature Factor
Differences between two factors by work type are shown from
analysis on whether accident type, location, cause, and height have
significant difference according to temperature factor by work
type. In temperature and accident height, difference is shown in
drainage, earth moving, and paving. In temperature and accident
location, it is shown in tunnel. Accident type and accident cause
have no meaningful result with temperature in all work types.
Meanwhile, tunnel, earth moving, and paving show relationship
to accident height according to temperature, with tunnel showing
that temperature and accident location have a correlation. These
factors and variables of work type, showing a significant result,
should be monitored closely among jobsite accident factors
(Table 14).
Results and Discussion
For the difference analysis of time, work environment, and accident
environment, and number of accident victims by work type, vari-
ance analysis (F, p-value, α < 0.1) has been performed and
presents the magnitudes of factors and variables with meaningful
results according to work type (Tables 4–6), for the difference
analysis of accident environment factors and accident type
(Tables 7–9), and for the difference analysis of accident environ-
ment factors according to time/work environment factors. Cross-
tabulation [χ2, p-value, (α < 0.1)] has been performed and presents
the magnitudes of significant factors and variables according to
work type (Tables 10–14). The results shown in Table 15 are the
major factors and variables that need to be addressed for the pre-
vention of worksite accidents.
1. Major worksite safety-management factors according to differ-
ence in time/work/accident factors and number of accident
victims by work type are shown in Table 15. Statistically
shown results have the following meanings:
• Because installations have a relationship with number of
accident victims according to month and accident type,
these factors and variables should be controlled;
• In tunnel, it is necessary to focus on day and hour;
• In drainage, prevention should be focused according to
hour, accident type, and accident location;
• In earth moving, month, day, accident cause, and accident
height are factors that have complex relationships with
number of accident victims and, as such, need to be con-
trolled as a part of safety precautions;
• In paving, because accident cause has a relationship with
number of accident victims, it is necessary to apply safety
precautions after a close investigation of accident
cause; and
• In structure, the factors of month, day, hour, weather,
accident type, accident cause, and accident height have
complex relationships with number of accident victims,
which means there are major accident factors that need
to be managed.
2. Major safety-management factors according to difference in
accident type and accident environment by work types are
as follows (Table 15):
Table 14. Difference Analysis of Temperature Factor and Accident Environment by Work Type
Accident
environment
Work type
Tunnel (χ2 ¼ 10.115; p ¼ 0.039a) Earth moving (χ2 ¼ 55.836; p ¼ 0.000b) Paving (χ2 ¼ 6.857; p ¼ 0.077c)
Temperature
(°C) Height
Temperature
(°C) Height
Temperature
(°C) Height
Accident
height
20–30 Surface level > 1–5 m 20–30 Surface level > 1–5 m > over 10 m 20–30 Surface level
10–20 Surface level > 1–5 m 10–20 Surface level > 1–5 m > over 10 m 10–20 Surface level > 1–5 m
0–10 Surface level > 1–5 m Above 30 Surface level > 1–5 m 0–10 Surface level
Above 30 Surface level 0–10 5–10 m Above 30 Surface level
Below 0 Surface level Below 0 5–10 m
Tunnel [χ2 ¼ 7.862; p ¼ 0.097 (p < 0.1)] Temperature (°C) Location Accident location 20–30, 10–20 Tunnel 0–10 Tunnel Below 0 Tunnel Above 30 Tunnel, wing wall Note: Only work types, factors, and variables with significant results are shown. ap < 0.05. bp < 0.01. cp < 0.01. JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / JUNE 2013 / 637 • The significant work types of accident cause and accident type are structure, installations, drainage, and earth moving; • Drainage shows a difference in accident type according to accident location; and • All work types show statistically meaningful results in the difference of accident type according to accident height. Such important analyses that can become a basis to trace the causal relationships of accident factors and to make research designed for the task. It is necessary to further investigate psy- chological, corporeal, environmental, and physical causes, such as the correlations between accident cause and accident type. 3. Major worksite safety-management factors according to differ- ence in time/work environment and accident environment by six work types are as follows (Table 15): • Installations have relationship with accident type according to month; • In tunnel, accident cause has difference according to weather and accident location has difference according to temperature; • In drainage, accident cause has difference according to temperature and accident height; • In earth moving, the factors that have relationship with number of accident victims are accident type, accident cause, and accident height according to month; accident cause and accident height according to weather; and acci- dent height according to temperature; • In paving, temperature and accident height are important factors; and • In structure, month and accident cause, day and accident type, hour and accident type, and accident height have relationships with accident occurrence. As such, it is confirmed that accidents occur as different fac- tors show complex relationships by six work types. In addition, Table 1 shows construction safety-management research with findings and factors. The research results show dis- agreements in findings and safety factors. This clearly shows a difference between this paper and other research. As shown in Table 15, Huh et al. (2010) presented the relationship between the convertible number of accident victims and significant factors such as time, accident cause, and accident type and found that accident types are closely related to accident cause and height. Although the Huh et al. (2010) research was limited to structure works, it may be suitable for other types of construction. According to this paper’s findings, as shown in Table 15, contributing factors vary with work type. Thus, this paper determines that the Huh et al. (2010) research results should be used in structure-construction works. This suggests that different factors should be considered for each work type. This paper supports that there are six work types that should be focused on in highway construction projects. The contribution of this research is to provide project personnel with safety factors to be considered in highway construction work types to help in accident prevention. It is predicted that accident management can be done effectively if the factors showing difference between accident factors are managed. This finding can be used to predict the probability of accidents based on six major highway work types, to prepare work type–specific training for workers, and to develop a strategy to reduce the number of accidents during highway construction. Table 15. Major Worksite Accident Factors according to Significant Difference Analysis between Accident Factors by Six Work Types (p) Work classification Significant difference, number of accident victims Significant difference of work environment and accident type Significant difference of time/work environment and accident environmentTime/work factor Accident environment factor Structure Month (0.063)a Accident cause (0.003)b Accident cause and accident type (0.000)b Month and accident cause (0.069)a Day (0.063)a Accident type (0.000)b Accident height and accident type (0.000)b Day and accident type (0.055)a Hour (0.036)c Accident height (0.000)b Hour and accident type (0.011)c Weather (0.062)a Hour and accident height (0.022)c Drainage Hour (0.000)b Accident type (0.000)b Accident cause and accident type (0.001)b Accident location (0.056)a Accident location and accident type (0.002)b Temperature and accident height (0.039)c Accident height and accident type (0.000)b Tunnel Day (0.035)c Accident height and accident type Weather and accident cause (0.000)b Hour (0.000)b Temperature and accident location (0.097)a Installations Month (0.019)c Accident type (0.002)b Accident cause and accident type (0.000)b Month and accident type (0.055)a Accident height and accident type (0.000)b Earth moving Month (0.019)c Accident cause (0.070)a Accident cause and accident type (0.007)b Month and accident type (0.014)c Day (0.061)a Accident height (0.002)b Accident height and accident type (0.012)c Month and accident cause (0.095)a Month and accident height (0.080)a Weather and accident height (0.001)b Temperature and accident height (0.000)b Paving Accident cause (0.059)a Accident height and accident type (0.013)c Temperature and accident height (0.077)a Prior research [Huh et al. (2010)] Adopted a research hypothesis: Structure Hour Accident cause, accident type, accident height Accident cause, accident type, and accident height No analysis ap < 0.1. bp < 0.01. cp < 0.05. 638 / JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / JUNE 2013 Conclusions and Suggestions This research has empirically analyzed whether accident factors have correlations with accident occurrences in each of six work types to support effective prevention and management of accidents at highway construction sites. The authors used a completely new set of nine factors and 248 variables. The statistical significance of the variables was measured, and another aspect of acknowledging and responding to construction accidents in highway construction was proposed. The results show that the magnitudes of accident factors and variables in six work types appear differently and have complex relationships among accident factors. Such results show that it is unreasonable to generally apply accident factors from one work type to all work types. Thus, accident prevention and management have to be applied differently according to the magnitudes of accident factors and variables for each work type. Major single and compound factors for intensive safety management that have mean- ingful results in six work types are summarized as follows: 1. Drainage; • Hour, accident type, accident location, accident cause and accident type, accident location and accident type, accident height and accident type, and temperature and accident height; 2. Tunnel; • Day, hour, accident height and accident type, weather and accident cause, and temperature and accident location; 3. Installation; • Month, accident type, accident cause and accident type, and month and accident type; 4. Earth moving; • Month, day, accident cause, accident location, accident cause and accident type, and accident height and accident type; 5. Paving; • Accident cause, accident height and accident type, and tem- perature and accident height, and; 6. Structure; • Month, day, hour, weather, accident cause, accident type, accident height, accident cause and accident type, accident height and accident type, month and accident cause, day and accident type, hour and accident type, and hour and accident height. The significance of this study is, first, that it presents a basis to apply the adopted approach and contributing factors for specific safety management to the six respective work types. Second, by analyzing these cases of accident occurrences empirically, the re- search model has identified accident factors that have relationships with each work type. Finally, the six work types and the environ- ment classifications of accidents identified in this study can be applied in similar highway construction projects. These findings can be used to predict the probability of acci- dents in six major highway work types, to aid in accident recog- nition and responses, to prepare customized training for workers, and to develop a strategy to reduce the number of accidents during highway construction. The contribution of this research is to pro- vide project personnel for the first time with managing factors to be considered in highway construction work types. Most construction takes place outside, and so weather will be a factor on most construction projects. Because the proposed model includes weather information in addition to accident information, this model will fit any project with work that is taking place in the elements. Because time information was considered, this model could be used to plan out safety programs based on days of the week. Although the study only looked at highway work, the im- portant aspect is that some days of the week are better than others for doing complicated construction work. From a planning perspec- tive, project managers can try to avoid planning work for days and times of day when accidents are more prevalent. If there are correlations among accident factors in highway con- struction, an epidemiologic survey of the causes has to be imple- mented for realistic safety-management responses. Thus, the authors suggest additional investigation to identify the causal rela- tionships of psychological, physical, environmental, and mana- gerial factors in accident environments, with worksite managers and construction workers as the target group, by using the factors and variables presented in this research. Also, the authors would like to suggest research on decision-making processes for active accident prevention and development of a safety-management system that can also be used as a training tool. Acknowledgments The authors would like to express their appreciation to the Korean Express Corporation for the valuable data used in this research. References Ahn, Y. S., Bena, J. F., and Bailer, A. J. (2004). “Comparison of uninten- tional fatal occupational injuries in the Republic of Korea and the United States.” Inj. Prev., 10(4), 199–205. Arditi, D., Ayrancioglu, M., and Shi, J. J. (2005). “Worker safety issues in night-time highway construction.” Eng. Construct. Architect. Manage., 12(5), 487–501. Arditi, D., Lee, D. E., and Polat, P. (2007). “Fatal accidents in nighttime vs. daytime highway construction work zones.” J. Saf. Res., 38(4), 399–405. Bryden, J. E. (1999). “Serious and fatal injuries to workers on highway construction projects.” Transportation Research Record 1657, Transportation Research Board, Washington, DC, 42. Dedobbeleer, N., and Beland, F. (1991). “A safety climate measure for construction sites.” J. Saf. Res., 22(2), 97–103. Fang, D. P., Huang, X. Y., and Hinze, J. (2004a). “Benchmarking studies on construction safety management in China.” J. Constr. Eng. Manage., 130(3), 424–432. Fang, D. P., Xie, F., Huang, X. Y., and Li, H. (2004b). “Factor analysis- based studies on construction workplace safety management in China.” Int. J. Proj. Manage., 22(1), 43–49. Gambatese, J. A., Behm, M., and Hinze, J. (2005). “Viability of designing for construction worker safety.” J. Constr. Eng. Manage., 131(9), 131–139. Hassanein, A., and Moselhi, O. (2004). “Planning and scheduling highway construction.” J. Constr. Eng. Manage., 130(5), 638–647. Hinze, J. (1992). “Role of designers in construction worker safety.” J. Constr. Eng. Manage., 118(4), 677–684. Hinze, J., and Raboud, P. (1988). “Safety on large building construction projects.” J. Constr. Eng. Manage., 114(2), 286–293. Hong, S. H. (2004). “Construction for safety management information model using the concept of design for safety.” Korea J. Constr. Eng. Manage., (5), 109–117. Huang, X., and Hinze, J. (2003). “Analysis of construction worker fall accidents.” J. Constr. Eng. Manage., 129(3), 262–271. Huh, W. C., Kim, Y. A., Hwang, U. S., and Kim, Y. S. (2010). “An analysis of accidents in the expressway structure construction.” Korean J. Constr. Eng. Manage., 11(3), 97–104. Korea Expressway Corporation (KEC). (2008). Construction Rep., KEC, Seoul. Korea Expressway Corporation (KEC). (2009). Highway Construction Guide Specification: Civil Engineering, KEC, Seoul. Korea Meteorological Administration. (2011). “Climate information.” ⟨http://www.kma.go.kr/weather/climate/average_6⟩ (Nov. 15, 2011). Korea Ministry of Employment and Labor (KMOEL). (2009). Statistics of Labor Rep., KMOEL, Seoul. JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / JUNE 2013 / 639 http://dx.doi.org/10.1136/ip.2003.004895 http://dx.doi.org/10.1108/09699980510627171 http://dx.doi.org/10.1108/09699980510627171 http://dx.doi.org/10.1016/j.jsr.2007.04.001 http://dx.doi.org/10.1016/0022-4375(91)90017-P http://dx.doi.org/10.1061/(ASCE)0733-9364(2004)130:3(424) http://dx.doi.org/10.1061/(ASCE)0733-9364(2004)130:3(424) http://dx.doi.org/10.1016/S0263-7863(02)00115-1 http://dx.doi.org/10.1061/(ASCE)0733-9364(2005)131:9(1029) http://dx.doi.org/10.1061/(ASCE)0733-9364(2004)130:5(638) http://dx.doi.org/10.1061/(ASCE)0733-9364(1992)118:4(677) http://dx.doi.org/10.1061/(ASCE)0733-9364(1988)114:2(286) http://dx.doi.org/10.1061/(ASCE)0733-9364(2003)129:3(262) http://dx.doi.org/10.6106/KJCEM.2010.11.3.97 http://dx.doi.org/10.6106/KJCEM.2010.11.3.97 http://www.kma.go.kr/weather/climate/average_6 http://www.kma.go.kr/weather/climate/average_6 http://www.kma.go.kr/weather/climate/average_6 http://www.kma.go.kr/weather/climate/average_6 Korea Ministry of Employment and Labor (KMOEL). (2011). “Employ- ment low information system.” ⟨http://www.wage.go.kr/time/ timesearch_tot.jsp⟩ (Nov. 15, 2011). Korea Occupational Safety and Health Agency (KOSHA). (2004). Major Accident Cause Analysis Rep., KOSHA, Seoul. Korea Occupational Safety and Health Agency (KOSHA). (2005). Major Accident Cause Analysis Rep., KOSHA, Seoul. Korea Occupational Safety and Health Agency (KOSHA). (2006). Major Accident Cause Analysis Rep., KOSHA, Seoul. Korea Occupational Safety and Health Agency (KOSHA). (2007). Indus- trial Accident Rep., KOSHA, Seoul. Korea Occupational Safety and Health Agency (KOSHA). (2009). Indus- trial Accident Rep., KOSHA, Seoul. Lehtola, M. M., and Van Der Molen, H. F. (2008). “The effectiveness of interventions for preventing injuries in the construction industry: A systematic review.” Am. J. Prev. Med., 35(1), 77–85. Li, Z. (2010). “Highway work zone safety audits for safety improvements.” Eng. Construct. Architect. Manage., 17(5), 512–526. Lim, J. Y., Han, K. K., and Kim, S. K. (2008). “A study of client’s role for safety management at construction site.” Korea Inst. Build. Constr., 8(5), 75–84. Mohan, S., and Zech, W. C. (2005). “Characteristics of worker accidents on NYSDOT construction projects.” J. Saf. Res., 36(4), 353–360. New York State Dept. of Transportation (NYSDOT). (2002). “New York State Dept. of Transportation construction safety program.” 1990–2001 Summary Rep., Office of Construction, Albany, NY. Occupational Safety and Health Administration (OSHA). (2009). The OSHA Database 2009–2010, U.S. Dept. of Labor, Washington, DC. Sawacha, E., Naoum, S., and Fong, D. (1999). “Factors affecting safety performance on construction sites.” Int. J. Proj. Manage., 17(5), 309–315. Sin, S., and Lee, M. (2010). Construction of middle-aged and elderly workers’ safety and health, ChoungMunGak, Seoul. Son, C. B., Park, C. S., Hong, S. H., and Choi, S. I. (2002). “An evaluation of safety management performance in head office and job site of large construction firms.” Archit. Inst. Korea, 18(6), 135–142. Teo, E. A. L., Ling, F. Y. Y., and Chong, A. F. W. (2005). “Framework for project managers to manage construction safety.” Int. J. Proj. Manage., 23(4), 329–341. Toole, T. M. (2002). “Construction site safety roles.” J. Constr. Eng. Manage, 128(3), 203–210. U.S. Bureau of Labor Statistics. (2004). “Census of fatal occupational in- juries profiles data.” CFOI Rep., U.S. Dept. of Labor, Washington, DC. U.S. Bureau of Labor Statistics. (2005). “Census of fatal occupational in- juries profiles data.” CFOI Rep., U.S. Dept. of Labor, Washington, DC. U.S. Bureau of Labor Statistics. (2006). “Census of fatal occupational in- juries profiles data.” CFOI Rep., U.S. Dept. of Labor, Washington, DC. 640 / JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / JUNE 2013 http://www.wage.go.kr/time/timesearch_tot.jsp http://www.wage.go.kr/time/timesearch_tot.jsp http://www.wage.go.kr/time/timesearch_tot.jsp http://www.wage.go.kr/time/timesearch_tot.jsp http://www.wage.go.kr/time/timesearch_tot.jsp http://www.wage.go.kr/time/timesearch_tot.jsp http://dx.doi.org/10.1016/j.amepre.2008.03.030 http://dx.doi.org/10.1108/09699981011074592 http://dx.doi.org/10.5345/JKIC.2008.8.5.075 http://dx.doi.org/10.5345/JKIC.2008.8.5.075 http://dx.doi.org/10.1016/j.jsr.2005.06.012 http://dx.doi.org/10.1016/j.ijproman.2004.09.001 http://dx.doi.org/10.1016/j.ijproman.2004.09.001 http://dx.doi.org/10.1061/(ASCE)0733-9364(2002)128:3(203) http://dx.doi.org/10.1061/(ASCE)0733-9364(2002)128:3(203) Copyright of Journal of Construction Engineering & Management is the property of American Society of Civil Engineers and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. Institutional determinants of construction safety References Ju, C., & Rowlinson, S. (2014). Institutional determinants of construction safety management strategies of contractors in Hong Kong. Construction Management & Economics, 32(7/8), 725-736. doi:10.1080/01446193.2014.909048 Abstract: Workplace safety in the construction industry of Hong Kong is regulated by a mix of enforcement and performance-based approaches. The two approaches are underpinned by different institutional structures and lead to divergent safety practices. To examine how contractors strategically respond to the complex institutional demands for safety performance, contractors’ day-to-day site safety practices were investigated. Safety practice data were obtained from 62 open-ended interviews and project archives in a case study. Different supervision patterns, i.e. enforcement and localized approaches were found to coexist on site. Discrepancies were found between workers’ self-reported safety awareness and safety awareness assessed by their supervisors. The evidence suggests that contractors implemented compromise and avoidance strategies. The complex institutional environment, especially the incompatible progress and safety requirements, was found to be a key determinant of mixed site safety practices. Institutional theory is explored as a possible theoretical perspective to explain contractors’ safety management strategies. An institutional level change of safety management strategies is suggested. [ABSTRACT FROM AUTHOR] Institutional determinants of construction safety Institutional determinants of construction safety management strategies of contractors in Hong Kong CHUANJING JU* and STEVE ROWLINSON Department of Real Estate and Construction, University of Hong Kong, Pokfulam Road, Hong Kong, China Received 12 November 2013; accepted 24 March 2014 Workplace safety in the construction industry of Hong Kong is regulated by a mix of enforcement and perfor- mance-based approaches. The two approaches are underpinned by different institutional structures and lead to divergent safety practices. To examine how contractors strategically respond to the complex institutional demands for safety performance, contractors’ day-to-day site safety practices were investigated. Safety practice data were obtained from 62 open-ended interviews and project archives in a case study. Different supervision patterns, i.e. enforcement and localized approaches were found to coexist on site. Discrepancies were found between workers’ self-reported safety awareness and safety awareness assessed by their supervisors. The evidence suggests that contractors implemented compromise and avoidance strategies. The complex institutional environ- ment, especially the incompatible progress and safety requirements, was found to be a key determinant of mixed site safety practices. Institutional theory is explored as a possible theoretical perspective to explain contractors’ safety management strategies. An institutional level change of safety management strategies is suggested. Keywords: Contractor, Hong Kong, institutions, strategy-as-practice. Introduction Performance-based occupational safety and health legislation was introduced by the Hong Kong govern- ment in 1995. In contrast to the prescriptive approach that emphasizes ‘standard setting and enforcement’, the new approach imposes ‘a set of general duties’ on relevant bodies and promotes ‘industry self-regulation’ (Rowlinson, 2003, p. 23). The current legislation is a combination of the two approaches (Labour Department, 2011). Under such a legislative framework, contractors must comply with prescriptive rules of safe conduct. Meanwhile, they need to employ proactive safety management approaches to satisfy a variety of safety requirements or expectations of the stakeholders, for instance the client, regulators and trade associations. The aim of this study is to examine how contractors cope with complex safety demands from an institutional theory perspective. Oliver’s (1991) conceptual framework that predicts how organizations react strategically to institutional influence is adopted. Theoretical background Resource dependence perspective The traditional task environment based on resource dependence theory emphasizes the roles of markets, resources, and competition in determining organiza- tional processes and outcomes (Pfeffer and Salancik, 1978). Within-firm decision-making and external stra- tegic factors are seen to determine resource selection and accumulation (Oliver, 1997a). Managerial deci- sion-making within an organization is guided by an eco- nomic rationality and motives of efficiency, effectiveness and profitability (Conner, 1991). The external strategic factors that impact on organizations include ‘buyer and supplier power, intensity of compe- tition, and industry and product market structure’ (Oliver, 1997a, p. 698). From a resource-based per- spective, economic motives drive resource procurement decisions, and economic factors in the firm’s competi- tive and resource environments drive firm conduct and outcomes (Oliver, 1997a). Although the resource *Author for correspondence. E-mail: jcjing@hku.hk © 2014 Taylor & Francis Construction Management and Economics, 2014 Vol. 32, Nos. 7–8, 725–736, http://dx.doi.org/10.1080/01446193.2014.909048 mailto:jcjing@hku.hk http://dx.doi.org/10.1080/01446193.2014.909048 dependence view provides insightful explanations of organization behaviours, it makes the role of institu- tional environment in both decision-makers’ choices and organizational processes implicit. Institutional perspective Neo-institutional theory emerged in the late 1970s, as a reaction against the view of organizations as rational, responding exclusively to economic pressures for resources (Suddaby et al., 2013). Institutional theory suggests that organizations respond to pressures in their social and symbolic environment rather than simply economic pressures (Meyer and Rowan, 1977). From an institutional perspective, organizations operate within a social framework of rules, norms, values and taken-for-granted assumptions about what constitute appropriate behaviours (Meyer and Rowan, 1977; Zucker, 1977; DiMaggio and Powell, 1983; Powell and DiMaggio, 1991). One important tenet of institu- tionalism is structural isomorphism that describes how the structures and practices of organizations in a certain institutional