Content uploaded by Ruoyu Jin
Author content
All content in this area was uploaded by Ruoyu Jin on Apr 29, 2016
Content may be subject to copyright.
Multilevel Safety Culture and Climate Survey for Assessing
New Safety Program
Qian Chen, A.M.ASCE
1
; and Ruoyu Jin
2
Abstract: Construction safety culture is becoming a typical proactive safety performance measurement and strategy for safety improvement.
However, there is no accepted industry wide definition for safety culture. There is also no clear distinction betw een safety culture and safety
climate. So far, little research has investigated how safety programs can be developed to improve safety culture and climate, which in turn
can be used to assess the effectiveness of these programs. This paper presents a study that examined the multilevel safety culture and climate
to assess a newly launched safety program. Specifically, three different questionnaires were used to survey 71 top executives, 229 site man-
agement personnel, and 350 field workers, respectively. This study found wide acceptance of the program across all three groups , strong
management accountability in enforcing safety, and a positive general safety climate among workers. The results demonstrated the program’s
overall effectiveness in building positive safety culture and climate despite a few identified weaknesses. Based on these findings and insightful
employee feedback, this paper offers generalized recommendations to help industry practitioners develop high-quality safety programs. The
presented survey instruments and analysis methods can also be used to holistically assess safety programs and related safety culture and
climate. DOI: 10.1061/(ASCE)CO.1943-7862.0000659. © 2013 American Society of Civil Engineers.
CE Database subject headings: Construction industry; Safety; Climates.
Author keywords: Safety culture; Safety climate; Safety program; Focus 4 Hazards.
Introduction
Safety performance has long been measured by reactive indicators
such as incident rates and experience modification ratings (EMRs).
These outcome-based measurements, however, have significant
weaknesses, such as ignorance of risk exposure, inaccurate report-
ing, and variability in accident severity (Zohar 1980; Mencke l and
Carter 1985; Glendon and McKenna 1995; Thompson et al. 1998).
Over the years, researchers have increasingly regarded proactive
indicators as superior because they focus on caus es and accident
prevention. Examples of proactive indicators include hazard iden-
tification, behavior-based safety, and safety culture/climate (DeJoy
1985; Hofmann et al. 1995; Guldenmund 2000). A positive safety
culture is now widely considered a contributing factor to superior
safety performance. Safety climate, on the other hand, serves as a
diagnostic tool for improvement efforts (Diaz and Cabrera 1997;
Choudhry et al. 2007; Melia et al. 2008).
Safety culture contains subcultures within different organiza-
tional levels [Schein 1996; Grote and Kunzler 2000; Neal et al.
2000; National Occupational Research Agenda (NORA) 2008].
Nevertheless, little research has assessed safety subcultures and
investigated any consistencies and differences between them.
Learning within an organization could be hindered by a lack of
alignment among subcultures (Schein 1996). Also, although some
safety programs have been found to be effective in reducing acci-
dents, their potential impacts on safety culture and climate are
not fully understood (Peyton and Rubio 1991; Center to Protect
Workers’ Rights 1993; Chen and Jin 2012). In addition, despite
the correlation between safety climate and the effectiveness of
safety programs (Zohar 1980; Cabrera et al. 1997), it has not been
clear how safety climate can be used to evaluate these programs.
This study aimed to fill these gaps by (1) performing a multi-
level survey of safet y culture and climate to assess the effectiveness
of a newly launched safety program; and (2) conducting a compar-
ative statistical study to check the alignment of safety cultur e and
climate in an organization. In this study, three multidimensional
questionnaires were used to survey top-level executives, site man-
agement personnel, and workers, respectively. This paper presents
survey results and identifies consistencies and differences across
the three organizational levels. It provides deeper insights into
how the safety program was implemented, perceived, and put into
effect. It also makes generalized recommendations to help ind ustry
practitioners develop strong safety programs and strengthen their
safety culture and climate.
Literature Review
Safety Culture and Safety Climate
There is some confusion regarding what safety culture and safety
climate are and whether they are distinct or interchangeable terms.
Several comprehensive reviews on existing definitions of safety
culture and climate demonstrate this confusion (Guldenmund 2000;
Wiegmann et al. 2002; Choudhry et al. 2007). However, a number
of researchers have made efforts to distinguish safety climate from
safety culture as two separate concepts, while still emphasizin g
the close relationship between them (Ekvall 1983; Schein 1992;
1
Assistant Professor, Construction Systems Management Program,
The Ohio State Univ., 590 Woody Hayes Dr., Columbus, OH 43210
(corresponding author). E-mail: chen.1399@osu.edu
2
Ph.D. Candidate, Construction Systems Management Program, The
Ohio State Univ., 590 Woody Hayes Dr., Columbus, OH 43210. E-mail:
jin.188@buckeyemail.osu.edu
Note. This manuscript was submitted on March 2, 2012; approved on
November 27, 2012; published online on November 29, 2012. Discussion
period open until December 1, 2013; separate discussions must be
submitted for individual papers. This paper is part of the Journal of Con-
struction Engineering and Management, Vol. 139, No. 7, July 1, 2013.
© ASCE, ISSN 0733-9364/2013/7-805-817/$25.00.
JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / JULY 2013 / 805
Berends 1996; Cox and Flin 1998; Mearns et al. 2003; Mohamed
2003; NORA 2008).
According to Mohamed (2003 ), safety culture is a top-down
organizational attribute approach to addressing safety management.
The National Occupational Research Agenda (NORA), a public
partnership program for innovative research and workplace safety
and health, also defined safety culture as the organizational prin-
ciples, norms, commitments, and values related to the operation of
safety and health. Safety culture determines the relative importance
of safety and other workplace goals (NORA 2008). Safety climate,
on the other hand, is concerned with workers’ perception of the role
of safety in the workplace and their attitude toward safety (Cox and
Flin 1998; NORA 2008). Safety culture is reflected in the safety
management system (SMS) and safety climate (Mearns et al. 2003).
Because of this relationship, workplace safety climate can be mea-
sured to assess organizational safety culture. In fact, safety culture
established at the organizational level needs to be perceived as valid
and important by employees in order to succeed (Cooper 2000).
Safety culture contains subcultu res. Grote and Kunzler (2000)
divided participants into different groups in their safety culture/
climate survey based on whether or not the employees had man-
agement functions. NORA (2008) also defined two levels of safety
culture within an organization: the top management level and the
supervisor/subunit level. The top management level develops and
defines safety policies and procedures, and the supervisor/subunit
level executes and practices these policies. Furthermore, Schein
(1996) classified three levels of subcultu re as executive culture, en-
gineering culture, and operators’ culture. Melia et al. (2008) studied
safety climate among four primary safety agents: the organization,
supervisors, coworkers, and workers. A close relationship was
found between the organization and supervisors. The relati onship
of safety culture/climate between management personnel and
workers, however, appeared to be more complex. So far, multilevel
safety culture and climate studies are sparse.
Measurement Tools for Safety Culture and Safety
Climate
Both qualitative and quantitative methods have been used to assess
safety culture and climate. Qualitative methods obtain in-depth
information through employee observations, focus group discus-
sions, historical information reviews, and case studies (Wreathall
1995). Quantitative methods numerically measure or score safety
culture or climate through interviews, surveys, and Q-sorts
(Wreathall 1995). Questionnaire surveys have been widely used to
study safety culture and climate because they are easy to perform
and cost very little (Dedobbeleer and Béland 1991; Mattila et al.
1994a; McDonald et al. 2000; Zohar 2000; Mearns et al. 2003).
Various safety culture/climate dimensions (or factors) have been
identified in different industry domains. After reviewing existing
safety literature, Zohar (1980) developed an eight-factor 40-item
questionnaire. It was then used to survey safety climate among
factory workers in the metal fabrication, food processing, chemical,
and textile industries in Israel. These factors included the perceived
importance of the safety-traini ng program, management attitude
toward safety, and the level of risk at the workplace.
Brown and Holmes (1986) tested Zohar’s measurement tool by
surveying a different sample from U.S. production workers. Be-
cause of inconsistent findings, they introduced a three-factor model
that included employee perception of management safety concerns,
management actions on safety, and of physical risks. Dedobbeleer
and Béland (
1991) fu rther tested this model by conducting a safety
climate survey among construction workers. Based on the results,
they created a two-factor safet y climate model, which examined
management commitment and worker involvement. A different
survey targeting safety management personnel on offshore oil and
gas installations was administered by Mearns et al. (2003). This
survey recorded personnel’s satisfaction with safety activities,
workforce involvement, communication about safety, safety-
related statements, and unsafe behavior.
Table 1 summarizes the 15 most frequently used safety culture/
climate dimensions. These are based on the review of 33 related
papers, including Zohar (1980), Brown and Holmes (1986), Cooper
and Phillips (1994), Diaz and Cabrera (1997), Cooper (2000), and
others. Most of these studies did not distinguish safety culture and
climate but saw them as interchangeable. Therefore, these dimen-
sions were used in both safety culture and climate surveys.
Relationship between Safety Culture/Climate and
Safety Management Programs
It is widely known that an effective safety management program
improves workplace safety performa nce while lowering cost, pre-
venting project delays, and building company image (Anton 1989;
Michaud 1995; Abdelhamid and Everett 2000; Rowlinson 2003;
Findley et al. 2004). Such a program could also potentially enhance
an organization’s safety culture or climate (Hakkinen 1995; Martin
1995; Aksorn and Hadikusumo 2008; Oh and Sol 2008).
According to Cooper (2000), an organization’s SMS (including
the safety program) is an environmental element of its safety cul-
ture. A close relationship also exists between safety management
and safety climate (Zohar 1980; Brown and Holmes 1986;
Dedobbeleer and Béland 1991; Coyle et al. 1995; Cabrera et al.
1997). Mattila et al. (1994b) found that foremen’s effort to improve
workers’ safety behavior led to a safer work environment. Zohar
(2000) specified that safety implementation at the supervisory level
could cause variation in safety climate among workers. It was also
noticed that safety culture could be improved by adopting a lead -
ership style that pays more attention to group members’ welfare
(Zohar 2002).
Cabrera et al. (1997) observed the relationships among safety
climate, safety management, and safety performance. They con-
cluded that a safety climate scale (e.g., scores) could help evaluate
the essential features of safe companies. In particular, safety climate
was found to correlate with the effectiveness of safety programs
(Zohar 1980). Furthermore, researchers identified multiple safety
culture/climate dimensions as success facto rs of safety programs
(Zohar 1980; Loushine et al. 2006; Aksorn and Hadikusumo
2008
). For example, Loushine et al. (2006) identified management
Table 1. Fifteen Most Frequently Used Safety Culture/Climate
Dimensions
Safety culture/climate dimensions Frequency (%)
Management attitudes/commitments 21 (64)
Safety procedures/policies/rules 15 (45)
Perception of risk 11 (33)
Attitudes toward safety 10 (30)
Communication 10 (30)
Safety training 9 (27)
Craft involvement 9 (27)
Safety prevention/investigation 9 (27)
Work environment and pressure 7 (21)
Status of safety committee/officers 6 (18)
Accountability 5 (15)
Workers’ perception of safety management 5 (15)
Safety implementation 4 (12)
Importance of safety compared to other goals 4 (12)
Workplace safety perception 3 (9)
806 / JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / JULY 2013
commitment and communication as the most frequently studied
success factors. Other major factors included the identification
of hazards or unsafe behaviors, measures of safety culture/climate,
and employee involvement. Aksorn and Hadikusumo (2008) found
16 critical success factors in four categories: worker involvement,
safety prevention and control system, safety arrangement, and
management commitment.
Fig. 1 displays a framework that was built upon the knowledge
gained through the literature review. This framework was based on
a comprehensive understanding of safety culture, safety climate,
safety program and their interrelationships. It served as the theo-
retical foundation for this study. The aforementioned safety culture/
climate dimensions and success factors were also incorporated into
the survey instruments developed in this study.
Background of the New Safety Program Investigated
The safety program studied in this paper was launched by a
regional general contractor (GC) in the U.S. building construction
industry at the end of May 2008 and has been ongoing since then. It
aims to reduce injuries and worker exposure to the Occupational
Safety and Health Admini stration (OSHA) Focus 4 Hazards.
These hazards are categorized as: falls, electrocution, struck-by,
and caught-in or -between [Occupational Safety and Health
Administration (OSHA) 2011]. In 2009, accidents falling into these
four categories were responsible for 63% of all fatal injuries in the
U.S. private construction industry [Bureau of Labor Statistics
(BLS) 2010]. The safety program introduced in this study focuses
on increasing the safety awareness and accountability of the GC’s
employees, all hired subcontractors (subs), and material suppliers
in Focus 4 Hazards-related fields while achieving positive changes
in employees’ safety attitude and behavior.