field become isomorphic over time (Meyer and Rowan, 1977; DiMaggio and Powell, 1983; Powell and DiMaggio, 1991). Conformity to institutional pressures or social expectations helps orga- nizations gain legitimacy, resources and survival capa- bilities (DiMaggio and Powell, 1983; Ashforth and Gibbs, 1990; Suchman, 1995). However, institutional theory was criticized for its inadequate attention to the role of managerial agency and strategic choice (Deephouse, 1999). In response to these criticisms, Oliver (1991) con- structed a conceptual framework based on convergent ideas of institutional and resource dependence perspec- tives. The framework predicts that organizations may respond to institutional requirements in a variety of strategic modes ranging from passive compliance to active manipulation. Oliver’s (1991) framework was adopted in this study to examine contractors’ safety management strategies for the following reasons. Linkages between the two perspectives and construction safety research Company safety records have increasingly important economic and business implications. Occupational accidents interrupt the production process, generate accident-related costs (Hinze, 2000), and undermine a company’s internal climate (Fernández-Muñiz et al., 2009). Serious accidents harm a company’s public image and reputation (Smallman and John, 2001), thereby reducing its organizational competitiveness (Fernández et al., 2000). Serious accidents may cause huge financial losses for contractors in Hong Kong. When contractors have caused or contributed to the occurrence of a serious incident on a construction site, they can be suspended from tendering for public works for one to 12 months (Development Bureau, 2009). From this point of view, it is in the construction com- panies’ best interests to take various actions to manage safety in the workplace (Wilson and Koehn, 2000). Nevertheless, the difficulties of managing safety in the workplace are evident (Zohar and Luria, 2003). First, the paradox of safe behaviours poses serious chal- lenges for management. Safe behaviours result in non- events (e.g. avoidance of low probability injury) but generate immediate costs (e.g. slower pace, extra effort), whereas unsafe behaviours bring about immedi- ate reinforcement (hindsight biased) and tangible ben- efits (e.g. increased pace, greater comfort) (Zohar and Luria, 2003). Therefore, when people perceive a low probability of an accident, they may ignore the safe working rules or procedures. Secondly, some procure- ment practices do not motivate contractors to put safety first, for instance ‘the high level of subcontract- ing’ and ‘the continued acceptance of the lowest ten- der’ (Rowlinson, 2003, pp. 13–14). On the contrary, contractors are incentivized to pursue speedier produc- tion pace and reduce investments in safety. Institutional interventions are thus introduced to promote safe working rules, norms and the value of safety through the mechanisms of regulative, normative and cultural-cognitive institutions (Scott, 2008). First, site safety practices are regulated by mandatory rules and regulations. Employers must ensure, as far as is reasonably practicable, the health and safety at work of their employees. Secondly, safety is regarded as a moral obligation by stakeholders who do not hold legal responsibilities, such as professional or trade associa- tions. These bodies play an important role in formulat- ing, transforming, and disseminating safety values and norms. Thirdly, the cultural-cognitive aspect of institu- tions is relevant to safety culture (Koch, 2013). Safety culture determines people’s ‘commitment to, and the style and proficiency of, an organisation’s health and safety management’ (Health and Safety Commission, Advisory Committee on the Safety of Nuclear Installa- tions, 1993, p. 23). Safety-culture programmes have been used as tools to enhance the commitment of the workforce and front-line supervisors to construction site safety (Sherratt et al., 2013). Based upon the above argument, it is essential to consider both resource-based and institutional perspec- tives in construction safety research. A resource-based perspective emphasizes contractors as active agencies and the roles of markets and clients in the process of safety management. An institutional perspective pays attention to the influence of regulations, collective 726 Ju and Rowlinson norms and cultural forces on safety management struc- tures and practices. The two perspectives provide com- plementary views of organizational forms and behaviours. Thus, it seems tenable and potentially meaningful to examine contractors’ safety management strategies building on Oliver’s framework. The Hong Kong context To analyse the institutional elements concerning con- struction safety in Hong Kong, Scott’s (2008) three pil- lars of institutions theory, namely regulative, normative and cultural-cognitive, are adopted as an analytical tool. It is worth noting that these three pillars of institu- tions are analytically distinct but impact on organiza- tions ‘in interdependent and mutual reinforcing ways’ (Scott, 2008, p. 50). Legal environment Despite a transition in safety legislation approach from prescriptive to performance-based, prescriptive legisla- tion and enforcement actions remain the essential instruments in influencing the safety performance of those contractors that are not persuaded to embrace safety management (Rowlinson, 2003). Statutory pro- visions governing work safety on construction sites are set out under the Factories and Industrial Undertakings Ordinance (FIUO) (Cap. 59), the Occupational Safety and Health Ordinance (OSHO) (Cap. 509) and their subsidiary legislation (Labour Department, 2011). Regulative institutions emphasize explicit regulatory processes, such as rule-setting, monitoring, and sanc- tioning activities (North, 1990). The impact of the legal environment on organizations’ conformity behaviours is discussed from two perspectives, namely regulatory stringency (Oliver, 1997b; Delmas and Toffel, 2008) and enforcement (Gainet, 2010; Short and Toffel, 2010). Regulatory stringency is defined as the complexity and burden of regulatory environment (Fennell and Alexander, 1987). High regulatory stringency refers to particularly extensive or severe regulatory pressures exerted on organizations (e.g. a large number of enforceable rules, regulations and codes) that poten- tially impede organization’s autonomy or efficiency (Oliver, 1997b). Biggs et al. (2011) found that ‘plenty of paperwork’ required by regulations distracts contrac- tors’ attention from specific safety problems and under- mines the integration of safety practices into normal construction processes. Regulatory stringency can be measured through the extent to which contractors per- ceive that workplace safety is or is not overregulated with many government codes, rules and regulations, and the extent to which contractors feel that the regu- latory environment does or does not reduce their dis- cretion to manage safety effectively (Oliver, 1997b). Enforcement is an important instrument in ensuring compliance with safety legislation. Site inspection is a primary enforcement tool used by the Labour Depart- ment (LD) in Hong Kong. The LD is empowered to initiate prosecutions, issue improvement notices (INs) and suspension notices (SNs), when breaches of the law or imminent risks are identified during site inspec- tions (Labour Department, 2011). Anderson (2007) argued that the effective enforcement of the legislation on construction health and safety issues is just as important as the law itself. The effectiveness of enforce- ment actions depends on the intensity (Gainet, 2010) and the capacity of enforcement bodies (Campbell, 2007). Campbell (2007) argued that if the enforcement capacity is developed on the basis of negotiation and consensus-building among contractors, government and other relevant stakeholders, contractors are more likely to comply with prescribed legal requirements. Normative institutional environment Normative aspects of institutions include both values and norms. Values are conceptions of the preferred or the desirable; norms specify how things should be done (Scott, 2008). Compared with regulative institutions that emphasize the ‘logic of consequentiality’, norma- tive institutions shift attention to the ‘logic of appropri- ateness’ (March and Olsen, 1989, p. 23). Normative institutions generally take form of ‘rules-of-thumb, standard operating procedures, occupational standards, and educational curricula’ (Hoffman, 1999, p. 353). Normative institutions exert influence on organiza- tional actions and beliefs via professional networks (e.g. professional and trade associations) and creden- tialing institutions (e.g. training institutions and univer- sities) (DiMaggio and Powell, 1983; Honig and Karlsson, 2004; Heugens and Lander, 2009). Many occupational groups, both professional and craft-based, generate and enforce work norms and actively promulgate standards and codes to govern con- duct (van Maanen and Barley, 1984; Brunsson and Jacobsson, 2000). In Hong Kong, safety guidelines, manuals or handbooks are developed and promoted through the collaboration of multiple bodies, including government departments (e.g. LD), industry statutory bodies (e.g. the Construction Industry Council (CIC), the Occupational Safety and Health Council (OSHC)), clients of public projects (e.g. the Housing Authority (HA)), and leading contractors. Specifically, the LD provides non-statutory guidelines to facilitate Safety management 727 compliance with the relevant safety and health legisla- tion (Labour Department, 2011). The CIC, as a statu- tory industry coordinating body, performs an important role in promoting industry self-regulation, and formu- lating codes of conduct and promoting such codes (Construction Industry Council, 2013). The OSHC is another statutory body which has a mandate to improve workplace safety and health (Occupational Safety and Health Council, 2013). The involvement of clients also facilitates the promulgation of safety standards. For example, the HA regularly convenes safety seminars and site safety forums with relevant stakeholders (http://goo.gl/uQl8u6). In addition, the Hong Kong Construction Association (HKCA) with most leading contractors as corporate members is active in disseminating safety initiatives and good safety practices among its members (Hong Kong Construc- tion Association, 2013). Credentialing institutions help to foster safe work norms through professional education and training. For instance, the Hong Kong Institution of Engineers (HKIE) collaborates with higher education institutions through accreditation of an engineering higher diploma programme to ensure the necessary education on occu- pational safety (Hong Kong Institution of Engineers, 2013). Cultural-cognitive institutional environment Cultural-cognitive elements of institutions stress both ‘the symbolic systems perceived to be objective and external to individuals’ and the subjective ‘interpreta- tion processes’ of individuals (Scott, 2008, p. 57). Cul- ture is the learning result of a group of people over a period of time. The deepest level of culture is in the cognition where a group shares ‘perceptions, language, and thought processes’ that become the ultimate deter- minant of ‘feelings, attitudes, espoused values, and overt behaviour’ (Schein, 1990, p. 111). Despite the fact that a clear consensus on the definition of safety culture is yet to evolve (Guldenmund, 2000), the con- cept generally depicts the set of beliefs, values, atti- tudes, and perceptions guiding people to behave in a safe manner (e.g. Cox and Cox, 1991; Hale, 2000). In order to improve safety performance on construction sites, it is crucially important to cultivate a safety cul- ture (Guldenmund, 2010; Fang and Wu, 2013). Many large contractors have implemented safety- culture programmes in the construction industry of Hong Kong. These safety-culture programmes (e.g. Gammon: Zero Harm, Leighton: Strive for Life) are led by a zero accident vision. In a zero accident vision, each individual’s commitment to site safety is essential. Indeed, both technical and social innovations are needed, as well as a paradigm shift in thinking for solv- ing existing safety problems (Zwetsloot et al., 2013). Some large contractors recognize that front-line super- visors play a critical role in promoting safety practices. They have launched a series of safety initiatives, such as front-line supervisor leadership training and best safety foreman incentive schemes. Supervisors gradually rec- ognize that the safety policing approach (i.e. blaming or punishment) has become less effective. Therefore, they need to seek new ways to communicate with work- ers on the basis of caring and respect. The market environment Both institutional and resource dependence theorists argue that organizations are less likely to resist external pressures when they are dependent on stakeholders who exert these pressures (Pfeffer and Salancik, 1978; DiMaggio and Powell, 1983). In practice, clients may exert significant influence on health and safety manage- ment undertaken by contractors through procurement strategies, contractual control, and incentive schemes. For example, historical safety records of construction companies are taken into account in the tender evalua- tion of HA projects. Points will be deducted if compa- nies have records of repeated breaches of safety regulations and serious accidents (Cheng et al., 2013). In contrast, clients in the private sector prefer to leverage the lowest tender price method to procure construction services, which is detrimental to site safety performance. Main contractors in Hong Kong usually employ a multi-tier subcontracting strategy to cope with the dramatic changing demand for in-house resources (Ng et al., 2009). However this strategy challenges main contractors during a market boom because of the inherent labour shortage and difficulties of coaching newcomers. According to the Labour Department (2013), the number of construction site workers increased by 40% from 51 000 in 2009 to 71 300 in 2012, indicating that many newcomers joined the industry. Managing new workers on construction sites poses serious safety challenges for management. In summary, contractors are exposed to a complex institutional environment regarding construction safety. Institutional complexity informs the multiplicity of institutional demands as well as the relative incom- patibility between demands (Greenwood et al., 2011). Construction sites, in particular, are under the scrutiny of multiple stakeholders including the LD, safety audi- tors, the client, and company head office. These parties exert coercive pressure on construction sites through enforcing safety regulations, prosecuting violations, and implementing contractual provisions and in-house 728 Ju and Rowlinson http://goo.gl/uQl8u6 policies. Contractors are also faced with normative pressure. The government and professional bodies pro- mulgate safety guidelines regarding some specific risks on sites, such as working in hot weather and tower crane operation. Site inspections will be carried out along with the promotional activities. Contractors are encouraged to participate in various safety competitions and promotional events organized by their clients, the OSHC, the CIC and the government. Although all ini- tiatives are proposed to improve site safety, some approaches have inherent conflicts, for example the leg- islative enforcement versus safety-culture programmes, the economic motives of unsafe behaviours versus social values of safety. Contractors’ strategic responses The complex institutional environment inevitably gen- erates challenges and tensions for organizations exposed to it (Greenwood et al., 2011). Organizations may exercise some level of strategic choice (Friedland and Alford, 1991; Whittington, 1992; Clemens and Cook, 1999; Seo and Creed, 2002; Dorado, 2005). Oliver (1991) classified the strategic choices into five types: acquiescence, compromise, avoidance, defiance, and manipulation (the sequence indicates an increasing level of resistance to institutional demands). Acquiescence strategies indicate full conformity to institutional demands or expectations. When con- fronted with incompatible institutional demands, orga- nizations are less likely to employ an acquiescence strategy (Oliver, 1991). This is because full conformity to some demands implies a violation of the contradict- ing ones (Oliver, 1991; Kraatz and Block, 2008; Pache and Santos, 2010). Instead, organizations may adopt compromise, avoidance, defiance and manipulation strategies (Oliver, 1991). Given the complex institutional demands for safety in Hong Kong, we assume that contractors are more likely to adopt compromise and avoidance strategies. Defiance and manipulation as explicitly resistant strat- egies are less likely to be adopted in light of the strong coercive pressure from the government and the high dependence on relationships with clients. Compromising Research has pointed to compromise as a viable strat- egy for organizations facing competing demands (Oliver, 1991; Kraatz and Block, 2008). Specifically, organizations may attempt to balance, pacify, or bar- gain with external constituents (Oliver, 1991). Balance is an organizational attempt to achieve parity among or between multiple stakeholders and internal interests. An organization that employs pacifying tactics devotes most of its energies to appeasing or placating the insti- tutional source or sources it has resisted. Bargaining is a more active form of compromise than pacifying. Bar- gaining tactics involve the effort of the organization to exact some concession from an external constituent in its demands or expectations (Oliver, 1991). Avoidance When faced with competing demands, organizations may also adopt avoidance strategies by concealing their nonconformity, buffering themselves from institu- tional pressures, or escaping from institutional rules or expectations (Oliver, 1991). Concealment tactics involve disguising nonconformity behind a façade of acquiescence. Buffering refers to an organization’s attempt to reduce the extent to which it is externally inspected, scrutinized, or evaluated by partially detach- ing or decoupling its technical activities from external contact (Pfeffer and Salancik, 1978; Boxenbaum and Jonsson, 2008). A more dramatic avoidance response to institutional pressures toward conformity is escape. By embracing such tactics, an organization may exit the domain within which pressure is exerted or signifi- cantly alter its own goals, activities, or domain to avoid the necessity of conformity altogether (Oliver, 1991). Research method The aim of the study is to investigate how contractors respond to the complex institutional demands for safety performance. The strategic responses of contractors were investigated building on a practice approach (Jarzabkowski et al., 2007). The practice approach focuses on the way that actors interact with the social and physical features of context in the everyday activities that constitute practice (Jarzabkowski, 2004). Besides, organizational forms and management practices are manifestations of, and legitimated by, institutional demands (Greenwood et al., 2010). Thus, a detailed examination of safety practices could inform management strategies as well as the underlying institutional requirements or expectations. A case study approach which allows for an in-depth investigation of safety practices within their real-life context was employed (Yin, 2003). This case study was part of an action research project. The broader research project was carried out in a large-scale railway project in Hong Kong, aiming to identify site safety issues, diagnose the underlying causes, propose and evaluate intervention strategies. The research project Safety management 729 comprised two phases (i.e. pre- and post-intervention phases). The empirical evidence presented in this study was drawn from the pre-intervention phase, given that the focus of investigation was on site safety practices. Data collection At the project level, safety practices include safety- related management decisions, implementation of the safety management system, and day-to-day actions of individuals at work. Two data collection techniques, project documents review and open-ended interviews, were adopted in order to capture a full perspective of site safety practices. Project documents provide, in a formal manner, a big picture of overall site safety man- agement as well as project background information. Open-ended interviews, in a complementary manner, provide situated information of safety-in-practice. The research team was authorized to access various archives, including project organization structure, safety meeting minutes, accidents statistics and investi- gation reports, site management meeting minutes, and site inspections analysis. Open-ended questions were designed to probe into site safety issues. Participants were asked to describe their perceptions of site safety management, safety problems, and possible causes. In total, 62 interviews with site personnel were con- ducted. The demographic information on interview participants is summarized in Table 1. The interviews lasted for five to 40 minutes depending on participants’ willingness to share their views and on their time conve- nience. Interviews with supervisors were tape-recorded and transcribed verbatim. In order to encourage work- ers to speak freely about site issues, interviews with workers were carried out in an informal manner, and notes were taken. Data analysis A thematic analysis approach was adopted to identify, analyse and report patterns of meanings (Braun and Clarke, 2006). Themes were generated from raw data through an inductive coding strategy. Higher order cat- egories were developed through clustering lower level themes (Creswell, 2009). While identifying the main themes, the interview raw data were read and coded. During the process of clustering lower level themes, a recurrence of com- ments about several issues was observed. Six categories of safety issues emerged from the coding process (see Table 2). At a subsequent step of analysis, the emergent issues were grouped in accordance with organizational identities of interview participants. Participants were classified into five groups as presented in Table 1. Supervisors refer to people who are on the upper level of the organizational hierarchy or those who have actual control over or influence on workers. The perceptions of different groups were compared at this stage. Finally, the linkage between site safety practices and contractors’ safety management strategies was estab- lished. Institutional analysis of contractors’ safety man- agement strategies was conducted. Case study Project background The railway project, valued at approximately HK$1.34 billion, involves one station and a 650-metre long over- run tunnel. The construction work commenced in March 2010 and was expected to complete in June 2014. The project is located at the centre of a commu- nity and is highly visible to the public. The main con- tractor (MC) is regarded as one of the leading construction companies in Hong Kong. The client with the Hong Kong government as the sole shareholder implements stringent safety management standards. The targeted accident frequency rate for the year 2012 is less than 0.30 (number of reportable accidents per 100 000 man-hours worked). However, owing to the complex site conditions, stringent health, safety and environmental (HSE) requirements, and separated construction sites, the MC faces huge safety manage- ment challenges. Table 1 Summary of interview participants Groups Organizational positions Client supervisors (6) Engineer (3); Inspector (3) Main contractor supervisors (7) Site manager (1); Engineer (2); Foreman (3); Supervisor (1) Main contractor workers (11) Worker (11) Subcontractor supervisors (3) Supervisor (3) Subcontractor workers (35) Leading hand (2); Worker (33) Note: Numbers in brackets indicate the number of participants in that group/position. 730 Ju and Rowlinson At the time that the case study commenced in April 2012, there were nine subcontractors (SCs) working concurrently on site. The accident frequency rate of the project was 0.42 which was the highest among the client’s projects. The causes pointed out by the client included poor housekeeping, and ‘walking by’ safety issues. Findings Six categories of safety issues emerged. These are: supervisory safety leadership; safety practices; poor planning; resources provision; safety awareness, knowl- edge, and competence of workers; and communication (see Table 2). The frequency of safety issues as men- tioned by interview participants is shown in Table 2, which indicates the importance of those safety issues. Competing goals of progress and safety Competing priorities, those of progress versus safety, were evident on this project. Table 3 shows the number of participants who mentioned ‘progress pressure’, ‘bal- ance safety and progress goals’ or ‘safety is in conflict with production’. The client had a high expectation of project safety performance, set up challenging safety targets, and adopted a systematic safety management approach to monitor the contractor’s safety perfor- mance. For example, the client carried out site inspec- tions regularly and conducted DNV (Det Norske Veritas) safety audits every six months. The client also employed a significant number of inspectors to monitor closely project quality, progress, safety, and workman- ship. Meanwhile a major proportion of participants in the groups of supervisors mentioned that they were faced with the pressure of project progress. Among 16 supervisory participants, 63% mentioned that it was not realistic to achieve both targets. In contrast only 22% of workers reported the competing goals of pro- gress and safety in their daily work. The data suggest that management staff experienced stronger production pressure than workers. How supervisors coped with the competing goals was probed by examining their super- visory safety leadership. Mixed safety supervision patterns Supervisory safety leadership was coded corresponding to participants’ perceptions of their supervisors’ responses toward site safety problems (e.g. violation of safe operating procedures). Contradictory safety supervision patterns of the MC and SC supervisors were found. MC supervisors mainly adopted an enforcement approach when carrying out site inspec- tions. Supervisors embracing such an approach strictly implemented safety rules, and they would stop the work process immediately on spotting safety violations. Comments regarding supervision of the MC supervi- sors are quoted as examples. We are faced with different pressure from the MC and my boss … MC supervisors emphasize compliance with safety procedures. (SC supervisor) If we don’t take sufficient safety measures, the MC won’t let us start. (SC worker) In contrast, SC supervisors employed a more localized supervision pattern. Instead of bearing ‘safety is the pri- ority’ in mind, almost all supervisors from SCs focused on production and progress as reported by workers. It is worth noting that a small proportion of MC supervisors also shared this supervision pattern. Workers employed by SCs commented that: The major responsibility of foremen and supervisors of subcontractors is production. (SC worker) Table 2 Emergent safety issues from interviews Categories of safety issues � Supervisory safety leadership (52) � Localized approach (28) (MC 9, SC 18, Client 1) � Enforcement approach (24) (MC 21, Client 3) � Safety practices (40) � Management words are consistent with actions (19) � Management words are inconsistent with actions (13) � Workers follow codes of safe conduct (4) � Workers violate codes of safe conduct (4) � Poor planning (36) � Progress and safety are in conflict (20) � Work access and housekeeping (9) � Insufficient planning and design (7) � Resources provision (22) � Provision of personal protective equipment and other safety facilities (13) � Shortage of workers and supervisors (9) � Safety awareness, knowledge, and competence of workers (21) � Workers take safety ownership (self-reported) (14) � Lack of safety awareness (reported by supervisors) (7) � Communication (10) � Communication between MC and SC (5) � Communication among SCs (3) � Communication within organization (2) Note: Numbers in brackets indicate the frequency of issues mentioned by interview participants. Safety management 731 Sometimes they (their own supervisors) turn a blind eye to unsafe behaviours. (SC worker) One important feature of the localized supervision pat- tern is that supervisors preferred to use their own expe- rience and safety knowledge instead of prescribed rules to assess the risks in the working environment and workers’ operations. This supervision pattern may explain why the ‘walking by’ phenomenon was fre- quently observed. One SC supervisor mentioned that ‘as long as the work can be done safely according to my experience, I will let it go.’ Another example is quoted from a MC engineer: Workers know they are safe or not. They have longer experience than me. They are more familiar with the site conditions than me. It is unnecessary for them to fully follow safety prescriptions. Discrepancies in workers’ safety awareness Discrepancies were found between workers’ self- reported safety awareness and safety awareness reported by their supervisors. Workers’ self-reported safety awareness is quite coordinated and positive. Workers thought that they had their own way of han- dling site issues and they were safety conscious. A majority of workers mentioned that they would refuse to work if the site environment was unsafe according to their experience and knowledge. In their own words: We will refuse to work if we think the situation is dan- gerous. Life and safety are the priority. (SC worker) It is worth nothing if I lose my life. (MC worker) Safety is mainly our own business. We should be very careful of ourselves and our co-workers. (SC worker) Since workers took safety personally, some of them complained that they were overregulated by the MC supervisors. For instance, a worker from the MC com- mented that ‘we could benefit from the strict safety management measures, but we are annoyed about being interrupted again and again.’ In contrast, most supervisors, including foremen, engineers and managers, commented that workers’ safety awareness was low because workers tended to take shortcuts instead of fully complying with safety procedures. A foreman from the MC, for example, mentioned that ‘Some older workers think they have rich site experience. Actually their safety awareness is low. If nobody is watching them, they won’t follow safety procedures.’ The discrepancies are probably because the basis upon which safety awareness is formed and assessed is different. Workers’ awareness of safety is generated from day-to-day practices in the workplace where situ- ational factors must be considered. Therefore, workers’ logic of actions has a certain degree of flexibility and is pragmatic. Supervisors make an assessment of the safety awareness of workers according to their safety knowledge that is gained from professional education and continuous professional development. The assess- ment is thus to some extent derived from a rigid enforcement logic. Contractors’ strategic responses: compromise Both mixed safety supervision patterns and discrepan- cies of workers’ safety awareness indicate contractors’ strategic responses to the contradiction of legitimacy and efficiency (e.g. safety versus progress). The data suggest that contractors in this project adopted com- promise and avoidance strategies to cope with compet- ing goals of progress and safety. This agrees with previous arguments that organizations facing compet- ing institutional demands or inconsistencies between institutional expectations and internal organizational objectives may adopt the strategy of compromising or decoupling (Oliver, 1991; Westphal and Zajac, 2001; Fiss and Zajac, 2006; Kraatz and Block, 2008). Table 3 Competing goals of ‘progress and safety’ Groups Number of participants Percentage (%) Quote Client (6) 5 83 ‘The progress pressure is from the client. The milestones set up by the top management are not realistic.’ MC supervisors (7) 3 43 ‘Under progress pressure, it is not realistic to achieve safety standards and progress target at the same time.’ MC workers (11) 4 36 ‘Sometimes we need to complete the order immediately and no time to do it safely.’ SC supervisors (3) 2 67 ‘When we got the contract from the MC, the progress of the project had already fallen behind.’ SC workers (35) 6 17 ‘When we are under progress pressure, we may ignore safety procedures.’ Note: Numbers in brackets indicate the total number of participants in that group. 732 Ju and Rowlinson Compromise strategies may take three forms: bal- ance, pacifying and bargaining (Oliver, 1991). The comments of the client regarding site safety manage- ment indicate that contractors tried to obtain an acceptable balance between progress pressure and safety target. The prevalence of a localized safety super- vision pattern among SC supervisors also implies the attendance of compromise strategies. Two participants from the client side are quoted as examples of compro- mise strategies: Sometimes they (contractors) need to balance the pro- gress and safety. If not a major safety concern, they will let workers go ahead … Everyone is looking at the pro- gress. Sometimes they have to give up something … (Client senior inspector) Some near-miss incidents happened because of viola- tion of safety procedures that were specified in the Method Statement and Construction Risk Assessment. However site personnel cannot wait. They skip some safety procedures. (Client engineer) The ‘acceptable’ balance is a result of bargaining between contractors and the client. For example, an interviewee from the client mentioned that: If we concern too much about safety, there will be no progress at all … As for minor problems that won’t lead to serious accidents, we will walk by even though con- tractors don’t follow our requirements. In addition, pacifying tactics were also adopted by con- tractors. Contractors chose to conform to at least the minimum standards as prescribed by safety regulations. When the client proposed higher safety requirements, contractors would take follow-up actions to meet the requirements. Taking the housekeeping and work access issues as examples, these two issues were repeat- edly mentioned by both workers and the client. These issues may not cause instant incidents or accidents, yet they have the potential to do so. However, the MC did not take appropriate actions until the client pointed out the problems. One engineer from the client commented that: Site safety issues are old issues, for instance the house- keeping. Actually the MC recognizes where the issues exist. However they would not take further action until pointed out by us during the weekly safety walk. Another engineer from the client provided a similar observation: They are not proactive in picking up safety problems. Every time we pick up problems for them. Site foremen or supervisors know the problems, due to lack of resources, there is no immediate response. Contractors’ strategic responses: avoidance Avoidance was another strategic response made by con- tractors to deal with institutional complexity. The phe- nomenon of decoupling the actual safety practices from the ‘paperwork’ of safety management systems indi- cates that avoidance strategies were adopted by con- tractors. On the one hand, a large safety management team, comprising safety manager, senior safety officer, site safety supervisor, safety training officer, site nurse, HSE trainee and site clerk, was assembled by the MC. The team’s major role was to maintain a ‘sound’ safety management system, and to ensure that safety-related works, documents, and activities meet regulatory and client’s requirements. On the other hand, the safety management practices adopted by the contractors were decoupled from the safety management systems. For instance, several problems, like insufficient work plan- ning and risk assessment, insufficient provision of safety resources, poor temporary works design, were not well managed. Several interviewees also mentioned that they actually did not see the contractor’s safety officers on site frequently despite a formal team being in place. This mismatch of formal safety approach and site safety practices implies a concealing tactic. One participant indicated that ‘when the DNV audit is on site, the safety condition is quite good; but when the audit is away, the safety condition declines.’ This also supports the appearance of concealing tactics; that is, site prac- tices are behind ‘a façade of acquiescence’ (Oliver, 1991, p. 154). The client was likely to accord legitimacy to the MC when the MC ‘shared its value’ or was ‘trustworthy’ (Suchman, 1995, p. 578). MC project leaders main- tained a very positive attitude toward improving site safety performance. Four of the six interview partici- pants from the client acknowledged the positive safety attitude of MC project leaders and their commitment to safety. The front-line supervisors of the MC pre- ferred an enforcement supervision approach which was aligned with the client’s expectation. Therefore the MC was able to buffer some minor safety issues from the client’s strict scrutiny. Conclusions The applicability of Oliver’s (1991) conceptual frame- work was explored by examining contractors’ safety management strategies in the Hong Kong construction industry through a case study. It was first found that the institutional environment of site safety in Hong Kong is Safety management 733 complex. Multiple stakeholders exert incompatible safety pressures on contractors through a variety of mechanisms, such as regulatory constraints, enforce- ment activities, normative force and cultural influence. At the project level, contractors confront competing goals of project progress that is enabled by market logic and safety performance that is demanded by multiple stakeholders. In order to deal with these incompatible demands, contractors may inevitably prioritize some interests at the expense of others. The second finding was that contractors’ safety practices are explicitly connected to the institutional environment. Under stringent safety requirements and clients’ tight project schedule demands, contractors compromise safety targets by balancing the progress and safety goals or bargaining with the client about safety standards. In order to avoid legal sanctions, the MC devotes great effort to following the formal safety management system. However, behind the façade of a legitimate safety management system, site practices take place in a localized manner where efficiency and safety are achieved with concessions. The decoupling of site practices from the formal safety management system ensures contractors a legitimate image. How- ever it indicates a suboptimal use of safety resources. Although the findings, for instance competing goals of progress and safety, are not novel to orthodox wisdom, the theoretical point of this study departs from the previous ones that emphasize how safety could be effectively ‘managed’. The problems observed in the workplace, e.g. mixed safety supervision patterns and discrepancies in workers’ safety awareness, reflect the institutional level issues. The other theoretical implication of the study points to the employment of strategy-as-practice perspective. A practice approach is helpful to enhance the under- standing of how people interpret and enact institutional demands placed on organizations as well as how they make decisions in the face of institutional constraints. Moreover, a more informed understanding of safety practices could help policymakers interpret the forces shaping current practices and thus help them to devise more appropriate rules. One practical implication from this study is a possi- ble institutional level change of safety management approach. For example, the conflicts between produc- tion and safety could be possibly solved by paying more attention to the planning process or design for safety. Institutional change may also emerge from the day-to- day work of individuals or organizations and may dif- fuse within an organization and beyond. One limitation of this study pertains to the use of unstructured data collection methods. As a conse- quence, it is not possible to provide solid evidence to support or reject the entire framework of Oliver (1991) and test the influence from the internal organizational factors, although two responses, namely compromise and avoidance, are evident in the data. It will be interesting to further examine the nature of institutional complexity (e.g. multiplicity and the degree of incompatibility) through more structured methodology and map a full array of institutional complexity playing on organizations, and to anticipate how organizations respond and whether they respond differently. References Anderson, J.M. (2007) Health and safety: matching legisla- tion and enforcement. Proceedings of the ICE: Management, Procurement and Law, 160(1), 11–15. Ashforth, B. and Gibbs, B. (1990) The double-edge of orga- nizational legitimation. Organization Science, 1(2), 177–94. Biggs, S., Davey, J.D. and Freeman, J.E. (2011) Safety lead- ers’ perceptions of safety culture in a large Australian con- struction organisation. Paper presented at the 1st International Conference in Safety and Crisis Manage- ment, EUC Cultural Centre, Nicosia, Cyprus, 24–28 June. Boxenbaum, E. and Jonsson, S. (2008) Isomorphism, diffu- sion and decoupling, in Greenwood, R., Oliver, C., Suddaby, R. and Sahlin-Andersson, K. (eds) The Sage Handbook of Organizational Institutionalism, Sage, London, pp. 78–98. Braun, V. and Clarke, V. (2006) Using thematic analysis in psychology. Qualitative Research in Psychology, 3(2), 77–101. Brunsson, N. and Jacobsson, B. (2000) A World of Standards, Oxford University Press, Oxford. Campbell, J.L. (2007) Why would corporations behave in socially responsible ways? An institutional theory of corpo- rate social responsibility. Academy of Management Review, 32(3), 946–67. Cheng, I., Ching, T. and Sing, Y.K. (2013) A roadmap for site safety in the construction of public housing in Hong Kong, in Liu, A.M.M., Rowlinson, S., Ng, T.S.T., Lu, W.W.S. and Chan, I.Y.S. (eds) Proceedings of the HKU- HKHA International Conference: Construction and Housing in the 21st Century, Hong Kong, 2–3 May, Faculty of Archi- tecture, The University of Hong Kong, Hong Kong, pp. 395–407. Clemens, E.S. and Cook, J.M. (1999) Politics and institu- tionalism: explaining durability and change. Annual Review of Sociology, 25, 441–66. Conner, K.R. (1991) A historical comparison of resource- based theory and five schools of thought within industrial organizational economics: do we have a new theory of the firm? Journal of Management, 17(1), 121–54. Construction Industry Council (2013) Functions of CIC, available at http://goo.gl/X17Wie (accessed 21 March 2014). Cox, S. and Cox, T. (1991) The structure of employee atti- tudes to safety: a European example. Work & Stress, 5(2), 93–106. 734 Ju and Rowlinson http://goo.gl/X17Wie Creswell, J.W. (2009) Research Design: Qualitative, Quantita- tive, and Mixed Methods Approaches, 3rd edn, Sage, Thou- sand Oaks, CA. Deephouse, D.L. (1999) To be different, or to be the same? It’s a question (and theory) of strategic balance. Strategic Management Journal, 20(2), 147–66. Delmas, M. and Toffel, M.W. (2008) Organizational responses to environmental demands: opening the black box. Strategic Management Journal, 29(10), 1027–55. Development Bureau (2009) Regulating action against con- tractors for occurrence of a serious incident or conviction for site safety or environmental offences. Technical Circu- lar (Works) No. 3/2009. DiMaggio, P.J. and Powell, W.W. (1983) The iron cage revis- ited: institutional isomorphism and collective rationality in organizational fields. American Sociological Review, 48(2), 147–60. Dorado, S. (2005) Institutional entrepreneurship, partaking and convening. Organization Studies, 26(3), 385–414. Fang, D.P. and Wu, H.J. (2013) Development of a safety cul- ture interaction (SCI) model for construction projects. Safety Science, 57, 138–49. Fennell, M.L. and Alexander, J.A. (1987) Organizational boundary spanning in institutionalized environments. Academy of Management Journal, 30(3), 456–76. Fernández, E., Montes, J.M. and Vázquez, C.J. (2000) Typology and strategic analysis of intangible resources: a resource-based approach. Technovation, 20, 81–92. Fernández-Muñiz, B., Montes-Peón, J.M. and Vázquez- Ordás, C.J. (2009) Relation between occupational safety management and firm performance. Safety Science, 47, 980–91. Fiss, P.C. and Zajac, E. (2006) The symbolic management of strategic change: sensegiving via framing and decoupling. Academy of Management Journal, 49(6), 1173–93. Friedland, R. and Alford, R.R. (1991) Bringing society back. In: symbols, practices, and institutional contradictions, in Powell, W.W. and DiMaggio, P.J. (eds) The New Institu- tionalism in Organizational Analysis, University of Chicago Press, Chicago, pp. 232–63. Gainet, C. (2010) Exploring the impact of legal systems and financial structure on corporate responsibility. Journal of Business Ethics, 95, 195–222. Greenwood, R., Díaz, A.M., Li, S.X. and Lorente, J.C. (2010) The multiplicity of institutional logics and the het- erogeneity of organizational responses. Organization Sci- ence, 21(2), 521–39. Greenwood, R., Raynard, M., Kodeih, F., Micelotta, E.R. and Lounsbury, M. (2011) Institutional complexity and organizational responses. Academy of Management Annuals, 5(1), 317–71. Guldenmund, F.W. (2000) The nature of safety culture: a review of theory and research. Safety Science, 34, 215–57. Guldenmund, F.W. (2010) (Mis)understanding safety cul- ture and its relationship to safety management. Risk Anal- ysis, 30(10), 1466–80. Hale, A. (2000) Culture’s confusions. Safety Science, 34, 1–14. Health and Safety Commission, Advisory Committee on the Safety of Nuclear Installations (1993) ACSNI Study Group on Human Factors. Third Report, Organising for Safety, HSC Books, London. Heugens, P.M.A.R. and Lander, M.W. (2009) Structure! Agency! (and other quarrels): a meta-analysis of institu- tional theories of organization. Academy of Management Journal, 52(1), 61–85. Hinze, J.W. (2000) Incurring the costs of injuries versus investing in safety, in Coble, R.J., Hinze, J.W. and Haupt, T.C. (eds) Construction Safety and Health Management, Prentice-Hall, Inc., Upper Saddle River, NJ, pp. 23-42. Hoffman, A.J. (1999) Institutional evolution and change: environmentalism and the US chemical industry. Academy of Management Journal, 42(4), 351–71. Hong Kong Construction Association (2013) Hong Kong’s Construction Industry Vision 2020, available at http:// www.hkca.com.hk/index.php (accessed 21 March 2014). Hong Kong Institution of Engineers (2013) Professional Accreditation Handbook for Engineering Higher Diploma Pro- grammes, available at http://goo.gl/3qZ4gu (accessed 21 March 2014). Honig, B. and Karlsson, T. (2004) Institutional forces and the written business plan. Journal of Management, 30(1), 29–48. Jarzabkowski, P. (2004) Strategy as practice: recursiveness, adaptation, and practices-in-use. Organization Studies, 25 (4), 529–60. Jarzabkowski, P., Balogun, J. and Seidl, D. (2007) Strategiz- ing: the challenges of a practice perspective. Human Rela- tions, 60(1), 5–27. Koch, C. (2013) From crew to country? Local and national construction safety cultures in Denmark. Construction Man- agement and Economics, 31(6), 691–703. Kraatz, M.S. and Block, E.S. (2008) Organizational implica- tions of institutional pluralism, in Greenwood, R., Oliver, C., Suddaby, R. and Sahlin-Andersson, K. (eds) The Sage Handbook of Organizational Institutionalism, Sage, London, pp. 243–75. Labour Department (2011) Legislative Council Panel on Man- power Hong Kong’s Safety Performance in the First Half of 2011 and Regulatory Framework for Construction Sector, Labour Department, HKSAR. Labour Department (2013) Legislative Council Panel on Man- power Hong Kong Measures to Enhance the Work Safety Per- formance in the Construction Industry, Labour Department, HKSAR. March, J.G. and Olsen, J.P. (1989) Rediscovering Institutions: The Organizational Basis of Politics, The Free Press, New York. Meyer, J.W. and Rowan, B. (1977) Institutionalized organi- zations: formal structure as myth and ceremony. American Journal of Sociology, 83(2), 340–63. Ng, S.T., Tang, Z. and Palaneeswaran, E. (2009) Factors contributing to the success of equipment-incentive subcon- tractors in construction. International Journal of Project Management, 27(7), 736–44. North, D.C. (1990) Institutions, Institutional Change, and Economic Performance, Cambridge University Press, Cambridge. Safety management 735 http://www.hkca.com.hk/index.php http://www.hkca.com.hk/index.php http://goo.gl/3qZ4gu Occupational Safety and Health Council (2013) Visions and Missions, available at http://www.oshc.org.hk/ (accessed 21 March 2014). Oliver, C. (1991) Strategic responses to institutional pro- cesses. Academy of Management Review, 16(1), 145–79. Oliver, C. (1997a) Sustainable competitive advantage: com- bining institutional and resource-based views. Strategic Management Journal, 18(9), 697–713. Oliver, C. (1997b) The influence of institutional and task environment relationships on organizational performance: the Canadian construction industry. Journal of Management Studies, 34(1), 99–124. Pache, A.-C. and Santos, F. (2010) When worlds collide: the internal dynamics of organizational responses to conflicting institutional demands. Academy of Management Review, 35 (3), 455–76. Pfeffer, J. and Salancik, G.R. (1978) The External Control of Organisations: A Resource Dependence Perspective, Harper & Row, New York. Powell, W.W. and DiMaggio, P.J. (1991) The New Institution- alism in Organizational Analysis, University of Chicago Press, Chicago. Rowlinson, S. (2003) Hong Kong Construction: Safety Management and the Law, 2nd edn, Sweet & Maxwell Asia, Hong Kong. Schein, E. (1990) Organizational culture. American Psycholo- gist, 45(2), 109–19. Scott, W.R. (2008) Institutions and Organizations: Ideas and Interests, Sage, Thousand Oaks, CA. Seo, M. and Creed, W.E.D. (2002) Institutional contradic- tions, praxis, and institutional change: a dialectical per- spective. Academy of Management Review, 27(2), 222–47. Sherratt, F., Farrell, P. and Noble, R. (2013) UK construc- tion site safety: discourses of enforcement and engagement. Construction Management and Economics, 31(6), 623–35. Short, J.L. and Toffel, M.W. (2010) Making self-regulation more than merely symbolic: the critical role of the legal environment. Administrative Science Quarterly, 55(3), 361– 96. Smallman, C. and John, G. (2001) British directors perspec- tives on the impact of health and safety on corporate per- formance. Safety Science, 38(3), 227–39. Suchman, M.C. (1995) Managing legitimacy: strategic and institutional approaches. Academy of Management Review, 20(3), 571–610. Suddaby, R., Seidl, D. and Lé, J.K. (2013) Strategy-as-prac- tice meets neo-institutional theory. Strategic Organization, 11(3), 329–44. van Maanen, J. and Barley, S. (1984) Occupational commu- nities: culture and control in organisations, in Staw, B. and Cummings, M. (eds) Research in Organizational Behaviour, JAI Press, Greenwich, CT, pp. 287–365. Westphal, J.D. and Zajac, E.J. (2001) Decoupling policy from practice: the case of stock repurchase programs. Adminis- trative Science Quarterly, 46(2), 202–28. Whittington, R. (1992) Putting Giddens into action: social systems and managerial agency. Journal of Management Studies, 29(6), 693–712. Wilson, J. and Koehn, E. (2000) Safety management: prob- lems encountered and recommended solutions. Journal of Construction Engineering and Management, 126(1), 77–9. Yin, R.K. (2003) Case Study Research: Design and Methods, 3rd edn, Sage, New Delhi. Zohar, D. and Luria, G. (2003) The use of supervisory prac- tices as leverage to improve safety behavior: a cross-level intervention model. Journal of Safety Research, 34(5), 567–77. Zucker, L.G. (1977) The role of institutionalization in cul- tural persistence. American Sociological Review, 42(5), 726–43. Zwetsloot, G.I.J.M., Aaltonen, M., Wybo, J.-L., Saari, J., Kines, P. and Beeck, R.O.D. (2013) The case for research into the zero accident vision. Safety Science, 58, 41–8. 736 Ju and Rowlinson http://www.oshc.org.hk/ Copyright of Construction Management & Economics is the property of Routledge and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. Abstract Introduction Theoretical background Resource dependence perspective Institutional perspective Linkages between the two perspectives and construction safety research The Hong Kong context Legal environment Normative institutional environment Cultural-cognitive institutional environment The market environment Contractors` strategic responses Compromising Avoidance Research method Data collection Data analysis Case study Project background Findings Competing goals of progress and safety Mixed safety supervision patterns Discrepancies in workers` safety awareness Contractors` strategic responses: compromise Contractors` strategic responses: avoidance Conclusions References An operational model of safety management References Nenonen, S. (2013). An operational model of safety management for service providers in manufacturing industry.Service Industries Journal, 33(1), 99-114. doi:10.1080/02642069.2011.600442 Abstract: Service providers encounter problems in trying to meet safety requirements set by legislation and multiple customers. This article introduces an operational model for service providers to promote and ease the management of safety. The model guides providers organising safety management within their own company and at customers' worksites, and encourages utilising good safety performance as an asset. The operational model is constructed by paying attention to the relevant legislation, encountered safety management problems, contributory factors of accidents, and the needs of organisations operating in manufacturing. According to the user reviews, the model's novel way of combining theoretical information, legal requirements, practical examples and tools, and a particular service provider viewpoint is a beneficial approach to the introduction of safety management. [ABSTRACT FROM PUBLISHER] An operational model of safety management An operational model of safety management for service providers in manufacturing industry Sanna Nenonen∗ Industrial Management, Tampere University of Technology, PO Box 541, Tampere 33101, Finland (Received 3 February 2011; final version received 21 June 2011) Service providers encounter problems in trying to meet safety requirements set by legislation and multiple customers. This article introduces an operational model for service providers to promote and ease the management of safety. The model guides providers organising safety management within their own company and at customers’ worksites, and encourages utilising good safety performance as an asset. The operational model is constructed by paying attention to the relevant legislation, encountered safety management problems, contributory factors of accidents, and the needs of organisations operating in manufacturing. According to the user reviews, the model’s novel way of combining theoretical information, legal requirements, practical examples and tools, and a particular service provider viewpoint is a beneficial approach to the introduction of safety management. Keywords: safety management; manufacturing operations; legal requirements; performance evaluation Introduction Outsourcing and safety Recent rapid conversion to outsourcing has increasingly transferred operations previously carried out in-house to being executed by external service providers. The utilisation of service providers is common, particularly in the manufacturing industry (Ali-Yrkkö, 2007; Eurostat, 2009). Due to outsourcing, employees of the service providers operate alongside the personnel of the customer company. In these types of multi-employer work- places, the operations carried out by one provider may influence not only the customer com- pany’s own employees’ safety but also the safety of other people who are in the workplace. Therefore, all employees operating in the same workplace, regardless of whether they rep- resent the customer or the service provider, are obligated to do their part to take care that their performance does not endanger the safety of any other employees (Council Directive 89/391/EEC, 1989; Finnish Occupational Safety and Health Act 2002/738, 2002). In addition to legal obligations, customers set requirements for the safety performance of pro- viders. The possible negative economic, ethical and image effects resulting from poor safety performance of providers have increased customer interest in provider safety management (European Agency for Safety and Health at Work, 2000; Zimmerman, 2005). In addition, providers often are expected to prove the consideration of safety in their performance by means of safety indicators and established practices, or even by certified safety management systems (Sulasalmi, Latva-Ranta, & Ylijoutsijärvi, 2003). ISSN 0264-2069 print/ISSN 1743-9507 online # 2013 Taylor & Francis http://dx.doi.org/10.1080/02642069.2011.600442 http://www.tandfonline.com ∗Email: sanna.nenonen@tut.fi The Service Industries Journal Vol. 33, No. 1, January 2013, 99 – 114 Guidelines for safety management systems provide information about the systematic management of safety that is also applicable in provider companies. However, practice has shown that small providers in particular do not have adequate resources for enforcing these instructions or they are considered as too burdensome to implement. In addition, many provider companies are stated to be unable to develop adequate safety management systems themselves (Matthews & Rowlinson, 1999). Nevertheless, practical and easy-to- use guidelines producing instructions for the management of safety in service production have not been readily available to date. Objectives of the study In an attempt to support provider safety management, the present study introduces an oper- ational model of safety management directed at service providers. The model is con- structed primarily for providers operating for manufacturing companies, but it can also be utilised in other provider companies. The operational model discusses the management of safety within the provider’s own activities and in operations performed at the custo- mer’s premises. The operational model pays particular attention to the special features of service production by taking into account the problems encountered, typical accident factors and the relevant legislation. Another aim is to review the long-term utilisation and utility of the model in the management of safety in provider companies. The review is conducted by a questionnaire discussing the utilisation of the operational model in manufacturing companies. Safety management in service production Introduction to safety management Safety management has been considered to be a key element in reducing hazards in the workplace (Frick & Wren, 2000). Safety management aims to support the systematic and continuous prevention of situations that may be harmful to people, work environment and property (Heinrich, Petersen, & Roos, 1980) through systematic managerial processes (Frick & Wren, 2000). Thus, safety management relates to those practices, roles and func- tions associated with remaining safe (Kirwan, 1998). Efficient safety management has been presented to require, for example, positive safety attitudes and culture, defining responsibilities and authorisations (Booth & Lee, 1995; Kirwan, 1998), management com- mitment to safety, involvement of employees (Mearns, Whitaker, & Flin, 2003), func- tional communication channels (Booth & Lee, 1995; Kirwan, 1998; Mearns et al., 2003), identification of risks (Booth & Lee, 1995; Frick & Wren, 2000) and integration of safety issues to everyday operations and other management decisions (Frick & Wren, 2000). Hence, implementation of systematic safety management requires well-defined prin- ciples and procedures. The activities constituting safety management in practice can be considered as a safety management system (Kirwan, 1998). The safety management system can be devised independently by the organisation realising safety management or through the general guidelines of safety management systems (Hämäläinen & Anttila, 2008). Several guidelines for managing safety are readily available, but the best-known models for the implementation of safety management are BS 8800, 2004, Occupational health and safety management systems – Guide; OHSAS 18001, 2007, Occupational health and safety management systems – Specification; and ILO-OSH, 2001, Guidelines to health and safety management systems. These guidelines, which 100 S. Nenonen provide a general viewpoint for safety management systems, follow largely similar frame- works. Safety management is guided through a review of initial status; compilation of safety policy; organisation, planning, and implementation of safety management; and auditing and review of safety performance (see BS 8800, 2007; OHSAS 18001, 2007; and ILO-OSH, 2001). On the other hand, the Safety Checklist Contractors (SCC) is a cer- tifiable safety management system particularly directed at industrial manufacturers and service providers. The SCC discusses safety management through 12 elements including such factors as safety policy, risk management, training, healthcare, and investigation of accidents (SSVV – Centraal College van Deskundigen, 2008). The previously mentioned safety management guidelines point out elements that require attention in safety manage- ment, but the discussion can be considered quite general. Furthermore, it has been stated that implementation of safety management according to the available guidelines is burden- some (Makin & Winder, 2008) and not that straightforward (Mitchison & Papadakis, 1999), particularly for smaller organisations which are common among service provider companies. The legislation does not necessitate the adoption of a safety management system, but sets requirements for several fields of safety management (Council Directive 89/391/EEC, 1989; Finnish Occupational Safety and Health Act 2002/738, 2002). These kinds of requirements are, for example, an obligation of identification and control of hazards, and for organising induction training of employees. Therefore, in order to respond to the requirements of legislation, organisations are required to implement safety manage- ment covering all the operations on the worksite. (Hämäläinen & Anttila, 2008). Safety management problems regarding outsourcing Companies outsource their operations to service providers for various reasons. Organis- ations may be keen to focus on their core operations (Abdel-Malek, Kullpattaranirun, & Nanthavanij, 2005; Kakabadse & Kakabadse, 2002; McIvor, 2005), allocate available resources more effectively (Arditi & Chotibhongs, 2005), obtain special skills not found within their own organisation (Kakabadse & Kakabadse, 2002), or acquire cost savings (Kakabadse & Kakabadse, 2002; Kremic, Tukel, & Rom, 2006), among others factors. At its best, outsourcing can also be beneficial from a safety point of view (Walters & James, 2009). For example, service providers may have more competent personnel and specialised equipment for certain work tasks (Arditi & Chotibhongs, 2005). However, effi- cient safety management in multi-employer worksites requires particular consideration, such as utilisation of safety-oriented service providers (Holmes, Lingard, Yesilyurt, & De Munk, 1999; Shafer, 2008), organising common orientations and safety trainings between customer and provider employees (Downey, 1995; Mynttinen, 2006), ensuring effective communication between different parties sharing the worksite (Hinze & Raboud, 1998; Mynttinen, 2006), and effective planning and scheduling of different per- formers’ work tasks (Shafer, 2008). Despite the possible benefits for safety, many studies have shown that outsourcing typically has adverse effects on the safety of employees operating at worksites shared by multiple employers (e.g. Blank, Andersson, Lindén, & Nilsson, 1995; Mayhew & Quinlan, 1999; Rousseau & Libuser, 1997). The negative influence of the utilisation of an external workforce has been presented as being related to complex and dynamic net- works resulting in changes in work organisations, such as changes in relationships (Mayhew & Quinlan, 1999), inappropriate flow of information (Beale, 2003), fragmenta- tion of management (Mayhew, Quinlan, & Ferris, 1997), and poor organisation and timing The Service Industries Journal 101 of work tasks (Mayhew & Quinlan, 1999). The probability of problems increases when the structure of the operator network becomes more complicated (Beale, 2003). Service providers have been found to have a high impact on safety (Langford, Rowlinson, & Sawacha, 2000; Toole, 2002) and, therefore, many providers play a critical role in the management of safety at multi-employer worksites (Beale, 2003). However, service providers in particular have been reported to be likely to encounter problems in the organisation and in the implementation of their safety management (Matthews & Rowlinson, 1999; Wilson & Koehn, 2000). The special features of service production, for example, operating with several customers who have different working cultures and practices and in varied and sometimes unfamiliar worksites and locations, pose problems and risks that are different from conventional industrial operations (Lind, Nenonen, & Luoto, 2006). Thus, the management of safety is particularly complex within work tasks in which the work environment, the working conditions, and thus also the risks change continuously (Välimaa, Varonen, Lappalainen, & Ketola, 2001). The constantly changing work features result in there being insufficient time for assessing risks, determin- ing training needs, and setting up, applying, and monitoring the effectiveness of safety measures (Papadopoulos, Georgiadou, Papazoglou, & Michaliou, 2010). Problems may arise in provider companies regarding the management of safety, for example due to ten- dering systems neglecting the safety costs (Hinze & Raboud, 1998; Langford et al., 2000), limitations in providers’ resources allocated for safety (Holmes et al., 1999; Mayhew & Quinlan, 1999; Mayhew et al., 1997), the complexity of ensuring adequate flow of infor- mation with the customer and other providers (Mynttinen, 2006; Väyrynen, 2003), the dis- tribution of responsibilities between different parties (Loosemore, Dainty, & Lingard, 2007; Toole, 2002), and deficiencies in providers’ worksite-specific hazard identification (Mynttinen, 2006; Trethewy, Atkinson, & Falls, 2003). In addition, the complexity of managing safety in the provider companies increases because the safety measures provi- ders implement need not only meet legal requirements but also those set by the customer companies (Wilson & Koehn, 2000). Compilation and review of the operational model Development process The responsibility for the development of the operational model was with the Tampere University of Technology research team. The development was implemented in close cooperation with some Finnish companies that provide services for organisations operat- ing in the manufacturing industry. Eight voluntary provider companies aiming to enhance their safety management participated in the model development process by commenting and steering work realised by the research team. These companies supplied maintenance and repair services, machinery manufacture, installation, and modification, property main- tenance, cleaning, industrial sanitation, security, up-keeping, and information system ser- vices. The Finnish Maintenance Society, Promaint, an association promoting Finnish maintenance, also took part in the process. The cooperating companies employed from a few dozen to several thousand personnel. Some of them operated mainly with Finnish companies but some of them were international operators. In addition to the provider com- panies, one customer of each participating provider was able to take part in the develop- ment process. Two of the participating companies had the role of both service provider and customer in the development process. The development process of the operational model was highly flexible and both the structure and the content of the model evolved continuously to respond to the needs 102 S. Nenonen emerging during the process. However, the development process can be roughly con- sidered to consist of five phases, even though the borders of the different phases are only vague due to the adaptability of the process (Figure 1). First, the requirements for the model were defined in cooperation with the service providers and the manufacturing companies. Second, the structure and the critical fields of safety management to be built into the model were designed and finalised with the participating companies. Third, the content of the model was produced by utilising the information obtained, among others, from literature, legislation, interviews, guidebooks, and standards. Fourth, the developed operational model was field-tested in the participating companies, and fifth, on the basis of the test experiences, the model was revised into the final form. More detailed descriptions of the different phases are presented in the following sections. Requirement specification Preliminary requirements for the operational model were defined by the research team when preparing the project proposal. These requirements were based on a literature review and results of previous research projects conducted on the topic. These require- ments were revised and supplemented according to the suggestions of the participating companies to form the final set of requirements. The opinions of the companies regarding the requirements were obtained by interviews (Nenonen & Vasara, in press). The main idea was to develop information material which supports the implementation of safety management at organisations providing services for the manufacturing industry and that promotes safety-related cooperation with the customer companies. The material to be developed was required to focus on the challenges in the management of safety while taking into account the work tasks performed by different operators and enabling the development of uniform practices in different units of the organisation. It should also take into consideration the relevant legislation and guide the management of safety when operated with multiple operators in varying worksites, all having their own working methods and cultures. The operational model was required to be easy to use, prac- tical, applicable at different worksites, and to be adaptable to the framework of common safety management systems. In addition, the model was hoped to be suitable for both large and smaller companies and to include both wide-ranging information on the implemen- tation of safety management and a compact listing of the issues needing to be taken into account in everyday operations. The specific themes that the operational model was hoped to cover were risk management, responsibilities, competence, communication, dangerous work tasks, licences, insurance, working abroad, and foreign employees. Conceptual design The requirements and wishes presented by the participating companies were utilised in designing the first draft of the structure and content themes of the operational model. Figure 1. Development process of the operational model. The Service Industries Journal 103 This conceptual design realised by the research team was introduced to the companies and, taking into account their comments, the researchers developed the design further to meet the emerging needs. The final concept, on which the actual operational model construction was based, was defined to adapt the principle of continuous development and consider operations per- formed at customers’ sites. It was decided that the structure of the operational model would consist of general safety objectives, a review of the current state of safety, planning and implementation of safety of provider company’s own performance and when operat- ing at the customer’s site, and evaluation of safety performance. The designed concept was constructed to take into account the main problems of managing safety in provider com- panies operating in the manufacturing industry. As the themes to be built into the model, 20 factors were defined which were considered to be challenging to manage and/or were reported to contribute to accidents in the manufacturing industry. These factors were found out by interviews carried out in the participating companies and by a review of accident statistics covering accidents which occurred in Finnish manufacturing companies (Nenonen, 2011; Nenonen & Vasara, in press). Legislation related to the problematic factors was also included in the design. Factors included were recruitment and professional skills, induction, occupational instruction and guidance, safety training, safety instructions and operating instructions, flow of information and cooperation, protection and safety devices, hazard identification and risk management, work tasks needing special attention, licenses, emergency preparations, aberrations and accidents, occupational health care, documentation, inspections and repairs, contracting, operating abroad, foreign employees, temporary work, and insurance. Content provision The content provision and realisation of the concept into the complete operational model were carried out in consecutive stages of content production realised by the research team, as well as commenting by the participating organisations. In these stages, a version con- sisting of the partial contents of the model was delivered to the participating companies for their comments. Using the feedback, the content of the model was developed and sup- plemented for the next version. The researchers produced the content of the operational model by utilising information obtained mainly from the literature, legislation, company interviews, a questionnaire, guidebooks, and standards. The main fields of the information gathered were relevant paragraphs of law, the integration of safety in service production, the theory of implemen- tation of safety management in general and within problematic fields, the management of safety in shared workplaces, good practices of implementation of safety measures, and tools (e.g. checklists and work permit forms) for ensuring safety in everyday operations. Pilot testing The utility and functionality of the operational model and its compliance with the require- ments were examined by pilot testing carried out at the participating companies. The testing lasted 4 months and was executed in two phases. At first, the companies tested by themselves the partially constructed operational model, which included about two- thirds of the material to be included in the model. The model was then supplemented to contain the rest of the content and was delivered to the companies halfway through the 104 S. Nenonen testing period. The experiences of using the model were recorded during interviews carried out in the piloting companies. The piloting companies considered the operational model to be a very workable instruction set. They particularly assessed the legislation fields as important and different tools as functional. Some of the companies appreciated the extent of the material. Changes that the piloting companies wanted in the operational model were additional work permit forms, clarification of the structure and outline of different fields, and minor corrections to some parts of the text in order to make it more understandable and to highlight the impor- tant features. Final deployment The research team revised the operational model according to the feedback received from the piloting companies after the testing period. The main changes were the addition of extra work permit forms and changes in the representation and outline of the material in order to make the model easier to follow. Other changes made were mainly minor inser- tions into the text, updating of paragraphs of law, and correction of wordings and misspell- ings. The finished operational model was delivered to the participating companies for their approval before its final release. The model was published as a web document available for all companies needing the information in question (Nenonen, Vasara, & Litmanen, 2008). Review of long-term utility The long-term utility of the operational model was evaluated using a questionnaire carried out 17 months after the official release of the model. The aim of the questionnaire was to capture the opinions of the model users on the success and applicability of the model. The adequacy of the content was also reviewed. The questionnaire was sent to the representa- tives of those companies who had participated in the operational model development process. The questionnaire included questions concerning the extent of the use of the different parts of the model, the situations in which the model was used and who was using it, the benefits obtained from using the model, the coverage of the information presented, and the compliance with the requirements. Most of the questions were closed with predetermined answer alternatives but there were also a couple of open questions seeking more detailed information about the issues discussed in the closed questions. More detailed information about the questions and answer alternatives can be found in the results section. The review was carried out as a web survey. The e-mail inviting responses to the ques- tionnaire was sent to the representatives of 14 manufacturing industry companies of which six operated as service providers, six as customers, and two as both providers and custo- mers. From these, representatives of three companies (one provider, one customer, and one provider – customer) were not reachable because their e-mail addresses were no longer active due to changes in employment. In total, seven responses were received of which five were from representatives of provider companies and two from those working for cus- tomer companies. Thus, the respondents of the questionnaire covered half of the compa- nies that took part in the development process and those respondents reachable represented 83% of the service providers and 33% of the customers. The data were processed using the statistical program SPSS 16.0 and summarised by using descriptive statistics. The Service Industries Journal 105 Operational model of safety management for service providers Structure and content of the operational model According to the requirements made for the operational model, the structure of the model follows the frameworks and the principles of continuous improvement of common man- agement guidelines (see e.g. BS 8800, 2004; ISO 9001, 2008; ISO 14001, 2004; OHSAS 18001, 2007). The operational model is divided into three parts (Figure 2). The first part of the model discusses the definition of the safety objectives and a review of the present state of safety management in the companies under consideration. The second part of the model embodies the planning and implementation of safety manage- ment procedures inside the companies and at the worksites. In the third part, safety per- formance is evaluated and improvements needed are executed. After the first iteration of the phases, the safety performance of the company is continuously improved by repeat- ing the different phases. In addition to these three parts, the operational model contains forms discussing subjects reviewed in parts I and II. The contents of the different parts are elaborated in the following sections. Part I – objectives and present situation The first part of the operational model encourages companies to review their safety objec- tives as well as their customers’ needs of safety performance. The organisation’s own safety objectives are defined to address the safety level that is aspired to be reached (e.g. a zero accident goal or a certain number of incident notifications). Secondly, on Figure 2. The structure of the safety management operational model. 