The primary components of this safety program are illustrated in
Fig. 2. The program consists of three basic elements: (1) 100% eye
protection for all the workers on the GC’s jobsites; (2) daily “hud-
dle” (tool-box) meetings for the GC and subs’ crews to go through
their daily plans and potential safety hazards; and (3) an account-
ability system centered on enforcing safety by reporting workers’
Safety culture
Organizational principles, norms,
commitments, and values related to
the operation of safety and health.
Safety subculture
Safety principles, norms,
commitments, and values accepted
by middle management subgroups.
Safety climate
Shared worker perception and
attitude toward safety, reflecting
safety culture in the workplace.
Top management
Develops and defines
company safety policies
and procedures.
Middle management
Executes and practices
safety policies and
procedures.
Worker
Follows safety rules and
perceives the importance of
safety in the workplace.
Organizational hierarchy
Safety program (part of the SMS)
Affect the implementation of the
safety program and be used to
diagnose its effectiveness
Impact safety subculture
Impact safety climate
An environmental element of
safety culture
Affect the development and
implementation of the safety program
Affect the implementation of the
safety program
Fig. 1. Theoretical framework for understanding safety culture, safety climate, and safety program
20 Non-negotiable Behaviors
Three Basic Elements
1) Working 6 feet above the ground or higher with no approved fall protection;
2) Exposed holes left unprotected or improperly protected;
3) Blatant, clear and/or intentional misuse of fall protection equipment/systems;
4) Altering or disabling any component of a fall protection system;
5) Improperly using a stepladder.
6) Disregarding red danger tape or barrier/crossing without proper authorization;
7) Dropping debris off buildings without proper chutes or approved alternates;
8) Riding on equipment not designed for multiple passengers;
9) Using damaged equipment that may inadvertently fire or altering safety guards;
10) Exposing person(s) to overhead struck-by hazards.
11) Working in unprotected trenches or excavations (5 feet or greater);
12) Not surveying for utilities before digging;
13) Operator not barricading/protecting swing radius and/or pinch points;
14) Operator exposing employee(s) to an overhead load during excavation;
15) Reckless operation of equipment.
16) Working within 10 feet from power lines;
17) Exposing oneself or others to live bare electrical conductors (>50 volts);
18) Working on live electric without proper PPE and procedures;
19) Unauthorized altering, by-passing or removing safety guards;
20) Not using a GFCI when using temporary wiring (including extension cords).
Consequences of Violation
Second-time Violation
- For the GC’s employees: separated
from the GC for 30 days without pay
and benefits;
- For subs’ employees: Dismissed
from all the GC’s projects for 1 year.
First-time Violation
1) Removed for the day when
violation is observed and reported;
2) Return to work the following day
morning and lead the “huddle”
meeting on non-negotiables;
3) Sign the engagement letter of the
safety program.
Subs with More Violations
1) Create a correction plan;
2) Provide a full-time safety
specialist or competent person;
3) Contract terminated for non-
compliance.
1) 100% eye protection for
all the workers on the GC’s
jobsites;
2) Daily “huddle” meeting
required for the GC’ and
subs’ crews;
3) Accountability for
accidents, incidents, and near
misses centered on 20 non-
negotiable unsafe behaviors.
FallsStruck-byCaught-in-
between
Electrical
Fig. 2. Primary components of the studied safety program
JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / JULY 2013 / 807
violations falling into the 20 non-negotiable unsafe behaviors iden-
tified in the new safety program. These 20 non-negotiables are
evenly distributed into the four Focus 4 Hazard categories. The
consequences of a violation can vary depending on whether this
is the first or second time the employee has committed a violation.
Subs with a higher number of violations mu st implement required
improvement measures at no additional cost to the owner or to the
GC. Otherwise, their contracts will be terminated.
This program differs from many other safety programs by
requiring project managers, superintendents, safety coordinators,
engineers, and other site management personnel to report safety
violations and enforce consequences. Ideally, this will lead to
shared safety responsibilities and better safety enforcement. Chen
and Jin (2012) presented this program in more detail in an earlier
study and confirmed its effectiveness in reducing accidents/
incidents and unsafe behaviors for a 27-month study period. How-
ever, the manner in which this program impacted safety culture and
safety climate in the workplace has not been investigated.
Research Methods
The goal of this research was to assess the safety program’s effec-
tiveness in building positive safety culture and climate. Data was
collected through a multilevel safety culture and climate survey.
Specifically, three different questionnaires were developed to
study: (1) safety culture of the GC’s top executives; (2) subculture
of site management personnel on the GC’s jobsites; and (3) safety
climate of craft workers, respectively. Each of these questionnaires
consisted of similar dimensions, namely awareness (only for
workers), acceptance, accountability, cost/schedule impact, and
general safety culture/climate. Questions were asked in multiple
choice, Likert scale, and open-ended formats.
Table 2 provides further details on how these dimensions were
defined and how they related to the existing safety culture/climate
dimensions summarized in Table 1. For each dimension, some ex-
ample questions (one per each type of questionnaire) were also pro-
vided. The questionnaires were designed to collect comprehensive
information about the program-related safety culture and climate so
that a holistic assessment could be performed. Although the struc-
ture of each questionnaire was simplified by covering only four or
five specific dimension s, the questions were structured in a manner
that also covered many of the other safety climate/culture dimen-
sions defined previously.
During the questionnaire development process, all researchers
passed online qualification training for human subject-related re-
search. The questionnaires and survey procedures wer e reviewed
and approved by the University Institutional Review Board (IR B).
To ensure wide worker participation and meaningful feedback, re-
searchers administered the surveys face-to-face. They visite d 31
construction sites in four states, where the GC’s six regional offices
were located. Top executives (senior managers and their superiors)
and site management personnel (project managers, superintend-
ents, engineers, safety coordinators, foremen, and trade leaders)
were surveyed either online or face-to-face, when their schedules
allowed. All questionnaire surveys were conducted anonymously
to protect participants’ rights and privacy.
In this study, the individual dimensions of each type of ques-
tionnaire were analyzed separately. Specifically, multiple choice
and Likert scale questions were evaluated by first calculating the
Table 2. Further Details on the Safety Culture and Climate Dimensions Examined in This Study
Dimensions examined Previously identified dimensions related Example questions
Awareness Employees’ awareness of the
safety program and its
components
Safety procedures/policies/rules; communication;
safety training; craft involvement; accident
prevention
W: Are you aware of the 20 non-negotiable
behaviors?
Acceptance Employees’ perceptions of the
program’s content,
implementation, impact,
benefits, pros and cons
Management attitudes/commitments; status of
safety committee/officers; accountability; safety
procedures/policies/rules; attitudes toward safety;
craft involvement; accident prevention/
investigation; workers’ perception of safety
management; safety implementation; workplace
safety perception
T: Do you agree that the safety program was
introduced and implemented correctly and
effectively?