106 S. Nenonen the basis of the customer needs review, providers can take into account the typical custo- mer needs and requirements relating to the safety aspects of activities and the organis- ation’s safety level (e.g. to go below a given accident frequency or to undertake certain safety training). This way, the provider companies can plan their own performance to fulfil both the organisation’s internal requirements and the customers’ needs and require- ments. In addition, the operational model encourages organisations to formulate their objectives as a safety policy in accord with common safety management systems and to follow a path that achieves the objectives set. In the first part of the model, instructions for organisations to assess the status of their own operations and their safety levels are also produced. During the initial status review, the operations performed for customers, which are core operations (e.g. maintenance and repair, property maintenance, and industrial sanitation) as well as supporting (e.g. customer service, safety training, and quality control) and enabling operations (e.g. sales services, logistics and information management) required for the implementation of the core oper- ations are recorded. The purpose of this review is to identify all the existing operations in order to observe their critical factors for ensuring safety (e.g. sufficiency of the employees’ professional skills and suitability of the tools utilised). As a result of the review, organis- ations obtain an insight into the spectrum of operations required for the organisation’s activities and can identify the problematic factors of the different operations in the light of safety. The review assists organisations to arrange their safety performance by paying attention to all their fields of activities and their safety critical factors. Part II – implementation of safety management The second part of the model conducts the implementation of safety management. The subject is approached through 20 critical fields of multi-employer worksites’ safety from recruitment and professional skills to acquiring insurance (Figure 2). The implemen- tation of these critical fields is discussed in the light of legislation and the organisation of safety operations in provider companies as well as at the varying worksites. The paragraphs of law presented describe the legislative requirements, setting down the basic level of safety performance. The sections of law included in the operational model concern the common legislation relating to the activities of organisations providing operations for other organisations operating in the manufacturing business. Industry- specific legislation is not included (e.g. specific law for nuclear power or the transportation industry). The actual model presents summaries of the legislative requirements, but the complete paragraphs of law can be found in the appendix of the model. The organisation of safety operations is guided by describing general information and practical examples of the implementation of safety measures, first as a part of the provi- ders’ own performance and after that at their customers’ sites. The first part guides provi- der companies in the development of the safety of the activities within their own organisation with an eye to obtaining safety performance at such a level that both their own organisation and their customers perceive it as respectable. By means of the infor- mation related to the safety issues of operating on customers’ sites, the provider companies can supplement their measures of safety management according to the requirements and special features of operating at customers’ worksites. The operational model encourages providers to work in close cooperation with their customers in the management of safety in order to be aware of their customers’ needs and wishes and thereby be able to produce safe services at varying worksites and at the same time fulfil the legal requirements. The Service Industries Journal 107 Part III – performance evaluation and development The third part concentrates on the evaluation of safety performance and on the continuous development of safety measures in the provider organisation. In this part, the subject is approached in a similar way to the previous part. The instructions about measurement and auditing of safety performance begin with a review of the relevant legislation, con- tinue to general information and practical examples and finally present methods for imple- menting the measures. Organisations can apply the information and methods of performance evaluation and development presented in the operational model to assess the effects of the implemented safety measures as well as to define the required develop- ment actions. Forms In addition to the three information parts, the operational model includes a section contain- ing forms for different purposes. Four types of forms are built into the model: summary forms, checklists for different phases of activities, permit forms, and a safety level check- list. These forms include general models for implementing and documenting safety measures. Forms are also available as electronic documents, enabling companies to adapt them to fit the special features of their activities. Summary forms for all the 20 critical fields discussed in the second part and the measurement and auditing of safety performance presented in the third part are formu- lated. Forms are compiled in accordance with the issues presented in the second and third parts and they consist of listings of common factors that organisations operating in the manufacturing industry need to take into account to ensure safety. The summary forms can be used as tools in managing safety in everyday operations. Checklists for different phases of activities consist of forms for the negotiating phase, the delivery phase, and the finishing phase. The first of these is a summary of safety-related issues that already need to be arranged when negotiating a contract before any work per- formance. The second checklist discusses issues affecting safety that require attention when actually performing work tasks at the customer’s site. The third form concludes with a listing of things that should be reviewed after the tasks have been performed or the operations at the site have been finished, enabling the discovery of any possible short- comings and the development of future performance into a safer direction. The permit form section contains forms to be filled in for the most common industrial work tasks that require permits. Forms are compiled for hot work, tank work, driving a forklift truck, lifts, and electrical installations. The safety level checklist is a listing of safety-related issues that the service provider should discuss with customers before performing work tasks. According to the checklist, the provider can show the safety level and safety measures implemented in the organis- ation. At the same time, the provider can assure that it is aware of those safety issues relat- ing to the customer company’s activities that can affect the safety of the provider company’s own employees. Long-term utility of the operational model The operational model was utilised to some extent in six of the respondent companies, of which four were providers and two customers. The model was not used in one service pro- vider due to the limited time resources available at the time for the implementation of the information presented in the model. The companies that had exploited the operational 108 S. Nenonen model had most commonly utilised the legislation and the organising parts. All the provi- der companies and other customers mentioned that they had applied the legislation part presented in the model. The information about organising safety management and the forms were also utilised more commonly in provider companies. Seventy-five percent of the providers had applied the organising information and half had applied the forms. The corresponding shares in the case of customers are 50% and 0%. The most commonly utilised field in the legislation and organising parts was protection and safety devices but also information related to recruitment and professional skills, licenses, temporary work in the case of legislation and safety instructions and operating instructions, flow of infor- mation and cooperation, aberrations and accidents, documentation, and foreign employees were commonly used. The extent of model utilisation differed between providers and cus- tomers; providers had applied on average information from 12 critical fields but customers only from five. Provider companies had also utilised the model more frequently than cus- tomers. The providers had utilised the operational model, for example to increase knowl- edge, as backup when compiling instructions and when formulating supply contracts. On the other hand, customers had applied the model in agreement negotiations and nego- tiations concerning details of work tasks. The people who mainly mentioned having utilised the operational model were those in safety organisation and at a higher level as well as customer managers, head of supplies, and service managers. The respondents representing organisations that had applied the model considered that it had been relatively valuable for the development of their organisation’s safety manage- ment and somewhat valuable for the cooperation implemented with their customers. Respondents brought out that the operational model helped to understand unfamiliar and obscure issues of safety management, it gave new ideas about the implementation of safety management in the organisation, and it pulled the scattered information together and clarified the field of safety management. One reason for the assessment of lesser use- fulness was that improvements of safety management had not had that important state in the organisation due to organisational changes that were in progress. Another aspect was that cooperation on safety issues between different parties sharing the workplace had not been common and therefore the development process according to the operational model was only just in its early stages. The users of the model perceived that the content of the model was comprehensive but still unnecessary issues were not included. Only one respondent, even though he also considered the information included in the model was extensive, mentioned that accident prevention measures and the special features of multi-employer worksites could possibly have been highlighted more. The opinions on the usefulness and adequacy of the content were similar between providers and customers. The operational model’s compliance with the requirements was obtained from those respondents who had utilised the model and therefore had a better view of it. These respon- dents considered that the model fulfilled the requirements rather well because on a 1 – 5 scale (poor – well), the model got a score of 3.8. The customers gave slightly better grades for the requirement compliance than did the providers but the difference was only a couple of tenths. The highest scores received were for the operational model’s comprehensive approach in safety (4.5), aid in the construction of uniform practices in different units of the organisation (4.3), role in the increase of knowledge on legislation requirements (4.2), adaptability according to the needs of the organisation in question (4.0), encourage- ment to the continuous development of safety (4.0), and usefulness in the development of safety management with partners (3.8). The weakest, but still positive, scores received were for the consideration of the work tasks performed by different operators (3.5) and cooperation with other partners (3.3) as well as worksite-specific adaptability (3.3). The Service Industries Journal 109 Discussion Review of the operational model The operational model presented in this article follows the previously published safety management guidelines by sharing the idea of continuous improvement and the main phases of the management process. However, it also has several factors that are not built into other guidelines. For example, if compared with the most well-known safety management guidelines having a general approach, the operational model has a specific viewpoint on provider companies. This enables concentration on and more detailed review of particular issues important to and problematic for providers. The SCC checklist has a provider approach but the scale of review is more concise and certification orientated than in the operational model developed. Another noticeable difference is the review extent and framework. The operational model combines theoretical information about important safety management issues, relevant legislation, practical examples, and forms supporting safety management in everyday operations. For example, other safety manage- ment guidelines require that companies take the legal requirements into account and adapt their performance to respond to these, but they do not provide information on the require- ments due to their general approach. Moreover, extra material supplementing the theoreti- cal information built in the operational model is not really included in other guidelines. In addition, the model encourages providers to cooperate with their customers regarding safety and to pay attention to the typical safety requirements their customers have. Thus, the operational model developed offers a new kind of information and approach for safety management compared with previous guidelines. However, it is not inconsistent with the other guidelines but instead supplements the instructions available. Providers who have utilised the previous guidelines are undoubtedly familiar with most of the infor- mation presented in the operational model, but they may benefit from the legislation summary and forms, for instance. On the other hand, companies which consider other guidelines too complex or difficult to implement may benefit from the practical approach of the operational model developed. The operational model can also be used as a practical means of implementing safety management on the way to utilising the more well-known models and certification of safety management systems. The field tests and interviews regarding the experiences indicated that the model fulfils quite well the requirements set in the requirement specification phase and fits the implementation of safety management in industrial services. Unfortunately, the organis- ations that had utilised the operational model were only in the beginning stages of the implementation process. Therefore, the effect of using the model for the organisation’s safety management could not be assessed within this study. However, the positive reviews can be considered as an implication of the model’s support for safety management in service provider organisations. Further, it is noteworthy that, even though the service production-specific factors were built into the model, the users wished that these issues could have been emphasised even more. This indicates that the service provider and pro- duction-specific approach in the safety management context would deserve more attention than it has gained to date. In addition, even though the user experiences were positive, the extent of the operational model can turn out to be a challenge for its utilisation, particularly in small companies. Adoption of the practices presented in the model is time-consuming, especially if safety management has not been considered earlier. Moreover, in short-term projects, the effective utilisation of the model requires that management of safety has already, to some extent, been executed. Another disadvantage of the model is that the legislation presented in the model is likely to be updated at some point in the future. 110 S. Nenonen Therefore, users have to follow the upcoming progress of regulations in order to ensure that the model is constantly up-to-date. Due to the potential challenges in utilisation, the adoption of the model has been facilitated by constructing a guidebook and training material including the most important issues of the model (Hyytinen, Vasara, & Nenonen, 2010; Vasara, Nenonen, & Hyytinen, 2010). These materials provide an intro- duction to the operational model and therefore ease the utilisation of model’s information, particularly in organisations with time limitations or lack of experience regarding safety management. Exploitation of the operational model The operational model of safety management was developed primarily for the use of service providers supplying operations for companies operating in the manufacturing industry. These kinds of providers are, for example, companies delivering maintenance, installation, and cleaning services. The main purpose of the model is to promote and ease the management of safety and the organisation of safety-related cooperation with the customers. Even though the operational model has been compiled from the viewpoint of provider companies and their needs and safety management problems, it can be utilised to a great extent in organisations providing other kinds of services and for companies oper- ating in lines of business other than the manufacturing industry. The results of the user survey show that parts of the model can also be applied in companies acquiring services from providers; for example, in contract negotiations or ensuring safety in cooperation with their providers. The significance of safety in service production is emphasised because in addition to the service provider companies, the customers’ organisations are also dependent on the safety of the service providers’ performance (Vassie & Fuller, 2003). However, providers rarely exploit the high quality of their safety performance as part of their sales appeal when marketing their services even though many customers consider a good safety level as a sign of quality in overall performance. Some customers may select their providers mainly, for example, on the basis of price (Langford et al., 2000) or availability (Lingard & Rowlinson, 2005), but some customers do not even accept tenders from com- panies whose safety performance does not reach a certain safety level (Lappalainen, Sauni, Piispanen, & Nurmi, 2005). Thus, the effective management of safety has been reported to constitute a sustainable competitive advantage, particularly if the client values safety. Even though the client would not put a high value on safety, they still appreciate, for example, working according to the schedule and budget, which also are affected by success in safety (Rechenthin, 2004). Hence, it is worthwhile for service providers to talk about their management of safety to their customers. This way, investments in safety can be turned to advantage in service tendering. By the means of the operational model developed, providers can show how they have implemented the management of safety in their organisation and how safety performance has been planned to be developed. If a provider can show that they operate according to the operational model, customers can see the provider’s eagerness to operate according to safe practices and they can believe that they will receive smooth, trouble-free, and scheduled services. Further needs The discussion regarding safety management in multi-employer worksites has concen- trated mainly on the customers’ possibilities to ensure proper safety management at the The Service Industries Journal 111 worksite, such as applying safety level as one criterion in provider selection or setting in requirements for a service provider’s safety performance. Only a minor part of multi- employer safety management research or guidelines have concentrated on the provider’s viewpoint. However, service providers have been considered to be not only significant contributors to the multi-employer worksite safety but also to those encountering signifi- cant problems in managing the safety of their operations. This study shows that service providers consider provision of service production-oriented safety management instruc- tion material as a beneficial support for their safety work. However, these types of service production-oriented materials have not been readily available to date. Therefore, this study provides a first approach to this topic. Service providers are widely utilised not only in industry but also in other lines of business, and the problems of provider safety management concern a great number of operators working at multi-employer work- sites. Therefore, it would be worthwhile to review the needs of service provider in more detail and to provide novel practices in order to respond to the emerging needs and to support providers’ attempts to improve their safety performance. References Abdel-Malek, L., Kullpattaranirun, T., & Nanthavanij, S. (2005). A framework for comparing out- sourcing strategies in multi-layered supply chains. International Journal of Production Economics, 97(3), 318–328. Ali-Yrkkö, J. (2007). Outsourcing in Finnish manufacturing – does industry matter (In Finnish, English abstract). Discussion Papers No. 1070. Helsinki: ETLA, The Research Institute of Finnish Economy. Retrieved November 8, 2010, from http://www.etla.fi/files/1715_ Dp1070 . Arditi, D., & Chotibhongs, R. (2005). Issues in subcontracting practice. Journal of Construction Engineering and Management, 31(8), 866–876. Beale, C.J. (2003). Factors influencing the safe management of contractors on major hazard installa- tions. Hazards XVII: Process Safety – Fulfilling Our Responsibilities (IChemE Symposium Series No. 149). (pp. 719 – 731). Rugby: Institution of Chemical Engineers. Blank, V.L.G., Andersson, R., Lindén, A., & Nilsson, C. (1995). Hidden accident rates and patterns in the Swedish mining industry due to involvement of contractor workers. Safety Science, 21(1), 23–35. Booth, R.T., & Lee, T.R. (1995). The role of human factors and safety culture in safety management. Proceedings of the Institution of Mechanical Engineers 1847 – 1996, 209(B5), 393–400. BS 8800. (2004). Occupational health and safety management systems – guide. London: British Standards Institution. Council Directive 89/391/EEC. (1989). Council directive on the introduction of measures to encou- rage improvements in the safety and health of workers at work. Luxembourg: The Council of the European Communities. Downey, J.M. (1995). Risks of outsourcing – applying risk management techniques to staffing methods. Facilities, 13(9/10), 38–44. European Agency for Safety and Health at Work. (2000). Occupational safety and health in market- ing and procurement. (Report Issue 304). Luxembourg: Office for Official Publications of the European Communities. Eurostat. (2009). International sourcing statistics. Retrieved November 8, 2010, from http://epp. eurostat.ec.europa.eu/statistics_explained/index.php/International_sourcing_statistics. Finnish Occupational Safety and Health Act 2002/738. (2002). Unofficial translation, section 49. Ministry of Social Affairs and Health. Retrieved July 10, 2011, from www.finlex.fi. Frick, K., & Wren, J. (2000). Reviewing occupational health and safety management – multiple roots, diverse perspectives and ambiguous outcomes. In K. Frick, P.L. Jensen, M. Quinlan, & T. Wilthagen (Eds.), Systematic occupational health and safety management perspectives on an international development (pp. 17–42). Amsterdam: Pergamon. Hämäläinen, P., & Anttila, S. (2008). Onnistuneen työterveys- ja työturvallisuusjohtamisen sisältö ja käytännöt (In Finnish). Tampere: Työsuojeluhallinto. 112 S. Nenonen http://www.etla.fi/files/1715_Dp1070 http://www.etla.fi/files/1715_Dp1070 http://epp.eurostat.ec.europa.eu/statistics_explained/index.php/International_sourcing_statistics http://epp.eurostat.ec.europa.eu/statistics_explained/index.php/International_sourcing_statistics Heinrich, H.W., Petersen, D., & Roos, N. (1980). Industrial accident prevention. New York, NY: McGraw-Hill. Hinze, J., & Raboud, P. (1998). Safety on large building constructions projects. Journal of Construction Engineering and Management, 114(2), 286–293. Holmes, N., Lingard, H., Yesilyurt, Z., & De Munk, F. (1999). An exploratory study of meanings of risk control for long term and acute effect occupational health and safety risks in small business construction firms. Journal of Safety Research, 30(4), 251–261. Hyytinen, T., Vasara, J., & Nenonen, S. (2010). Turvallisuusjohtamisen koulutusmateriaali, Teollisuusyrityksen ja toimittajien vÄlinen yhteistyö (In Finnish). Tampere: Tampere University of Technology. Retrieved October 15, 2010, from http://turva.me.tut.fi/tutkimus/ Turvallisuusjohtamismateriaalit/Tulostusversio050710 ILO-OSH. (2001). Guidelines to health and safety management systems. Geneva: International Labour Organization. ISO 9001. (2008). Quality management systems – requirements. International Organization for Standardization. ISO 14001. (2004). Environmental management systems – Requirements with guidance for use. International Organization for Standardization. Kakabadse, A., & Kakabadse, N. (2002). Trends in outsourcing: Contrasting USA and Europe. European Management Journal, 20(2), 189–198. Kirwan, B. (1998). Safety management assessment and task analysis – a missing link? In A. Hale & M. Baram (Eds.), Safety management. The challenge of change (pp. 67–92). Amsterdam: Pergamon. Kremic, T., Tukel, O.I., & Rom, W.O. (2006). Outsourcing decision support: A survey of benefits, risks, and decision factors. Supply Chain Management: An International Journal, 11(6), 467–482. Langford, D., Rowlinson, S., & Sawacha, E. (2000). Safety behaviour and safety management: Its influence on the attitudes of workers in the UK construction industry. Engineering, Construction and Architectural Management, 7(2), 133–140. Lappalainen, J., Sauni, S., Piispanen, P., & Nurmi, S. (2005). Good safety management on combined construction sites. Finnish Institute of Occupational Health. Retrieved May 20, 2008, from www.ttl.fi. Lind, S., Nenonen, S., & Luoto, M. (2006). Huolto- ja kunnossapitotöiden riskienhallinta (In Finnish) (Project Report). Tampere: Tampere University of Technology, Institute of Occupational Safety Engineering. Retrieved November 8, 2010, from http://turva.me.tut.fi/tutkimus/Huolto%20ja% 20kunnossapitotoiden%20riskienhallinta/HuoltoLoppuraportti0307062 . Lingard, H., & Rowlinson, S. (2005). Occupational health and safety in construction project man- agement. New York: Spon Press. Loosemore, M., Dainty, A., & Lingard, H. (2007). Human resource management in construction projects – strategic and operational approaches. London: Taylor and Francis. Cited in Loosemore? M., & Andonakis, A. (2007). Barriers to implementing OHS reforms – the experiences of small subcontractors in the Australian Construction Industry. International Journal of Project Management, 25(6), 579 – 588. Makin, A., & Winder, C. (2008). A new conceptual framework to improve the application of occu- pational health and safety management systems. Safety Science, 46(6), 935–948. Matthews, J., & Rowlinson, S. (1999). Partnering: Incorporating safety management. Engineering, Construction and Architectural Management, 6(4), 347–357. Mayhew, C., & Quinlan, M. (1999). The effects of outsourcing on occupational health and safety: A comparative study of factory-based workers and outworkers in the Australian clothing industry. International Journal of Health Services, 29(1), 83–107. Mayhew, C., Quinlan, M., & Ferris, R. (1997). The effects of subcontracting/outsourcing on occu- pational health and safety: Survey evidence from four Australian industries. Safety Science, 25(1 – 3), 163–178. McIvor, R. (2005). The outsourcing process – strategies for evaluation and management. New York, NY: Cambridge University Press. Mearns, K., Whitaker, S.M., & Flin, R. (2003). Safety climate, safety management practice and safety performance in offshore environments. Safety Science, 41(8), 641–680. The Service Industries Journal 113 http://turva.me.tut.fi/tutkimus/Turvallisuusjohtamismateriaalit/Tulostusversio050710 http://turva.me.tut.fi/tutkimus/Turvallisuusjohtamismateriaalit/Tulostusversio050710 http://turva.me.tut.fi/tutkimus/Huolto%20ja%20kunnossapitotoiden%20riskienhallinta/HuoltoLoppuraportti0307062 http://turva.me.tut.fi/tutkimus/Huolto%20ja%20kunnossapitotoiden%20riskienhallinta/HuoltoLoppuraportti0307062 Mitchison, N., & Papadakis, G.A. (1999). Safety management system under Seveso II: Implementation and assessment. Journal of Loss Prevention in the Process Industries, 12(1), 43–51. Mynttinen, M. (2006). Yhteisten työpaikkojen turvallisuus (In Finnish, English abstract). (Bachelor’s thesis, Jyväskylä University of Applied Sciences, Jyväskylä). Retrieved November 6, 2010, from http://www.tyosuojelu.fi/upload/2is9hffo Nenonen, S. (2011). Fatal workplace accidents in outsourced operations in the manufacturing industry, Safety Science. doi:10.1016/j.ssci.2011.06.004. Nenonen, S., & Vasara, J. (in press). Safety management in multi-employer worksites in the manu- facturing industry – opinions on cooperation and problems encountered. International Journal of Occupational Safety and Ergonomics. Nenonen, S., Vasara, J., & Litmanen, A. (2008). Turvallisuusjohtamisen toimintamalli teollisuuden palveluja tarjoaville yrityksille (In Finnish). Retrieved November 10, 2010, from http://turva. me.tut.fi/tutkimus/Palveluntuottajat/Toimintamalli OHSAS 18001. (2007). Occupational health and safety management systems. Specification. British Standards Institution. Papadopoulos, G., Georgiadou, P., Papazoglou, C., & Michaliou, K. (2010). Occupational and public health and safety in changing work environment: An integrated approach for risk assessment and prevention. Safety Science, 48(8), 943–949. Rechenthin, D. (2004). Project safety as a sustainable competitive advantage. Journal of Safety Research, 35(3), 297–308. Rousseau, D.M., & Libuser, C. (1997). Contingent workers in high risk environments. California Management Review, 39(2), 103–123. Shafer, C. (2008). Preconstruction safety, plan for safety excellence. Professional Safety, 53(12), 26–31. SSVV – Centraal College van Deskundigen VCA. (2008). SHE Checklist Contractors (Version 5.1). Netherlands: Author. Retrieved from http://www.vca.nl/content.aspx?guid=7A1B7BE4- DB4E-4865-8DFE-7FFC80AD5321. Sulasalmi, M., Latva-Ranta, J., & Ylijoutsijärvi, P. (2003). Turvallisuusjohtamisen kehittämiskokeilu teollisuuden toimittajayrityksessä (TUTTO-hanke). In M. Sulasalmi & J. Latva-Ranta (Eds.), Turvallisuusjohtaminen teollisuuden toimittajayrityksessä. Lähtökohtia ja kenttäkokemuksia (In Finnish) (pp. 22–71). Ministry of Employment and the Economy. Retrieved July 10, 2011, from http://www.mol.fi/mol/fi/99_pdf/fi/03_tutkimus_ja_kehittaminen/02_tykes/05_ aineistopankki/julkaisut/raportti26 . Toole, T.M. (2002). Construction site safety roles. Journal of Construction Engineering and Management, 128(3), 203–210. Trethewy, R.W., Atkinson, M., & Falls, B. (2003). Improved hazard identification for contractors in the construction industry. Journal of Construction Research, 4(1), 71–85. Välimaa, P., Varonen, U., Lappalainen, J., & Ketola, M. (2001). Risk assessment and safety culture in variable work environments (in Finnish) [English abstract]. Työ ja ihminen, 4 – 5, 231–241. Vasara, J., Nenonen, S., & Hyytinen, T. (2010). Turvallisuusjohtaminen teollisuuden palveluissa – Palveluntoimittajan opas (In Finnish). Helsinki: The Centre for Occupational Safety. Vassie, L.H., & Fuller, C.W. (2003). Assessing the inputs and outputs of partnership arrangements for health and safety management. Employee Relations, 24(5), 492–503. Väyrynen, S. (2003). Vahinkoriskien hallinta, turvallisuuskulttuuri ja johtaminen: Katsaus lähtö- kohtiin. In M. Sulasalmi & J. Latva-Ranta (Eds.), Turvallisuusjohtaminen teollisuuden toimit- tajayrityksessä. Lähtökohtia ja kenttäkokemuksia (In Finnish) (pp. 5–21). Ministry of Employment and the Economy, Retrieved July 10, 2011, from http://www.mol.fi/mol/fi/99_ pdf/fi/03_tutkimus_ja_kehittaminen/02_tykes/05_aineistopankki/julkaisut/raportti26 . Walters, D., & James, P. (2009). Understanding the role of supply chains in influencing health and safety at work. Wigston, Leicestershire: IOSH. Wilson, J.M., & Koehn, E. (2000). Safety management: Problems encountered and recommended solutions. Journal of Construction Engineering and Management, 126(1), 77–79. Zimmerman, D.S. (2005). Risks of involvement in contractor safety. Occupational Health & Safety, 74(1), 38–42. 114 S. Nenonen http://www.tyosuojelu.fi/upload/2is9hffo http://turva.me.tut.fi/tutkimus/Palveluntuottajat/Toimintamalli http://turva.me.tut.fi/tutkimus/Palveluntuottajat/Toimintamalli http://www.mol.fi/mol/fi/99_pdf/fi/03_tutkimus_ja_kehittaminen/02_tykes/05_aineistopankki/julkaisut/raportti26 http://www.mol.fi/mol/fi/99_pdf/fi/03_tutkimus_ja_kehittaminen/02_tykes/05_aineistopankki/julkaisut/raportti26 http://www.mol.fi/mol/fi/99_pdf/fi/03_tutkimus_ja_kehittaminen/02_tykes/05_aineistopankki/julkaisut/raportti26 http://www.mol.fi/mol/fi/99_pdf/fi/03_tutkimus_ja_kehittaminen/02_tykes/05_aineistopankki/julkaisut/raportti26 Copyright of Service Industries Journal is the property of Routledge and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. Process safety auditing References Morrison, L. M., Brown, L., Hazzan, M., & Traynor, J. (2014). Process safety auditing techniques – a U.S. perspective.Loss Prevention Bulletin, (235), 21-23. Abstract: This paper discusses some of the effective auditing techniques operating companies and consultants have used to identify process safety issues and drive improvements in process safety. From the development of the second edition of the CCPS Guidelines for Auditing Process Safety Management Systems, the subcommittee has identified a number of methods and example questions that can be used to audit specific process safety topics. The topics covered in the book include the elements of process safety as defined by CCPS' Guidelines for Risk-Based Process Safety as well as other subjects such as EPA's Risk Management Plan (RMP) and Process Safety Management (PSM)/RMP Applicability. [ABSTRACT FROM AUTHOR] Process safety auditing Loss Prevention Bulletin 235 February 2014 | 21 Good safety practice Process safely auditing techniques - a U.S. perspective LisaM. Morrison, BP; Laurie Brown, Eastman Chemical Company; Mike Hazzan, AcuTech Consulting Croup; John Traynor, Evonik Degussa Corporation Summary This paper discusses some ofthe effective auditing techniques operating companies and consultants have used to identify process safety issues and drive improvements in process safety. From the development ofthe second edition ofthe CCPS Guidelines for Auditing Process Safety Management Systems, the subcommittee has identified a number of methods and example questions that can be used to audit specific process safety topics. The topics covered in the book include the elements of process safety as defined by CCPS' Guidelines for Risk-Based Process Safety as well as other subjects such as EPA's Risk Management Plan (RMP) and Process Safety Management (PSM)/RMP Applicability. K e y w o r d s : Auditing, process safety, safety management Process safety culture Process safety culture, one ofthe new process safety elements identified within risk-based process safety, is a critical component of any PSM program. The common attitudes, work habits, customs and assumptions about process safety that create the environment in which the PSM program exists will determine, to a large extent, whether a well-designed PSM program can be successfully implemented. To accomplish a meaningful audit of this topic requires an examination of values. One example of an effective technique to audit process safety culture is to attend a safety committee meeting, or better yet, a process safety committee meeting, if held during the period ofthe audit. Evaluating the tone ofthe discussions is one piece of information the auditor can use to determine the level of process safety culture at the site. Another piece of evidence comes from observation - if safety bulletin boards are covered with graffiti, there may be issues. As always, the auditor should be careful in reaching conclusions based on only one or two sources - interviews across the facility at all levels, along with record reviews (e.g., meeting minutes) are necessary to fully inform the auditor. PSM/RMP applicability This aspect of PSM/RMP programs is often not audited. However, the decisions made about what processes and equipment are to be included, or more importantly which are to be excluded, is crucial to a process safety program being complete and in compliance. Sometimes a facility actively manages the inventory of PSM- or RMP-covered chemicals to maintain less than the threshold quantity, thus maintaining a process that does not fall under formal regulatory requirements. It is important to review the current inventory of such chemicals, as well as any regular inventory checks that are made overtime and information on amounts purchased, to be able to determine that such inventory controls are working. Management of Change When auditing Management of Change (MOC), auditors usually review the change log and then evaluate whether the change process meets MOC requirements (company and regulatory requirements as well as those required by the facility's own MOC procedure). It is important also to determine whether all changes have gone through the MOC process. One source of information is to interview employees, since employees should be aware of MOC definitions and program requirements to help ensure that inadvertent changes are not made. Work orders, daily logs, and action items from incident investigations and process hazard analyses should also be reviewed. Theauditor looks for indicators of change- words like 'new,' 'changed,' 'improved,' 'removed,' 'upgrade,' and then works with site personnel to determine if these are changes or not. Finally, the auditor searches the MOC database to confirm if these changes were included. Temporary changes are a subset of change that merit a special review. Two auditing techniques to determine if the facility's MOC system adequately treats temporary changes are to 1 ) check to see how many temporary changes have exceeded their pre-determined expiration date and 2) field check closed temporary changes to confirm that they have been returned to their original condition. Additionally, auditors assigned the MOC element should carefully examine whether the audited facility has evaluated changes forthe following health, safety and other important impacts: • Re-analysis and re-approval of changes that exceed the established time period; • Current vent, reliefand flare capability; • Industrial hygiene requirements; © Institution of Chemical Engineers 0260-9576/14/$!7.63 -̂ 0.00 IChemE 22 I Loss Prevention Bulletin 235 February 2014 • Existing environmental permits and other regulatory requirements; • Revisionsto the spare parts list; • Revision of the emergency action plan and/or emergency response plan; and • Ghanges to RMP Program Level or changes in the submitted risk management plan. Asset Integrity (Al) and reliability The list in the facility's computerised maintenance management system may not necessarily include all PSM- covered equipment. The auditor should select equipment from an equipment list, a piping and instrumentation diagram (P&ID) or equipment identified in the field. Al deficiencies Al deficiencies may be found in several areas including out-of- specification inspection, testing and preventive maintenance (ITPM) results (e.g., wall thickness measurements on vessels, tanks, or piping that are at or below retirement thickness; rotating equipment vibration readings that are at the alert value; instrumentation that cannot be calibrated properly; equipment that is operating outside its approved limits, such as throughput that is higher than designed, or temperatures and pressures that are beyond their safe upper or lower limits; bypassed or removed safety features). Auditors should review ITPM records to determine what deftciencies exist and when they were first identifted, to see that they were addressed in a timely manner. Training and performance assessment When reviewing training records under the competency element, one could ask for a record of everyone who has been trained from the training documentation database. Or the auditor could browse the information in the database. It is preferable, however, to start with an employee list that includes job titles. From this list, an appropriate sample of difterent types of employees can be chosen, and the required training determined. At this point, a request of all training from the training database will provide the auditor with information that starts from a more relevant population set. Process Safety Competency The Process Safety Gompetency (PSG) element is one of the elements from GGPS' Guidelines for Risk-Based Process Safety that has no corresponding element in OSHA PSM or EPA RMP. Encompassing organisational learning, it is a combination of three interrelated actions: (1) continually improving knowledge and competency, (2) ensuring that appropriate information is available to people who need to know it and (3) consistently applying what has already been learned. One key element that should be audited under PSG is that the basis for past design, operational and maintenance decisions is documented in a retrievable manner. The auditor could start by asking if such information is documented somewhere and review the available documentation. The auditor could also start by identifying some decisions that were previously made - these decisions can be found from several sources, including incident investigations, previous PSM audits, process hazard analyses, operational review documentation, management of change documentation, commissioning documentation and pre-startup safety reviews. Examples of some of these decisions include why the steps of a procedure are in a particular order; the use of a particular brand, make or model of equipment or why the use of a particular material supplier is preferred. Interviewing is another method for determining whether decisions have been made and documented - just asking why often elicits this information. Then the auditor can follow up to look for documentation. Workforce involvement The OSHA PSM element Trade Secrets has been included within Workforce Involvement because it fits well with the other aspects regarding ensuring the workforce has access to applicable information. This subject is not often given much time in process safety audits, but just including the following question when otherwise interviewing employees on Workforce Involvement may yield some important information: "Gan you tell me about any situation where you requested information about the process and you were turned down? if so, what reason was given?" Process Knowledge Management A great deal of time is, and should be, spent auditing the element of Process Knowledge Management. Process Knowledge Management encompasses a number of requirements. One important aspect that should be reviewed is whether or not facility processes are being operated within the safe upper and lower limits. The auditor needs to review the safe upper and lower limit documentation, being sure to check the original equipment information as well as any other documentation that the facility has developed to summarise such information. Then the auditor should review operating procedures and limits within the Distributed Gontrol System (DGS), if one exists, to see that both the procedures and the DGS indicate that the process is being operated within its safe limits. In doing so, the auditor can also complete part of the operating procedures assessment. Another important part of auditing this element involves a review of safety systems. The auditor will need to determine what safety systems exist at the site. A good idea is to go over a list of typical safety systems with knowledgeable personnel at the site to determine which safety systems are being used. This can be done during the preparation phase of the audit or during the audit itself. An example of one such list follows: • Gontrols including: - safety instrumented systems, - indicators, - alarms, - trips and, - interlocks. • Systems that are intended to detect or suppress reactions or chemical releases, including: - quench systems, - rapid neutralisation systems, - reaction kill injection systems and - vapor cloud knockdown systems. IChemE © Institution of Chemical Engineers0260-957ó/14/$17.63 + 0.00 Loss Prevention Bulletin 235 February 2014 | 23 • Systems that are intended to mitigate vapour releases, for example HF deluge systems. • Secondary containment systems. • Inerting systems. • Fire protection equipment, including: - fire water system, - hydrants, - monitors, - deluge systems, - sprinkler systems, - foam systems, - special extinguishing systems (e.g., halon) and - fireproofing. • Explosion or blast panels. • Explosion suppression systems (e.g., Fenwal). • Uninterruptable power sources. • Flame arrestors. • Detection systems (hydrocarbon, oxygen, carbon monoxide, phosgene, etc.) • Ventilation (local and room). • HVAC systems for control rooms and rooms where temperature must be controlled for safety reasons. • Relief systems, including relief valves, rupture discs, relief system piping, relief system containment (dump tanks, burp tanks), depressuring systems (e.g., flare system) and dump systems. Once a complete list of safety systems is developed, the auditor can sample from this list and review whether information exists describing how the system works, setpoints and control features. Operating procedures The list of safety systems that is developed for the Process Safety Knowledge element can also be used when reviewing operating procedures. The auditor can review the list to determine if every applicable safety system and its function are described in the operating procedures. Hazard Identification and Risk Assessment Another area that deserves a lot of time auditing is the Risk- based Process Safety (RBPS) element of Hazard Identification and Risk Assessment (HIRA). The first part of this element requires that all hazards are identified. An auditor needs to bring a combination of experience with the process being audited and preparation to be able to effectively review this area, if an auditor has a deep background in the process being audited, that is often enough to be able to sit down with the Hi RA studies and review them to see if the typical hazards are identified and addressed. Without this deep background, an auditor needs to prepare by studying the process through whatever documents are available. In some audits, the facility presents an overview of the process - a lot of information can be gathered from this overview. The facility tour is another source of information about the process and its hazards, as isthefacility safety orientation. So, if during the facility overview the auditor learns that high pressure nitrogen is being fed to the three reactors, whether or not the hazard of high pressure was addressed in the HIRA studies can be reviewed. Or, if an auditor notices a large accumulation of dust in the process area, he or she can ask about the nature of the dust and also review the HIRA studies to see if this potential hazard has been addressed. A key area of HIRA is the identification of safeguards. When documented in a HIRA study, it is implicit that the safeguards exist and are functional. This assumption should be checked. The auditor can take a sample of safeguards found in the HIRA studies and check that they are in the field. The auditor may be able to determine ifthey are functioning by this field inspection, but in addition, the auditor should determine if the safeguards are included in some type of test or inspection activity during the Ai portion of the audit. Emergency management A critical part of an operating emergency management system is the emergency alarm system. The emergency alarm system needs to be perceivable by all who could be affected by any potential emergency. A useful activity during an audit is to determine in advance, or upon arrival at the site, the date and time of the periodictest of the emergency alarm system. It is often done weekly, which can be used to the audit team's advantage. The lead auditor should station members of the audit team around the facility in noisy areas or other areas where it may be difficult to perceive the emergency alarm system, such as isolated or more remote areas. The auditors should be wearing the typical personal protective equipment worn by the facility employees, such as hearing protection, and, when the alarm goes off, determine ifthey can perceive the alarm, through sound, light or other method. Incident investigation Before an incident can be properly investigated, the incident must be reported. To determine whether all process safety incidents are being reported, the auditor has a number of resources to consult. Besides interviewing a broad sample of employees on whether or not they believe or have knowledge that all incidents are being reported, the auditor can review logs and daily reports. Action item closure A PSM audit offers numerous opportunities to review the adequate closure of action items. Action items are generated from incident investigations, HIRA studies, PSM compliance audits, MOC reviews, Pre-Startup Safety Reviews, commissioning reviews and different engineering studies, such as relief valve and electrical classification reviews. Each of these sources offers numerous items for auditors to review. The auditor should verify that the action documented actually addresses the issue or risk identified in the recommendations and thatthe action was completely closed in the field, and not 'closed on a plan' or 'closed on intent'. This activity warrants several dedicated time slots during the typical PSM audit. Conclusion A review of the auditing techniques included in this paper by audit managers and team leaders, as well as process safety managers, will provide opportunities to improve process safety audit programs as well as process safety management reviews. © institution of Cfiemical Engineers 0260-957ó/14/$17.63 + 0.00 IChemE Copyright of Loss Prevention Bulletin is the property of Institution of Chemical Engineers and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. Interrelationships among Highly Effective Construction Injury Prevention Strategies References Hallowell, M. R., & Calhoun, M. E. (2011). Interrelationships among Highly Effective Construction Injury Prevention Strategies. Journal Of Construction Engineering & Management, 137(11), 985-993. doi:10.1061/(ASCE)CO.1943-7862.0000354 Abstract: Preventing injuries to workers is essential to effective organizationalmanagement in the construction industry. It is generally accepted that achieving a high level of safety performance requires the implementation of an effective safety program. In an effort to provide guidance for the development of an effective safety program, previous studies have focused on identifying highly effective injury prevention strategies and quantifying their individual impacts on safety performance. Despite these advancements in knowledge, the synergistic interrelationships among highly effective injury prevention strategies remain unknown. The objective of this paper is to describe the results of a Delphi study that quantified the interrelationships (i.e., pairwise cross impacts) of highly effective and commonly implemented injury prevention strategies. Analysis of the data indicates that the site safety manager, worker participation and involvement, a site-specific safety plan, and upper management support and commitment are the most central elements in an effective program. It is expected that the results of this study can be used by practicing professionals when designing a new safety program, enhancing an existing program, or conducting economic analyses of safety-related investments. [ABSTRACT FROM AUTHOR] Interrelationships among Highly Effective Construction Injury Prevention Strategies Interrelationships among Highly Effective Construction Injury Prevention Strategies Matthew R. Hallowell, A.M.ASCE1; and Matthew E. Calhoun2 Abstract: Preventing injuries to workers is essential to effective organizational management in the construction industry. It is generally accepted that achieving a high level of safety performance requires the implementation of an effective safety program. In an effort to provide guidance for the development of an effective safety program, previous studies have focused on identifying highly effective injury prevention strategies and quantifying their individual impacts on safety performance. Despite these advancements in knowledge, the synergistic inter- relationships among highly effective injury prevention strategies remain unknown. The objective of this paper is to describe the results of a Delphi study that quantified the interrelationships (i.e., pairwise cross impacts) of highly effective and commonly implemented injury pre- vention strategies. Analysis of the data indicates that the site safety manager, worker participation and involvement, a site-specific safety plan, and upper management support and commitment are the most central elements in an effective program. It is expected that the results of this study can be used by practicing professionals when designing a new safety program, enhancing an existing program, or conducting economic analyses of safety-related investments. DOI: 10.1061/(ASCE)CO.1943-7862.0000354. © 2011 American Society of Civil Engineers. CE Database subject headings: Construction industry; Injuries; Occupational safety; Risk management. Author keywords: Safety; Risk management. Introduction It is no secret that the construction industry accounts for a dispro- portionate injury rate. In 2002, the estimated direct and indirect costs of fatal and nonfatal construction injuries totaled $13 billion [Bureau of Labor Statistics (BLS) 2009]. Fortunately, safety per- formance in the construction industry has improved significantly in the past 40 years. In fact, between 1973 and 2004 the fatality rate decreased from 71 to 11.1 per 100,000 workers [National Safety Council (NSC) 2006]. Nevertheless, the industry continues to have the greatest number of fatalities and in 2009 accounted for nearly 20% of all workplace deaths from all industries combined (BLS 2009). Additionally, in 2005, the fatality rate in the U.S. con- struction industry was three times greater than the fatality rate in other developed countries, such as Sweden (4.4) and Switzerland (4.8) (Center for Construction Research and Training 2008). In the United States, the federal agency tasked with carrying out research and making recommendations for the prevention of work-related injuries is the National Institute for Occupational Safety and Health (NIOSH). In 1996, NIOSH initiated the National Occupational Research Agenda (NORA) to conduct research that identifies critical occupational safety and health issues. Recently, in an effort to become more effective in preventing fatalities, injuries, and illnesses in construction, the NORA Construction Sector Council developed 15 strategic goals for the National Construc- tion Agenda. One of the strategic goals was aimed at improving the effectiveness of safety and health management programs in construction and increasing their use in the industry (NORA 2008). In an effort to address that goal, there have been a number of studies that discuss the elements that constitute an effective safety program. For example, recent researchers have quantified the effectiveness of the independent implementation of various injury prevention strategies (Rajendran and Gambatese 2009; Hallowell and Gambatese 2009). Despite these recent advancements in knowledge, the synergis- tic impacts among highly effective injury prevention strategies remain unknown. Such information is essential to the development of an optimal safety program because