S: Were you provided with enough training to
understand the program?
W: Do you feel that the GC’s safety program
contributes to a safer work site?
Accountability Employees’ perceptions of their
safety responsibilities, safety
enforcement, and consequences
of violations
Safety procedures/policies/rules; communication;
safety training; craft involvement; work
environment and pressure; accountability;
workers’ perception of safety management;
importance of safety compared with other goals
T: Do you agree that the responsibilities
associated with your job have increased due to the
implementation of the safety program?
S: Are you comfortable with explaining the
program to employees and business partners?
W: Do you understand how you will be held
responsible for a non-negotiable violation?
Cost/schedule
impact
Employees’ perceptions of the
program’s impact on cost,
schedule, and productivity
Safety procedures/policies/rules; work
environment and pressure; safety implementation;
importance of safety compared with other goals
T: Does the safety program affect the daily
operating budget of your projects?
S: Are there any tasks that take longer to
accomplish since the inception of the program?
W: Do you believe the implementation of the
program slows down work progress?
General safety
culture/climate
Employees’ general safety
awareness, attitude, and risk
perception
Perception of risk at the workplace; attitudes
toward safety; craft involvement; accident
prevention/investigation; work environment and
pressure; accountability; workplace safety
perception
T: Based on your perception and observations,
which specific trade(s) listed below tend to have
more safety violations than others?
S: Is the safety violation related to the experience
and training of the employee?
W: Would you risk getting hurt to
“get the job
done”?
Note: The capitalized and bolded letter denotes from which questionnaire the example question was drawn. T, S, and W denote top executives, site
management, and workers, respectively.
808 / JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / JULY 2013
percentage of respondents to each of the predefined answers. Next,
scores ranging from −1 to 1 were assigned to answers that repre-
sented negative, neutral, and positive perceptions of the safety
program or its impact. Then, the mean scores (falling between
−1 and 1) and standard deviations (SD) for these questions were
computed. For some multiple choice questions that only contained
neutral and positive answers, their mean scores ranged from 0 to 1
instead of from −1 to 1. Based on the calculated proportions or
mean values, statistical analyses were conducted using the infer-
ence concerning proportions or two means one-tailed test (Johnson
2005). The analyses were performed at a significant level of 0.05 to
compare survey results for consistencies and differences.
The open-ended responses collected in this study were analyzed
using content analysis. Specifically, repeatedly used keywords or
phrases were identified through a method of coding for explicit
concepts. Relational analysis was performed to verify the relevance
of such keywords or phrases as well as to determine their relation-
ships and categorization. Lastly, frequency counts and statistical
analysis were conducted to generate valuable insights.
Results and Discussions
This section presents survey results from the three selected groups
following each individual safety culture or climate dimension. The
consistencies and differences identified across organizational levels
are also discussed.
Demographics of Survey Participants
During the period from November 2010 to February 2011, the re-
search team collected 650 completed questionnaires from 71 top
executives, 229 site management personnel, and 350 workers (in-
cluding 11 Spanish-speaking workers). Table 3 shows the regional
distribution of survey participa nts.
Of the site management personnel surveyed, 69% were from
the GC’s organization; the remaining 31% were hired by subs.
Subs’ personnel primarily consisted of foremen for different trades,
e.g., carpentry, electrical, plumbing, and sheet metal. In contrast,
only 31% of workers surveyed worked for the GC; the
remaining 69% were subs’ workers on the GC’s jobsites. Figs. 3
and 4 display these workers’ age groups and trade distribution.
Dimension One: Awareness
In this study, only workers were asked about their awareness of the
new safety program. Their responses are summarized in Table 4.A
high general awareness of the program existed among workers
surveyed (based on a mean value of 0.99 on a 0-to−1 scale). How-
ever, only 74% of them displayed full awareness of the three basic
elements, and only 33% showed full awareness of the 20 non-
negotiable behaviors defined in the program. This indicates the
increased difficulty for workers to learn and remember detailed
safety rules. Workers who admitted some awareness of the 20
non-negotiable behaviors (N ¼ 325) were asked to identify how
they obtained the information. It was found that 53% learned from
both training (classroom and/or orientation video) and visuals
(posters, handouts, and hardhat stickers), whereas 35% learned
from training only, and 12% learned from visuals or other methods.
This study investigated four factors that could affect workers’
learning outcome related to the 20 non-negotiables. First, workers
were divided into subgroups according to (1) how they obtained the
information (i.e., what types of training they had received); (2) their
age range; (3) whether they had committed a violation before; and
(4) whether they had worked for the GC before. Then, the percent-
ages of workers who were aware of either “all 20 non-negotiables”
or “all or most of the 20 non-negotiables” were calculated, respec-
tively, for each subgroup. Finally, statistical analyses were per-
formed to determine whether any difference existed between the
learning outcomes of the paired subgroups.
As shown in Table 5, when the percentage of workers knowing
all 20 non-negotiables was analyzed, the joint use of training and
visuals in workers’ learning process was proven to be statistically
more effective than “visuals or other.” Older workers (above 50)
were also more effective in learning non-negotiables than young
and middle-aged workers. Among workers who had committed a
violation before, a higher percentage was fully aware of all non-
negotiables. Similar conclusions were reached after assessing the
percentage of workers knowing all or most of the non-negotiables.
Furthermore, “training” was found to be less effective than “training
and visuals” but more effective than “visuals or other.” Workers who
had worked for the GC before also knew more non-negotiables.
These results suggest that the use of multiple formats in safety
education can improve the learning effectiveness and outcome.
The effectiveness of safety training for young and middle-aged
workers needs to be assessed and strengthened. Implementing
an accountability system (in which workers are held responsible
for behavioral violations) and maintaining long-term business
Table 3. Number of Participants in Each Region
Participant R1 R2 R3 R4 R5 R6 Total
Top executives 6 11 7 8 23 16 71
Site management 20 53 28 45 48 45 229
Workers 33 65 29 53 82 88 350
Younger
than 20
3%
20-29
27%
30-39
32%
40-49
23%
Above
50
15%
Fig. 3. Age groups of surveyed workers
0%
2%
4%
6%
8%
10%
12%
14%
16%
Percentage of workers
Fig. 4. Workers’ trade distribution
JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / JULY 2013 / 809
relationships with subs can also improve workers’ awareness of
detailed safety rules.
Dimension Two: Acceptance
All participants were surveyed with slightly different questions re-
garding their acceptance of the safety program. As Table 6 shows,
both executives and site management personnel had positive per-
ceptions of the safety program’s introduction and implementation.
Specifically, site management personnel agreed that they had re-
ceived enough training to understand the program. Communication
and collaboration between the office and the field was adequate.
Among the surveyed executives, 97% either agreed or strongly
agreed that the program had improved the company’s safety perfor-
mance, resulting in a high average score of 0.77 on a scale of −1
to 1. Among those who were involved in executive-level discussions
with subs regarding safety violations (N ¼ 44), 89% either agreed
or strongly agreed that such meetings had improved subs’ safety
performance. In addition, most executives (81%) perceived a pos-
itive change in employees’ safety attitude. To verify the accuracy of
these perceptions, this study analyzed information on the 20 non-
negotiable safety violations provided by the GC to see if there was
any noticeable improvement in behavior-based safety performance.
Fig. 5 shows the numbers of non-negotiable violations observed
during the 44-month study period (May 5, 2008 to December
2011). Of 1,113 total violations reported on the GC’s jobsites,
the top five mo st violated rules (9 44 or 84.8%) were the 6-ft rule,
improper use of step ladder, misuse of fall protection, altering/
disabling fall protection system, and GFCI (ground fault circuit
interrupter) protection as highlighted in the figure. The first four
were all associated with fall hazards, which were causes for nearly
one-third of fatal injuries in the U.S. private construction industry
(BLS 2010). There were six rules that were violated much less
frequently (<10 times), such as not barricading the swing radius,
exposing workers to overhead loads, and working within 10 ft of
power lines. It was noted that in some violations, more than one
individual was involved. Therefore, the total number of workers
who committed violations (1,374) was slightly larger than the total
number of violations observed. In addition, 41 workers committed
second-time violations. Of these, one came from the GC and the
rest came from subs. More second-time violations occurred in
2008 (17) and 2009 (11) than 2010 (5) and 2011 (8).
A trend analysis was performed to study annual safety violation
rates (SVR) for all 20 non-negotiables and each of the five most
violated rules during the 44-month study period. Introduced by
Chen and Jin (2012), the SVR analysis measured the number of
violations based on the annual working hours (40 h per week
for 50 weeks) of 100 full-time workers: SVR = number of viola-
tions × 200,000/employee hours worked. As illustrated in Fig. 6,
the observed SVRs for the 20 non-negotiable violations had re-
duced annually, leading to an overall reduction of 2.62 or 25.2%
(from 10.41 in 2008 to 7.79 in 2011). For the five most violated
rules, although some fluctuations existed in their annual SVRs, the
overall trends showed slight drops. The sum of these SVRs for
2011 was 6.45, a reduction of 2.08 (or 24.4%) from the sum for
2008 (8.53). This shows that the drop in SVRs associated with the
five most violated ru les was the primary contributor in the overall
SVR reduction (2.08 out of 2.62, 79.4%).
The dashed lines in Fig. 6 show annual SVRs calculated for the
GC, subs, and both the GC and subs combined. These SVRs were
based on the number of workers who committed safety violations
for each group. Obviously, the GC’s annual SVRs were signifi-
cantly lower than subs’. The SVR trends for the three groups were
also downward. The subs’ annual SVRs rose in 2009 (slightly) and
2010, but eventually dropped by 30.5% in 2011. This indicates that
the GC’s program had an overall positive impact on subs’ behavior-
based safety performance, even if it took longer to achieve notice-
able results. The results suggest that the safety program improved
contractors’ safety performance as perceived. It should, therefore,
be continuously promoted and enforced, especially among subs ’
workers, to sustain its positive effect.
Table 4. Workers’ Awareness of the Safety Program
Program awareness item Answer, assigned score, and percentage of respondents Range Mean SD
Aware of the GC’s safety program Yes = 1 (99%), no = 0 (1%) 0 to 1 0.99 0.10
Aware of the program’s three
basic elements
All of them = 1 (74%), two of them = 0.67 (14%), one of them = 0.33 (11%),
none of them = 0 (1%)
0 to 1 0.87 0.24
Aware of the 20 non-negotiable
behaviors
All of them = 1 (33%), most of them = 0.67 (40%), a couple of them = 0.33 (20%),
no idea about them = 0 (7%)
0 to 1 0.65 0.30
How the information of 20
non-negotiables was obtained
Training (classroom and/or orientation video (35%), visuals (posters, handouts, etc.)
(7%), both training and visuals (53%), other (5%)
NA NA NA
Note: SD = standard deviation; NA = not applicable.
Table 5. Factors That Might Affect Workers ’ Awareness of 20 Non-Negotiable Behaviors
Analysis item
How the information was obtained Age range
Committed a
violation before
New to the
GC’s jobs
Training and
visuals
N ¼ 174 (a)
Training
only
N ¼ 113 (b)
Visuals or
others
N ¼ 38 (c)
Above 50
N ¼ 51
(d)
30–49
N ¼ 191
(e)
Below 30
N ¼ 106
(f)
Yes
N ¼ 27
(g)
No
N ¼ 319
(h)
Yes
N ¼ 74
(i)
No
N ¼ 271
(j)
Aware of all 36% 32% 21% 43% 31% 25% 52% 29% 27% 32%
Z-score 0.757ða=bÞ 1.265ðb=cÞ 1.793ða=cÞ 1.647ðd=e
Þ 0.986ðe=fÞ 9.348ðd=fÞ 2.450ðg=hÞ −0.837ði=jÞ
p-value 0.224 0.102 0.037 0.049 0.162 0.000 0.007 0.201
Aware of all
or most
83% 75% 50% 86% 73% 64% 89% 71% 64% 74%
Z-score 1.681ða=bÞ 2.912ðb=cÞ 4.448ða=cÞ 1.994ðd=eÞ 1.549ðe=fÞ 2.871ðd=fÞ 2.010ðg=hÞ −
1.807ði=jÞ
p-value 0.047 0.002 0.000 0.023 0.061 0.002 0.022 0.035
Note: Values in bold indicate that statistically significant differences existed between the compared subgroups.
810 / JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / JULY 2013
The safety program’s impact on the five predefined project
elements (the GC’s safety culture, subs’ safety culture, employee
involvement with safety, housekeeping, and productivity) was
assessed using a 5-level Likert scale, which ranged from very neg-
ative (− 1 ) to very positive (1). The perceptions from top execu-
tives and site management personnel were consistent. As shown
in Table 7, both groups rated the program’s influence on each of
the five elements positively. The five elements were also ranked
in the same order, with the impact on the GC’s safety culture
being the highest and the impact on productivity being the lowest,
based on the average impact scores they received. Some statisti-
cally significant differences existed between the two levels of
management. The top executives observed a higher positive
impact of the safety program on both the GC’ and subs’ safety
cultures than the site management personnel did (see the bolded
p-values).