individual injury prevention strategies are rarely implemented in isolation. Furthermore, under- standing the potential synergistic effects of clusters of safety pro- gram elements may allow safety managers to formalize the design of a safety program when resources are limited. The objective of this paper is to describe the results of a Delphi study that quantified the interrelationships (i.e., pairwise cross impacts) of highly effective and commonly implemented injury prevention strategies. It is expected that these unprecedented data can be used to identify the elements that are central to the overall effectiveness of the program and to strategically design a safety program when a subset of elements is implemented. Literature Review Since the Occupational Safety and Health Act of 1970 assigned the responsibility of site safety to the employer, injury prevention strategies have become more common. In recent years, safety pro- grams have matured and improved in effectiveness, resulting in reduction in incident rates for the industry as a whole (Hill 2001). According to Rajendran (2006), there exist hundreds of unique safety program elements that are implemented on construction 1Assistant Professor, Dept. of Civil, Environmental, and Architectural Engineering, Univ. of Colorado at Boulder, 428 UCB, 1111 Engineering Drive, Boulder, CO 80309-0428 (corresponding author). E-mail: Matthew .Hallowell@colorado.edu 2Ph.D. Student, Dept. of Civil, Environmental, and Architectural Engineering, Univ. of Colorado at Boulder, 428 UCB, 1111 Engineering Drive, Boulder, CO 80309-0428. E-mail: Matthew.Calhoun@colorado.edu Note. This manuscript was submitted on January 27, 2010; approved on January 14, 2011; published online on January 17, 2011. Discussion period open until April 1, 2012; separate discussions must be submitted for indi- vidual papers. This paper is part of the Journal of Construction Engineer- ing and Management, Vol. 137, No. 11, November 1, 2011. ©ASCE, ISSN 0733-9364/2011/11-985–993/$25.00. JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / NOVEMBER 2011 / 985 http://dx.doi.org/10.1061/(ASCE)CO.1943-7862.0000354 http://dx.doi.org/10.1061/(ASCE)CO.1943-7862.0000354 http://dx.doi.org/10.1061/(ASCE)CO.1943-7862.0000354 http://dx.doi.org/10.1061/(ASCE)CO.1943-7862.0000354 work sites. As indicated, the objective of this study is to investigate the interrelationships among highly effective safety program ele- ments. Thus, to narrow the scope, the writers were faced with the challenge of identifying and defining highly effective injury pre- vention strategies. During the literature review, several studies were found that dis- cuss essential elements of a safety program (Liska and Goodloe 1993; Meridian 1994; Jaselskis et al. 1996; Findley et al. 2004; Hill 2001; Hinze 2006; Hallowell and Gambatese 2009). These publications identify 20 unique elements that are considered to be essential components of an effective safety program. Of these 20 elements, 13 were mentioned by at least four of the seven pub- lications. These 13 elements, listed in Table 1, were the foci of this study. The number of elements was narrowed to 13 because this number results in an appropriate scope (i.e., 156 pairwise ratings). As previously discussed, there exists a well-established body of literature that identifies the elements of an effective safety program. Recently, several researchers have found that some elements are more effective than others. Jaselskis et al. (1996) studied the effec- tiveness of safety program elements by measuring the ability of each element to decrease a company’s experience modification rate. Sawacha et al. (1999) built upon this research by conducting a factor analysis to determine the relative impacts of various safety program elements on injury rates. In a high-profile study conducted for the Construction Industry Institute (CII), Hinze (2002) studied the elements needed to achieve the zero injury objective. Recently, Molenaar et al. (2009) created a set of best practices by evaluating five latent variables through a structural equation model, and Hallowell and Gambatese (2009) used the Delphi method to quan- tify the individual ability of safety program elements to reduce the risk of construction injuries and illnesses. These studies have indicated that upper management support and commitment, safety staffing, preproject planning, safety education, worker involve- ment, employee recognition, substance abuse programs, accident investigations, strategic subcontractor selection and management, and job hazard analyses are the most effective strategies. A basic limitation of the existing literature is that the effective- ness of safety program elements has been studied and quantified without explicit reference to the impact that they have on one another. The only study that considers the synergy of individual safety program elements was Hinze (2002), who found that safety strategies are more effective when they are implemented in groups. For example, safety meetings tend to be more effective when upper managers actively participate, there exists appropriate staffing for safety, and workers are involved in management and planning. The limitation of this study was that the pairwise interrelationships were not considered, and the synergistic effects were not quantified. Research Methods To quantify the pairwise interactions among highly effective safety program elements, a cross-impact analysis was conducted by using the Delphi method. The specific research design, justification of the selection of this research method, and methods implemented to Table 1. Critical Elements of an Effective Construction Safety Program (After Hallowell and Gambatese 2009) Safety program element Description Upper management support Explicit acknowledgement from upper management that worker safety and health is a primary goal of the firm demonstrated by participation in regular safety meetings and committees and sufficient funding Subcontractor selection and mgmt Consideration of safety and health performance during the selection and management of subcontractors (e.g., prequalification and required compliance) Employee involvement and evaluation Including all employees in the formulation and execution of other safety elements and including participation and safe work behavior in evaluations Job hazard analyses (JHA) Review and recording activities associated with a construction process, highlighting potential hazardous exposures, and documenting safe work practices that prevent injury Project-specific training/meetings Establishing and communicating project-specific safety goals, plans, and policies before the construction phase of the project Frequent work site inspections Inspections performed internally by a contractor’s safety manager, safety committee, representative of the contractor’s insurance provider, or by an OSHA consultant to identify uncontrolled hazardous exposures Safety manager on site Employment of a safety and health professional (i.e., an individual with formal construction safety and health experience and/or education) whose primary responsibility is to perform and direct the implementation of safety and health program elements and serve as a resource for employees Substance abuse programs Identification and prevention of substance abuse of the workforce (includes random testing and testing after an injury) Safety and health committees Committee with the power to affect change and set policies, consisting of a diverse group including supervisors, laborers, representatives of key subcontractors, owner representatives, OSHA consultants, may be formed with the sole purpose of addressing safety and health on the work site S&H orientation/training Participation of all new hires or transfers in orientation and training sessions that have a specific focus on safe work practices and company safety policies Written safety and health plan Development of a documented plan that identifies project-specific safety objectives, unique hazards, and methods for achieving success Record keeping/analyses Regular reporting of the specifics of all accidents, including information such as time, location, work site conditions, and cause Emergency response planning Creation of a plan that documents the company’s policies and procedures in the case of a serious incident or catastrophe, such as a fatality or an incident involving multiple serious injuries 986 / JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / NOVEMBER 2011 minimize bias and enhance the validity and reliability of the results are described in detail in this section. Cross-Impact Analysis Cross-impact analyses have been performed as a part of many studies in the area of forecasting economics and engineering to es- timate interactions among events or elements that were previously considered independent of one another (Seung and Diekmann 2004; Gordon and Hayward 1968). There exist many versions of cross-impact analyses that have been developed by researchers and practitioners. In general, cross-impact analyses are performed by quantitatively estimating the impact each element of interest has on the other elements of interest (Blanning and Reinig 1999). One way to perform such an analysis is to engage knowledgeable people and solicit subjective estimates among the elements, usually result- ing in a matrix of conditional probabilities or percent impacts on a particular performance metric (Mitchell and Tydemann 1978). In such studies, the Delphi method is generally the research method of choice because it offers the researcher the ability to obtain con- sistent, valid, and reliable results (Helmer 1977; Turoff 1972; Ezner 1972). According to Helmer (1977), the major steps required to perform a cross-impact analysis are as follows: 1. Define the elements to be included in the analysis; 2. Develop cross-impact matrices to define the interdependencies between the elements; 3. Define how the interactions or interdependencies will be measured; 4. Estimate the interactions or interdependencies in the cross- impact matrix; 5. Perform the cross-impact calculations; and 6. Evaluate the results. Delphi Method The present study utilizes the Delphi method to perform the fourth step of the cross-impact analysis strategy outline by Helmer (1977). The interdependencies were measured as the percent in- crease or decrease in effectiveness that each program element had on the others. For example, the Delphi panelists were asked to rate the interaction between safety manager and job hazard analy- sis. This interaction was measured by rating the impact that a safety manager has on the effectiveness of job hazard analyses measured as a percent increase or decrease in base-level effectiveness. The Delphi method was selected for this analysis because it has been used successfully in previous cross-impact studies (Turoff 1972; Helmer 1977; Kendall 1977; Eschenbach and Geistauts 1985; Ren et al. 2007); the method offers the facilitator the oppor- tunity to employ controls to minimize the potential impact of judgment-based biases, and because Delphi is preferred when ob- jective data are unavailable, experimental research is unrealistic, and empirical evidence cannot be separated from confounding fac- tors (Linstone and Turoff 1975). This approach was also selected because of the relatively significant time investment required of participants due to the high volume of data desired (i.e., 156 ratings). Delphi differs from other expert-based studies because panelists are certified as experts by using objective criteria before the first round of surveys; multiple rounds of surveys are administered in an effort to achieve consensus within the group; feedback is pro- vided by the facilitator during each round; and panelists remain anonymous. One of the chief benefits of the Delphi method is the opportunity for the facilitator to reduce the potential impacts of judgment-based bias. Recently, Hallowell and Gambatese (2010) identified eight judgment-based biases that may lead to invalid or unreliable data and associated controls that may be implemented through the design and administration of the Delphi process. These biases include collective unconscious, contrast effect, neglect or probability, the Von Restorff effect, myside bias, recency effect, primacy effect, and dominance. To minimize the potential effects of these biases, the following four controls were implemented in this study, as suggested by Hallowell and Gambatese (2010): randomize the question order on the Delphi surveys; include rea- sons for outlying responses in controlled feedback between rounds; conduct multiple rounds of surveys with anonymous panelists; and report median ratings as the final results. In addition to the controls listed, the Delphi panel was provided with specific definitions of the safety program elements that are consistent with the literature reviewed. Because the Delphi method has been criticized by some due to the shortcuts and modification to the prescribed research method, a description of and justification for the expertise requirements, number of panelists, number of rounds, feedback provided, and targeted level of consensus are provided. Expertise Requirements To be considered for a Delphi study, the panelists must be certified as experts by using objective criteria before initiating the first round of data collection. Following the custom of practice, demographic data, obtained during an introductory survey, were used to qualify individuals as experts. As suggested by Rogers and Lopez (2002) and demonstrated by Rajendran and Gamabatese (2009), every panelist was required to meet at least four of the following eight characteristics related to the construction safety management to qualify as an expert for this study: 1. Primary or secondary author of at least three peer- reviewed journal articles on the topic of injury prevention in construction, 2. At least three presentations on a safety-related topic at a national conference, 3. Member of a national construction safety committee (ASCE site safety, CII Safety Community of Practice, CIB W099), 4. At least 5 years of professional experience in the construction industry with safety-management responsibilities, 5. Faculty member at an accredited institution of higher learning with a research or teaching focus on injury prevention in construction, 6. Author or editor of a book or book chapter on the topic of injury prevention in construction, 7. Advanced degree in the field of civil engineering, construction engineering, occupational safety and health, or other fields directly related to this study, from an institution of higher learning (minimum of a B.S.), and 8. Designation as a Professional Engineer (PE), Certified Safety Professional (CSP), Associated Risk Manager (ARM), or a Licensed Architect (AIA). The Delphi panel formed for this study was extremely well- qualified to address this research topic. The panel included six academics (i.e., professors or full-time researchers at academic in- stitutions who specialize in construction safety and health) and four professionals (i.e., individuals with five or more years of safety management responsibility). The education of the panel included seven panelists with a terminal degree (Ph.D.), two with a Master of Science (M.S.) degree, and one with a Bachelor of Science (B.S.), all in related fields. Additionally, the collective panel auth- ored 204 peer-reviewed journal papers, 13 books or book chapters, 348 peer-reviewed conference proceedings, and 49 trade publica- tions on the topic of construction safety and health. Although most of the panelists were employed at academic institutions during the JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / NOVEMBER 2011 / 987 study, the panel also had a wealth of professional experience, including a collective 108 years of professional experience related to construction safety. Furthermore, five of the panelists were reg- istered as PEs and three were CSPs. Finally, all panel members were actively participating on at least one national construction safety committee associated with the ASCE, the International Council for Research and Innovation in Building and Construction (CIB), or the CII. Number of Panelists Delphi studies range significantly in the number of expert panelists involved. The number of panelists required depends on the number of experts available on the particular topic, the expected volume of data, associated time requirements, and the ability and sophistica- tion of the facilitator. Brockhoff (1975) and Boje and Murnighan (1982) studied the impact of the number of panelists on the level of accuracy of the Delphi method and concluded that the appropriate number of panelists for the typical Delphi study ranges from eight to 15. For the present study the authors targeted an initial group of 15 experts in case one or more of the panelists defaulted during the series of survey rounds. Of the 15 participants who were tar- geted, 13 agreed to participate, and 10 successfully completed all survey rounds. Number of Rounds and Feedback Provided Delphi is characterized, in part, by the use of multiple rounds and feedback provided between rounds. Most literature indicates that the Delphi process should continue for as many rounds as it takes to achieve the desired consensus. However, other literature indi- cates that Delphi results are most accurate after Rounds 2 and 3 and become less accurate as a result of additional rounds (Dalkey 1971). Most researchers suggest three rounds because this number tends to be adequate for achieving consensus and implementing the controls to minimize bias. In this study, the authors elected to con- duct three rounds of surveys with detailed feedback provided at the beginning of Rounds 2 and 3. Providing adequate and strategic feedback allows expert panel- ists to anonymously consider the opinions and experiences of other members without having to be subjected to time-consuming discus- sions, which are also prone to dominance, myside, and collective unconscious biases. Feedback was provided in written feedback and quantitative statistics from previous rounds. To ensure adequate feedback that would promote consensus, this study involved con- trolled feedback at the beginning of Rounds 2 and 3. In Round 2, the median interaction ratings from the previous round were pro- vided to all panelists, in addition to their personal rating from the previous round. During Round 2, panelists were asked to provide reasons if they believed that the true value for a particular rating deviated more than 10% from the group median from Round 1. In Round 3, the panelists were provided with the median ratings from the second round, their personal rating from the second round, and the reasons provided by all panelists for outlying responses. Target Consensus The primary objective of any Delphi study is to achieve consensus among a certified group of experts with regard to a specific topic. Thus, measuring consensus is an integral component of the Delphi process. Most quantitative studies use standard deviation of abso- lute deviation to measure consensus depending on whether the facilitator elects to report the results in median or mean responses. In the present study, the writers set the goal of achieving an ulti- mate absolute deviation (i.e., average deviation from the median) of < 0:10. Survey Design As indicated, the Delphi panelists were asked to rate the interac- tions among safety program elements as measured by a percent in- crease or decrease in effectiveness. Each pairwise interaction was considered independently. Furthermore, the interactions were con- sidered to be two-way interactions. For example, the impact that a safety manager has on the effectiveness of job hazard analyses was considered independently from the impact that job hazard analyses have on the effectiveness of safety managers. For clarity, an exam- ple survey sheet from Round 1 of the Delphi process is provided as Fig. 1. Results The Delphi study was conducted over a 4-month period with approximately 1 month dedicated to each survey round. This rel- atively long duration was deemed necessary because each panelist was asked to provide 156 ratings per round for all three rounds. Out of the original 13 panelists, 10 completed all three survey rounds, resulting in a response rate of 77%. In total, 4,680 ratings were gathered from the expert panel. As previously indicated, one of the benefits of the Delphi method is the ability to achieve consensus within a group. After the first round, the average absolute deviation of all ratings was 19%. After three rounds the absolute deviation was 12%. Although the group did not come to the target consensus of 10%, a fourth round of surveys was deemed to be unnecessary because the medi- ans remained unchanged between Rounds 2 and 3, each Delphi panelist had already produced 468 ratings (156 ratings per round), and Delphi studies become increasingly less accurate after the third round (Dalkey 1971). To test for any statistically significant differences between the responses of professional and academic Delphi panelists, the authors performed a Wilcoxon rank sum test. The results indi- cated that there was no statistically significant difference in the re- sponses (p-value ¼ 0:78). The resulting pairwise cross impacts are provided in Table 2. These data represent the median ratings obtained from the 10 expert panelists who participated in all three rounds. From the information in this table, all of the cross impacts are positive. In other words, the results indicate that all of the injury prevention strategies have a positive relationship with one another. Only one interaction, the impact of record keeping and accident analyses on substance abuse programs, resulted in a 0% impact. Most of the interactions resulted in cross-impact ratings of between 30 and 60%. Table 3 was produced to highlight the 10 most significant syn- ergistic relationships. As one can see, a site safety manager plays a central role in enhancing the effectiveness of inspections, safety plans, safety training, worker participation, and job hazard analy- ses. Additionally, upper management support and commitment was found to significantly increase the effectiveness of the site safety manager, safety plan, subcontractor safety management, and safety and health committees. These results may not be surprising because effective executives and safety managers tend to be well-educated, have a job function that involves integrating and involving multiple organizational units, and are ultimately responsible for implement- ing and managing safety-related activities within the organization. Analysis In addition to the pairwise interactions previously highlighted, a simple analysis shows the degree to which specific elements 988 / JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / NOVEMBER 2011 contribute to the overall effectiveness of the safety program. Table 4 includes the following three measures: 1. Total contribution received by the other 13 elements, measured as the sum of all impacts on a particular element (i.e., sum of the columns in Table 2), 2. Total contribution made to the other 13 elements, measured as the sum of all impacts made by a particular element (i.e., sum of the rows in Table 2), and 3. The total level of centrality of a particular element as measured by the sum of the contributions made to and contributions received from the other elements. These three measures are unitless but provide relative scores that indicate the contributing impact, receiving impact, and overall centrality. This analysis uses the sum of both the impact that a par- ticular element has on the others and the amount that the element is impacted by others as a measure of centrality because it represents the relative increase in effectiveness that the inclusion of the element would have on the overall effectiveness of the safety program. These three measures give an indication of the impact that each element would have if integrated into a safety program that includes the remaining 12 elements highlighted in this study. These values would change, however, if one or more elements were omitted from a program but could be simply calculated by finding the sum of the rows and columns of the remaining matrix. From Table 4, it is clear that the elements most central to the safety program are the site safety manager, worker participation and involvement, and upper management support and commitment. The elements that are least central are emergency response plan- ning, substance abuse programs, and safety and health orientation. This supports previous research that has shown that upper manage- ment support and commitment, safety managers, and worker par- ticipation and involvement are essential elements of an effective injury prevention program. These findings build on this body of knowledge by quantifying synergistic interrelationships. Validation In an effort to validate the findings, structured phone interviews were completed with eight construction safety experts. Members of the ASCE Site Safety Committee and representatives from the Occupational Safety and Health Administration (OSHA) were targeted for participation. To minimize bias, potential participants were randomly selected from both contact lists. Fortunately, all eight individuals initially contacted agreed to participate in the validation effort, resulting in a response rate of 100%. Although the initial Delphi panel consisted of both academic and industry professionals, the validation panel was made up of experts who were actively managing safety in the construction in- dustry or conducting inspections and providing safety training on behalf of OSHA. Collectively, these eight industry experts had a total of 320 years experience in the construction industry. The indi- viduals who participated in the validation interviews were not a part of the initial Delphi panel. To ensure consistency, each interviewee was supplied before the interview date with the list of validation questions and the list of the safety program elements under investigation. The validation questions included the following: 1. Of the 13 injury prevention strategies listed, which five are impacted the most by the other elements? 2. Of the 13 injury prevention strategies, which five contribute the most to the effectiveness other elements? and 3. Of the 13 injury prevention strategies, which five pairwise safety program element interactions are most significant? DIRECTIONS: Put an “X” in the box that indicates the percent increase that a SITE-SAFETY MANAGER has on the effectiveness of each safety program elements listed. Safety Program Element N eg at iv e In fl ue nc e Percent increase that a SITE-SAFETY MANAGER has on the effectiveness of the indicated safety program element 0 10 20 30 40 50 60 70 80 90 100 >100
Project Safety Incentives
Site-Specific Safety Plan
Substance Abuse
Programs
Safety and Health
Committees
Training and Regular
Safety Meetings
Worker Participation and
Involvement
On-site First Aid
Safety and Health
Orientation
Job Hazard Analyses
Subcontractor Selections
and Compliance
Record Keeping and
Accident Analyses
Emergency Response
Planning
Upper Management
Support
Frequent Safety
Inspections
Fig. 1. Example Round 1 survey sheet
JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / NOVEMBER 2011 / 989
T
a
b
le
2
.