Table 6. Survey Participants’ Acceptance of the Safety Program
Program acceptance item Answer, assigned score, and percentage of respondents Range Mean SD
Top executives
Program introduced and implemented
correctly and effectively
Strongly agree = 1 (51%), agree = 0.5 (37%), neutral = 0 (10%),
disagree = −0.5 (2%)
−1 to 1 0.69 0.37
Improved safety performance (GC) Strongly agree = 1 (57%), agree = 0.5 (40%), neutral = 0 (3%) −1 to 1 0.77 0.28
Positive impact of meeting with subs Strongly agree = 1 (32%), agree = 0.5 (57%), neutral = 0 (11%) −1 to 1 0.60 0.32
Impact on the GC’s safety culture Very positive = 1 (60%), positive = 0.5 (40%) −1 to 1 0.80 0.25
Impact on employee involvement with
safety
Very positive = 1 (34%), positive = 0.5 (49%), neutral = 0 (16%),
negative = −0.5 (1%)
−1 to 1 0.58 0.37
Impact on subs’ safety culture Very positive = 1 (21%), positive = 0.5 (63%), neutral = 0 (16%) −1 to 1 0.53 0.31
Impact on housekeeping Very positive = 1 (13%), positive = 0.5 (49%), neutral = 0 (32%),
negative = −0.5 (6%)
−
1 to 1 0.35 0.39
Impact on productivity Very positive = 1 (9%), positive = 0.5 (41%), neutral = 0 (41%),
negative = −0.5 (9%)
−1 to 1 0.25 0.39
Result in a positive attitude change Yes = 1 (81%), no = 0 (19%) 0 to 1 0.81 0.49
Site management personnel
Enough training provided for them to
understand the program
Strongly agree = 1 (60%), agree = 0.5 (31%), neutral = 0 (6%),
disagree = −0.5 (2%), strongly disagree = −1 (1%)
−1 to 1 0.74 0.38
Proper communication between the
office and the field
Strongly agree = 1 (41%), agree = 0.5 (42%), neutral = 0 (13%),
disagree = −0.5 (2%), strongly disagree = −1 (2%)
−1 to 1 0.60 0.43
Impact on the GC’s safety culture Very positive = 1 (54%), positive = 0.5 (38%), neutral = 0 (6%),
negative = −0.5 (1%), very negative = −1 (1%)
−1 to 1 0.71 0.37
Impact on employee involvement
with safety
Very positive = 1 (35%), positive = 0.5 (50%), neutral
= 0 (13%),
negative = −0.5 (1%), very negative = −1 (1%)
−1 to 1 0.59 0.36
Impact on subs’ safety culture Very positive = 1 (21%), positive = 0.5 (51%), neutral = 0 (23%),
negative = −0.5 (5%)
−1 to 1 0.44 0.40
Impact on housekeeping Very positive = 1 (25%), positive = 0.5 (41%), neutral = 0 (31%),
negative = −0.5 (2%), very negative = −1 (1%)
−1 to 1 0.43 0.43
Impact on productivity Very positive = 1 (12%), positive = 0.5 (36%), neutral = 0 (39%),
negative = −0.5 (10%), very negative = −1 (3%)
−1 to 1 0.22 0.46
Impact on enforcing safety Make it easier = 1 (72%), no difference = 0 (21%), make it
more difficult = −1 (7%)
−1 to 1 0.64 0.62
What best describes the program Substantial improvement in safety = 1 (18%), workers are more aware of
the GC’s safety expectations = 0.67 (63%), safety awareness somewhat
increased = 0.33 (15%), no difference in safety on my projects
= 0 (4%)
0 to 1 0.65 0.24
Provide a platform to involve subs
and all workers
Strongly agree = 1 (47%), somewhat agree = 0.5 (49%), no = 0 (4%) 0 to 1 0.70 0.28
Daily huddle meeting keeps crew
focused on safety
Strongly agree = 1 (34%), somewhat agree = 0.5 (51%),
indifferent = 0 (13%), not at all = 0 (2%)
0 to 1 0.59 0.34
Workers
Awareness of safety increased Significantly = 1 (44%), somewhat = 0.5 (52%), not at all = 0 (4%) 0 to 1 0.70 0.28
Impact on a work site Contributed to a safer site = 1 (96%), not contributed to a safer site = 0 (4%) 0 to 1 0.96 0.20
Consistency of safety enforcement Enforced consistently = 1 (76%), not enforced consistently = 0 (24%) 0 to 1 0.76 0.43
How the safety program is enforced Genuine attempt to keep workers safe = 1 (75%), non-positive
perception(s) = 0 (25%)
0 to 1 0.75 0.43
Effect of “100% eye protection” Very effective = 1 (61%), effective = 0.5 (23%), neutral = 0 (9%),
ineffective = −0.5 (2%), very ineffective = −1 (4%)
−1 to 1 0.67 0.52
Effect of “20 non-negotiable violations” Very effective = 1 (41%), effective = 0.5 (26%), neutral = 0 (25%),
ineffective = −0.5 (4%), very ineffective = −1 (4%)
−1 to 1 0.48 0.54
Effect of
“daily huddle meeting” Very effective = 1 (33%), effective = 0.5 (30%), neutral = 0 (26%),
ineffective = −0.5 (4%), very ineffective = −1 (7%)
−1 to 1 0.39 0.58
Compared to other company’s safety
program
Better = 1 (51%), the same = 0 (43%), not as effective = −1 (6%) −1 to 1 0.45 0.60
Compared to other company’s jobsites Much safer = 1 (32%), safer = 0.5 (30%), the same = 0 (32%),
less safer = −0.5 (4%), much less safer = −1 (2%)
−1 to 1 0.41 0.50
Note: SD = standard deviation.
JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / JULY 2013 / 811
Among site management survey respondents, 72% confirmed
that the program made enforcing safety easier. When asked to de-
scribe the effects of the program, 96% held a positive view. They
believed that the program increased workers’ awareness of the GC’s
safety expectations (63%), improved safety on jobsites (18%), or
increased workers’ general safety awareness to some degree (15%).
Also, 96% strongly or partially agreed that the safety program pro-
vided a platform to involve subs and their employees and helped
with the implementation of safety plans. Site management had a
moderately positive perception of the effectiveness of daily huddle
meetings in keeping the crew focused on safety throughout the day.
The lower portion of Table 6 shows the data analysis results for
workers’ acceptance of the safety program. Ninety-six percent of
workers had a positive view of the safety program’s effectiveness in
increasing their safety awareness on the GC’s jobsites. The same per-
centage felt that the program contributed to a safer work site. Up to
75% of workers saw the program as a genuine attempt to keep them
safe. Among those who had worked on the GC’s jobsites previously
(N ¼ 257), 76% believed that the enforcement of the program was
consistent. The inconsistencies mentioned by workers were primarily
related to equal treatment of workers, accountability for unsafe acts,
and the enforcement of personal protective equipment (PPE).
Workers were asked to assess the effectiveness of the three basic
elements of the safety program in creating a safer work environ-
ment. Their feedback was collected using Likert scale questions.
The analysis results showed that the “100% requirement for safety
glasses” element was most effective followed by “20 non-negotiable
violations” and “required daily huddle meetings.” Workers were
also asked to compare this program with safety programs from
other companies they had worked for (or were currently working
603
168
86
48
13
26
14
13
12
7
25
12
11
2
1
39
16
9
44
0
100
200
300
400
500
600
6-foot rule
Improper use of step ladder
Misuse of fall protection
Altering/disabling fall protection system
Exposed hole
Crossing safety barriers
Riding on equipment
Using damaged tools or altering guards
Exposing workers to overhead struck-by
Dropping debris off building
Reckless operation of equipment
Not surveying utilities
Working in unprotected trench
Exposing workers to overhead loads
Not barricading the swing radius
GFCI protection
Exposing oneself to live bare electric
Working on live electric
Altering, by-passing or removing guards
Working within 10 ft of power lines
Fall Struck-by Caught-in-
between
Electrocution
Number of Violations Observed
Fig. 5. Number of 20 non-negotiable violations observed (May 2008
through December 2011)
10.41
9.91
9.26
7.79
12.70
12.09
11.80
8.79
2.69
1.04
1.23
0.90
15.35
15.37
16.29
11.32
0.00
1.00
2.00
3.00
4.00
5.00
6.00
7.00
8.00
9.00
10.00
11.00
12.00
13.00
14.00
15.00
16.00
17.00
2008 2009 2010 2011
Safety Violation Rate (SVR)
20 non-negotiable violations observed
6-foot rule
Improper use of step ladder
Misuse of fall protection
Altering/disabling fall protection system
GFCI protection
Workers comitting violations (GC & subs)
GC's workers
Subs' workers
Fig. 6. Trend of SVRs (2008–2011)
Table 7. Comparison of Management’s Perceptions on the Impact of the Safety Program
Item
Perception from top executives
Perception from site
management
Alignment between
two levels
Rank Mean SD Rank Mean SD Z-score p-value
Impact on the GC’s safety culture 1 0.80 0.25 1 0.71 0.37 2.355 0.009
Impact on employee involvement with safety 2 0.58 0.37 2 0.59 0.36 −0.202 0.421
Impact on subs’ safety culture 3 0.53 0.31 3 0.44 0.40 2.017 0.022
Impact on housekeeping 4 0.35 0.39 4 0.43 0.43 −1.578 0.057
Impact on productivity 5 0.25 0.39 5 0.22 0.46 0.451 0.326
Note: SD = standard deviation. Values in bold indicate that statistically significant difference existed between two levels of management.
812 / JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / JULY 2013
for). Approximately 51% felt that this program was better and 62%
felt much safer or somewhat safer on this GC’s jobsites.
Their written feedback disclosed that workers were impressed
about the clear safety rules defined by the program. As indicated by
many of them, their companies (primarily trade contractors) also
had similar safety requirements related to PPE (e.g., hard hats,
gloves, safety glasses, face shield, and 100% tie off), huddle meet-
ings, or specific rules about using ladders or operating equipment.
However, none of these companies had specified safety rules as
detailed as the GC’s 20 non-negotiables nor was safety enforced
through a well-designed accountability system like the GC’s. Some
subs’ workers further pointed out that their companies had already
started or were considering implementing a similar safety program.
Dimension Three: Accountability
All survey participants were asked about accountability under the
new safety program. Table 8 shows the survey results. The study
found that more than half of the executives (55%) felt increased
work responsibilities owing to the implementation of the safety
program. The result was very similar to the perception of site
management (56% felt the same way).
Up to 94% of executives said that they held their managers
accountable for enforcing safety, either aggressively or to some
degree. Ninety-nine percent believed that site managers were pre-
pared for their safety responsibilities. These perceptions were
consistent with the feedback from site management personnel—
83% expressed a strong feeling of accountability for workers’
safety and the remaining 17% felt a slight feeling of accountability.
Additionally, 98% of site management personnel replied that they
would comply with the program’s requirements and enforce safety
on jobsites regardless of their comfort level.
This percentage dropped to 92% when management personnel
were asked if they would actually address non-negotiable behav-
iors. Twenty-seven percent of site mana gement felt uncomfortable
with this nontraditional safety management role (including 25%
who felt uncomfortable but still enforced safety and 2% who felt
uncomfortable and tried to avoid it). It is thus necessary to find
ways to increase their comfort level. Regarding the procedures for
reporting violations and managing accountability, 81% of respond-
ents said that they were easy or very easy to implement, whereas
19% were uncomfortable with them. Feelings of discomfort could
cause site management to avoid enforcing safety. Enhancing these
procedures would increase the percentage of people enforcing
safety and reduce their workloads.
Accountability questions for workers tested their understanding
of safety responsibilities and consequences of violations. As Table 8
shows, 89% of workers claimed that they understood how they
would be held accountable if any non-negotiables were violated.