P
ai
rw
is
e
C
ro
ss
Im
pa
ct
s
of
C
on
st
ru
ct
io
n
S
af
et
y
P
ro
gr
am
E
le
m
en
ts
Im
pa
ct
on
E
m
er
ge
nc
y
re
sp
on
se
pl
an
F
re
qu
en
t
in
sp
ec
ti
on
s
Jo
b
ha
za
rd
an
al
ys
es
R
ec
or
d
ke
ep
in
g
an
d
ac
ci
de
nt
an
al
ys
es
S
af
et
y
an
d
he
al
th
co
m
m
it
te
e
S
af
et
y
an
d
he
al
th
or
ie
nt
at
io
n
S
it
e
sa
fe
ty
m
an
ag
er
S
it
e-
sp
ec
if
ic
sa
fe
ty
pl
an
S
ub
co
nt
ra
ct
or
m
an
ag
em
en
t
S
ub
st
an
ce
ab
us
e
pr
og
ra
m
s
T
ra
in
in
g
an
d
re
gu
la
r
sa
fe
ty
m
ee
ti
ng
s
U
pp
er
m
an
ag
em
en
t
su
pp
or
t
W
or
ke
r
in
vo
lv
em
en
t
P
er
ce
nt
in
cr
ea
se
in
ef
fe
ct
iv
en
es
s
Im
pa
ct
of
E
m
er
ge
nc
y
re
sp
on
se
pl
an
5
18
8
25
8
16
25
25
30
16
20
30
F
re
qu
en
t
in
sp
ec
ti
on
s
13
70
48
55
23
75
51
33
10
55
33
55
Jo
b
ha
za
rd
an
al
ys
es
25
65
30
25
15
60
79
25
5
35
50
55
R
ec
or
d
ke
ep
in
g
an
d
ac
ci
de
nt
an
al
ys
es
15
30
28
30
20
46
20
30
0
20
50
30
S
af
et
y
an
d
he
al
th
co
m
m
it
te
e
30
54
45
40
40
53
60
33
33
60
50
80
S
af
et
y
an
d
he
al
th
or
ie
nt
at
io
n
20
30
28
35
28
33
30
10
30
40
13
43
S
it
e
sa
fe
ty
m
an
ag
er
50
90
80
60
45
75
85
50
60
85
53
85
S
it
e-
sp
ec
if
ic
sa
fe
ty
pl
an
40
60
60
34
53
50
50
40
50
50
35
55
S
ub
co
nt
ra
ct
or
m
an
ag
em
en
t
25
45
48
35
30
25
40
41
25
28
30
41
S
ub
st
an
ce
ab
us
e
pr
og
ra
m
s
5
18
23
25
8
16
25
25
30
16
20
30
T
ra
in
in
g
an
d
re
gu
la
r
sa
fe
ty
m
ee
ti
ng
s
46
63
65
35
38
48
50
50
13
35
13
78
U
pp
er
m
an
ag
em
en
t
su
pp
or
t
55
64
50
65
79
58
80
80
80
70
66
70
W
or
ke
r
in
vo
lv
em
en
t
46
39
65
55
70
55
45
60
13
45
65
20
990 / JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / NOVEMBER 2011
Each phone interview was conducted one-on-one in a 1-h ses-
sion. When appropriate, the questions were clarified to ensure
proper validation. The results of the initial Delphi study were
not discussed with the interviewees until after responses to the ini-
tial questions had been provided. Table 5 summarizes the responses
from the validation interviews and compares the responses of the
interviewees with the initial Delphi results. Specifically, Table 5
shows the number of interviewees that selected each element as
one of the top five elements that receive support from the other
elements (Question 1) and the number of interviewees that selected
each element in the top five contributing elements (Question 2).
From the comparison between the initial Delphi results and the
results of the validation interviews, there was a great deal of con-
sistency. In fact, the five elements identified through the Delphi
analysis as those that receive the most support from other elements
were also the top five elements identified in the validation inter-
views. Similarly, four of the five elements initially identified as
the top contributing elements were also ranked in the top five in
the validation interviews. The safety and health committee was
identified as a top-five contributor in the analysis of the original
Delphi data but was not ranked in the top five in the interviews.
This lower rating reflects the interviewees’ belief that safety and
health committees tend to be less effective on small projects. Alter-
natively, the panel rated a site-specific safety plan as a top-five con-
tributor because it helps to establish a culture of safety during the
initial stages of the project. Despite this one conflict, the responses
Table 3. Most Significant Interrelationships of Construction Safety Program Elements
Contributing element Receiving element
Receiving element’s potential
percent increase in effectiveness
Site safety manager Frequent inspections 90%
Site safety manager Site-specific safety plan 85%
Site safety manager Training and regular safety meetings 85%
Site safety manager Worker participation and involvement 85%
Site safety manager Job hazard analyses 80%
Safety and health committee Worker participation and involvement 80%
Upper management support Site safety manager 80%
Upper management support Site-specific safety plan 80%
Upper management support Subcontractor selections and compliance 80%
Upper management support Safety and health committee 79%
Table 4. Measures of Synergistic Impact of Each Element on the Overall
Effectiveness of the Safety Program
Safety program element
Total contribution
received from
other elements
(relative unitless
score)
Total
contribution
made to other
elements (relative
unitless score) Sum
Site safety manager 573 818 1,391
Worker participation and
involvement
651 578 1,229
Upper management support 385 816 1,201
Site-specific safety plan 606 576 1,182
Frequent inspections 561 519 1,080
Training and regular safety
meetings
536 531 1,067
Safety and health committee 484 576 1,060
Job hazard analyses 578 469 1,047
Subcontractor selections and
compliance
380 413 793
Record keeping and accident
analyses
469 319 788
Safety and health orientation 431 338 769
Substance abuse programs 393 240 633
Emergency response plan 370 225 595
Table 5. Comparison of Validation Results with Initial Delphi Results (Validation of Questions 1 and 2)
Safety program element
Contributing elements Receiving elements
Number of validation
interviewees including
element in top five (Q2)
Rank from
initial Delphi
(See Table 4)
Number of validation
interviewees including
element in top five (Q1)
Rank from
initial Delphi
(See Table 4)
Site safety manager 7 1 3 4
Upper management support 7 2 0 11
Worker participation and involvement 6 3 5 1
Site-specific safety plan 6 5 6 2
Frequent inspections 5 7 7 5
Training and regular safety meetings 4 6 6 6
Job hazard analyses 2 8 5 3
Safety and health orientation 2 10 1 9
Safety and health committee 0 4 2 7
Subcontractor selections and compliance 0 9 2 12
Record keeping and accident analyses 0 11 2 8
Substance abuse programs 0 12 0 10
Emergency response plan 0 13 0 13
JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / NOVEMBER 2011 / 991
to the first two questions provided very strong validation of the
relative ranking.
The objective of Question 3 was to validate the interactions that
were identified as the most significant to a construction safety pro-
gram. A higher degree of variability was expected in the responses
because the validation interviewees were asked to identify the
top five interactions out of the possible 156 pairwise interactions.
Despite the high volume of options, seven of the 10 most signifi-
cant interrelationships were identified by 25% or more of the
validation panel. Table 6 summarizes the results of the validation
interviews and compares the results with the initial Delphi analysis.
Conclusions
The purpose of this study was to quantify the interrelationships be-
tween highly effective safety program elements by using a Delphi
panel of experts and to validate the findings with interviews of an
independent panel of practicing professionals. These experts came
to the consensus that the site safety manager, worker participation
and involvement, a site-specific safety plan, and upper management
support and commitment play a central role in a highly effective
safety program. The findings of this study are unique in that they
quantify the contributions made and received by each safety pro-
gram element, rather than investigating the effectiveness of injury
prevention strategies in isolation.
These results can be used to enhance safety programs by illu-
minating the significant interrelationships that have a higher prob-
ability of increasing the effectiveness of other elements within a
safety program. The authors suggest that practicing professionals
use the findings when selecting specific injury prevention strategies
for potential integration into an existing safety program because
elements that would contribute most to the existing program can
be easily identified with a simple analysis. Furthermore, the find-
ings can be used to direct management priority and funding when
establishing a new safety program.
An interesting conclusion of this study is that many of the injury
prevention strategies that have been found to be effective in isola-
tion (i.e., upper management support and the site safety manager)
also provide a high level of synergistic effects that enhance the ef-
fectiveness of other elements. Additionally, this study shows that
for some strategies, impact on the effectiveness of an overall safety
program depends more on the synergistic effects than base-level
effectiveness. Other strategies, such as substance abuse programs,
are very effective at reducing the potential for injuries, although
they have a relatively low impact on the effectiveness of other strat-
egies. When structuring an effective safety program, one should
consider both the base-level effectiveness of individual strategies
and their influence on the other elements in the existing program.
There exist several limitations in this study. First, this study
operated under the assumption that all safety program elements are
used consistently and effectively. However, in practice, there exist
many approaches to implementing each strategy. Additionally, a
second assumption was that the interactions are cumulative and that
there exist no diminishing returns as safety programs become more
complex. This assumption may reduce the reliability of the results
if diminishing returns do exist. Finally, the study only focused on
the 13 elements already identified by previous literature, which is
now nearly 5 years old. Therefore, the potential contributions of
new safety management techniques and emerging technologies
are not included in this analysis.
Future research that explores additional injury prevention strat-
egies would further our knowledge of synergistic safety program
development. Because expert ratings of element interaction were
the primary data source for this study, the authors suggest sensi-
tivity analyses on empirical databases as a method of validation.
Furthermore, the authors believe that gaining input from construc-
tion workers on this topic would produce interesting and valuable
results. The results presented in this study have the potential to be
used to populate a decision support system that would help con-
tractors to create the most effective safety program under various
constraints. Such a decision support system would reduce the need
for firms to rely only on intuition when creating safety programs.
Acknowledgments
The authors would like to express their gratitude to the Delphi
panelists and validation interviewees for their time, dedication,
and candor throughout the study.
References
Blanning, R., and Reinig, B. (1999). “Cross-impact analysis using group
decision support systems: An application to the future of Hong Kong.”
Futures, 31(1), 39–56.
Boje, D. M., and Murnighan, J. K. (1982). “Group confidence pressures
decisions.” Manage. Sci., 28(10), 1187–1196.
Brockhoff, K. (1975). “The performance of forecasting groups in computer
dialogue and face-to-face discussion.” The Delphi method: Techniques
and applications, Addison Wesley, Reading, MA, 291–321.
Bureau of Labor Statistics (BLS). (2009). “Census of fatal occupational
injuries summary, 2007.” Economic News Release, ⟨http://www.bls
.gov/news.release/cfoi.nr0.htm⟩ (Aug. 5, 2009).
Center for Construction Research and Training. (2008). The construction
chart book: The U. S. construction industry and its workers, 4th Ed.,
CPWR, Washington, DC.
Dalkey, N. (1971). “An elementary cross-impact model.” Technol.
Forecasting Social Change, 3, 341–351.
Table 6. Pairwise Interactions Ranked in the Top Five Interactions by Two or More Validation Interviewees
Contributing element Receiving element
Number of experts that
selected interrelationship
as significant
Rank in initial
Delphi analysis
(out of 156 interactions)a
Upper management support Site safety manager 4 5
Site safety manager Job hazard analyses 3 5
Site safety manager Training and regular safety meetings 3 2
Site safety manager Site-specific safety plan 3 2
Training and regular safety meetings Worker participation and involvement 3 12
Site safety manager Frequent inspections 2 1
Frequent inspections Job hazard analyses 2 14
Upper management support Site-specific safety plan 2 5
aNote that interactions with the same rating are given the same rank.
992 / JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / NOVEMBER 2011
http://dx.doi.org/10.1016/S0016-3287(98)00109-8
http://dx.doi.org/10.1287/mnsc.28.10.1187
http://www.bls.gov/news.release/cfoi.nr0.htm
http://www.bls.gov/news.release/cfoi.nr0.htm
http://www.bls.gov/news.release/cfoi.nr0.htm
http://www.bls.gov/news.release/cfoi.nr0.htm
http://www.bls.gov/news.release/cfoi.nr0.htm
http://www.bls.gov/news.release/cfoi.nr0.htm
http://dx.doi.org/10.1016/S0040-1625(71)80022-7
http://dx.doi.org/10.1016/S0040-1625(71)80022-7
Enzer, S. (1972). “Cross-impact techniques in technology assessment.”
Futures, 4(1), 30–51.
Eschenbach, T., and Geistauts, G. (1985). “A Delphi forecast for Alaska.”
Interfaces, 15(6), 100–109.
Findley, M., et al. (2004). “Safety program elements in construction: Which
ones best prevent injuries and control related workers’ compensation
costs?” Professional Safety, 49(2), 14–20.
Gordon, T. J., and Hayward, H. (1968). “Initial experiments with the cross
impact matrix method of forecasting.” Futures, 1(2), 100–116.
Hallowell, M. R., and Gambatese, J. A. (2009). “Construction safety risk
mitigation.” J. Constr. Eng. Manage., 135(12), 1316–1323.
Hallowell, M. R., and Gambatese, J. A. (2010). “Qualitative research:
application of the Delphi method to CEM research.” J. Constr. Eng.
Manage., 136(1), 99–107.
Helmer, O. (1977). “Problems in futures research: Delphi and causal
cross-impact analysis.” Futures, 9(1), 17–31.
Hill, D. C. (2001). Construction safety management planning, American
Society of Safety Engineers, Des Plaines, IA.
Hinze, J. (2002). “Safety plus: Making zero accidents a reality.” CII
Research Rep. 160-11, University of Texas at Austin, Austin, TX.
Hinze, J. (2006). Construction safety, 2nd Ed., Prentice Hall, Englewood
Cliffs, NJ.
Jaselskis, E. J., et al. (1996). “Strategies for achieving excellence in con-
struction safety performance.” J. Constr. Eng. Manage., 122, 61–70.
Kendall, J. (1977). “Variations of Delphi.” Technol. Forecasting Social
Change, 11(1), 75–85.
Linstone, H., and Turoff, M. (1975). The Delphi method: Techniques and
applications, Addison Wesley, Reading, MA.
Liska, R., and Goodloe, S. (1993). “Zero accident techniques: A report to
the construction industry institute.” Source Document #86, University
of Texas, Austin, TX.
Meridian Research. (1994). “Worker protection programs in construction:
Final report.” Meridian Research, Silver Spring, MD.
Mitchell, R. B., and Tydeman, J. (1978). “Subjective conditional probabil-
ity modeling.” Technol. Forecasting Social Change, 11, 133–152.
Molenaar, K. R., Park, J., and Washington, S. (2009). “Framework for
measuring corporate safety culture and its impact on construction safety
performance.” J. Constr. Eng. Manage., 135(6), 488–496.
National Occupational Research Agenda (NORA). (2008). “National con-
struction agenda.” ⟨http://www.cdc.gov/niosh/NORA/comment/
agendas/construction/pdfs/ConstOct2008 ⟩ (Dec. 1, 2008).
National Safety Council (NSC). (2006). Accident facts, Itasca, IL.
Rajendran, S. (2006). “Sustainable construction safety and health rating
system.” Ph.D. dissertation, Oregon State Univ., Corvallis, OR.
Rajendran, S., and Gambatese, J. A. (2009). “Development and initial
validation of sustainable construction safety and health rating system.”
J. Constr. Eng. Manage., 135(10), 1067–1075.
Ren, H., Yang, Q., and Chen, Q. (2007). “Evaluating Olympic technologies
with the integration of Delphi and cross-impact analysis.” Advances in
Management of Technology Int. Conf. Proc., 213–219.
Rogers, M. R., and Lopez, E. C. (2002). “Identifying critical cross-
cultural school psychology competencies.” J. Sch. Psychology, 40(2),
115–141.
Sawacha, E., Naoum, S., and Fong, D. (1999). “Factors affecting safety
performance on construction sites.” Int. J. Proj. Manage., 17(5),
309–315.
Seung, H., and Diekmann, J. (2004). “Judgment-based cross-impact
method for predicting cost variance for highly uncertain projects.”
J. Constr. Res., 5(2), 171–192.
Turoff, M. (1972). “An alternative approach to cross-impact analysis.”
Technol. Forecasting Social Change, 3, 309–339.
JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / NOVEMBER 2011 / 993
http://dx.doi.org/10.1016/0016-3287(72)90023-7
http://dx.doi.org/10.1287/inte.15.6.100
http://dx.doi.org/10.1016/S0016-3287(68)80003-5
http://dx.doi.org/10.1061/(ASCE)CO.1943-7862.0000107
http://dx.doi.org/10.1061/(ASCE)CO.1943-7862.0000137
http://dx.doi.org/10.1061/(ASCE)CO.1943-7862.0000137
http://dx.doi.org/10.1016/0016-3287(77)90049-0
http://dx.doi.org/10.1061/(ASCE)0733-9364(1996)122:1(61)
http://dx.doi.org/10.1016/0040-1625(77)90017-8
http://dx.doi.org/10.1016/0040-1625(77)90017-8
http://dx.doi.org/10.1016/0040-1625(78)90026-4
http://dx.doi.org/10.1061/(ASCE)0733-9364(2009)135:6(488)
http://www.cdc.gov/niosh/NORA/comment/agendas/construction/pdfs/ConstOct2008
http://www.cdc.gov/niosh/NORA/comment/agendas/construction/pdfs/ConstOct2008
http://www.cdc.gov/niosh/NORA/comment/agendas/construction/pdfs/ConstOct2008
http://www.cdc.gov/niosh/NORA/comment/agendas/construction/pdfs/ConstOct2008
http://www.cdc.gov/niosh/NORA/comment/agendas/construction/pdfs/ConstOct2008
http://dx.doi.org/10.1061/(ASCE)0733-9364(2009)135:10(1067)
http://dx.doi.org/10.1016/S0022-4405(02)00093-6
http://dx.doi.org/10.1016/S0022-4405(02)00093-6
http://dx.doi.org/10.1016/S0263-7863(98)00042-8
http://dx.doi.org/10.1016/S0263-7863(98)00042-8
http://dx.doi.org/10.1142/S1609945104000139
http://dx.doi.org/10.1016/S0040-1625(71)80021-5
Copyright of Journal of Construction Engineering & Management is the property of American Society of Civil
Engineers and its content may not be copied or emailed to multiple sites or posted to a listserv without the
copyright holder’s express written permission. However, users may print, download, or email articles for
individual use.