However, after verifying further, this study found that, in reality,
only 59% and 36% of workers knew the correct consequences for
first- and second-time violations, respectively. A higher percentage
Table 8. Survey Participants’ Accountability Related to the Safety Program
Program accountability item Answer, assigned score, and percentage of respondents Range Mean SD
Top executives
Involvement in safety discussions with
subs’ executives
More than 5 times (14%), 3–5 times (18%), 1–3 times (33%),
no involvement (35%)
NA NA NA
Responsibilities increase due to the
safety program
Strongly agree = 1 (23%), agree = 0.5 (32%), neutral = 0 (29%),
disagree = −0.5 (9%), strongly disagree = −1 (7%)
−1 to 1 0.28 0.57
Holding managers accountable for
enforcing safety
Aggressively = 1 (54%), somewhat = 0.5 (40%), no = 0 (6%) 0 to 1 0.74 0.30
Site leaders/managers’ preparation for
implementing the program
Adequately prepared = 1 (79%), somewhat prepared = 0.5 (20%),
no = 0 (1%)
0 to 1 0.89 0.20
Site management personnel
Frequency on explaining the program Daily (34%), weekly (45%), monthly (10%), other (11%) NA NA NA
Time spent on the program (per week) >10 h (13%), 7–10 h (10%), 4–7 h (24%), 0–4 h (37%), NA (16%) NA NA NA
Increase of job responsibility Significantly = 1 (19%), somewhat = 0.5 (37%), no = 0 (44%) 0 to 1 0.40 0.34
Feeling of accountability for workers’
safety
Strong feeling = 1 (83%), a little bit = 0.5 (17%), no = 0 (0%) 0 to 1 0.91 0.20
Comfort level of explaining the program Very comfortable and try to promote = 1 (38%), comfortable and do
what is required = 0.67 (53%), uncomfortable but do what is
required = 0.33 (7%), uncomfortable and try to avoid =
0 (2%)
0 to 1 0.76 0.22
Comfort level of enforcing safety Comfortable and enforce = 1 (89%), uncomfortable but
enforce = 0.67 (9%), comfortable but try to avoid = 0.33 (2%),
uncomfortable and try to avoid = 0 (0%)
0 to 1 0.95 0.14
Comfort level of addressing a
non-negotiable behavior
Comfortable and enforce = 1 (67%), uncomfortable but
enforce = 0.67 (25%), comfortable but try to avoid = 0.33 (6%),
uncomfortable and try to avoid = 0 (2%)
0 to 1 0.86 0.23
Procedures for reporting and managing
the accountability elements
Very easy to implement = 1 (15%), easy to implement = 0.5 (66%),
somewhat uncomfortable = −0.5 (18%), makes you very
uncomfortable = −1 (1%)
−1 to 1 0.38 0.47
Workers
Understanding of safety accountability Know the responsibility of violating the 20 Non-Negotiable
behaviors = 1 (89%), do not know the responsibility = 0 (11%)
0 to 1 0.90 0.35
Consequence of the first-time violation Removed from jobsite for one day = 1 (59%), verbal warning = 0 (22%),
written warning = 0 (10%), do not know = 0 (8%)
0 to 1 0.59 0.49
Consequence of the second-time violation Suspended for 30 days for the GC’s workers and banned from working
on any GC’s projects for one year for subs’ workers = 1 (36%),
wrong answer selected = 0 (51%), do not know = 0 (13%)
0 to 1 0.35 0.48
Note: SD = standard deviation; NA = not applicable.
JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / JULY 2013 / 813
of workers may have known the right consequences for first-time
violations because they occurred more frequently. Because the
occurrence of these violations is expected to reduce further, con-
tinuing education is still the best way to increase workers’ aware-
ness of their safety accountability and consequences of violations.
Dimension Four: Cost/schedule Impact
All participants were asked for their opinions whether the safety
program affected project cost, schedule and productivity, and to
what extent. As Table 9 shows, only 20% of executives thought
that the program affected their daily operating budget and cited that
additional time or cost was needed for training, compliance inspec-
tions, and safety materials and equipment. Of site management
personnel surveyed, up to 48% felt this impact. A statistically sig-
nificant difference was confirmed between the two proportions
(p-value = 0.000).
Most executives thought that the accountability system had
a moderate impact on field productivity, cost, and schedule.
Comparatively, site management felt a slightly greater but still
moderate impact. However, of the workers who directly performed
construction tasks, 66% felt that the program slowed down their
work progress. The negative influence perceived was statistically
greater than what the site management personnel had estimated
(p-value = 0.004).
Despite these potential disadvantages, many executives still
found the safety program a worthwhile effort. They believed that
it would prevent accidents and save the GC money in the long run.
Similarly, site management personnel deemed the program neces-
sary due to the priority it placed on safety. Although both levels of
management had a strong belief in the value of the safety program,
it would still be helpful for the program’s implementation if the
potential cost and schedule impact at the production level could
be properly addressed. It is necessary to solicit employee feedback
Table 9. Survey Participants’ Perceptions of the Cost/Schedule Impact of the Safety Program
Program cost/schedule impact item Answer, assigned score, and percentage of respondents Range Mean SD
Top executives
Impact on daily operating budget Yes = 1 (20%), no = 0 (80%) 0 to 1 0.20 0.40
Impact of the accountability system
on productivity, cost, and schedule
Substantially but necessary = 1 (8%), somewhat but necessary = 0.67 (59%),
substantially and process needs to be changed = 0.33 (3%), not at all = 0 (30%)
0 to 1 0.49 0.34
Site management personnel
Tasks take longer to finish Yes = 1 (55%), no = 0 (45%) 0 to 1 0.55 0.50
Impact on budget Yes = 1 (48%), no = 0 (52%) 0 to 1 0.48 0.50
Impact of the accountability system
on productivity, cost, and schedule
Substantially but necessary = 1 (13%), somewhat but necessary = 0.67 (67%),
substantially and process needs to be changed = 0.33 (4%), not at all = 0 (16%)
0 to 1 0.59 0.29
Workers
Slow down work progress (schedule) Significant = 1 (7%), moderately = 0.67 (22%), somewhat or significantly
at times = 0.33 (37%), no = 0 (34%)
0 to 1 0.33 0.30
Note: SD = standard deviation.
Table 10. General Safety Culture/Climate among Survey Participants
General safety culture/climate item Answer, assigned score, and percentage of respondents Range Mean SD
Top Executives
Trades having most violations
a
Drywall (54%), steel (46%), masonry (43%), electric (41%),
mechanical (19%)
NA NA NA
Site management personnel
Trades having most violations
a
Steel (52%), drywall (39%), masonry (34%), electric (27%),
mechanical (23%)
NA NA NA
Related to subs’ safety program Yes = 1 (48%), no = 0 (14%), do not know/NA = 0 (38%) 0 to 1 0.48 0.50
Related to the experience and training Yes = 1 (49%), no = 0 (13%), do not know/NA = 0 (38%) 0 to 1 0.49 0.50
Related to the risk level of the work Yes = 1 (52%), no = 0 (12%), do not know/NA = 0 (36%) 0 to 1 0.52 0.50
Workers
Trades having most violations
a
Roofing (40%), steel (32%), drywall (23%), masonry (23%),
electric (17%)
NA NA NA
Related to subs’ safety program Yes = 1 (27%), no = 0 (10%), do not know/NA = 0 (63%) 0 to 1 0.27 0.44
Related to the experience and training Yes = 1 (38%), no = 0 (9%), do not know/NA = 0 (53%) 0 to 1 0.38 0.49
Related to the risk level of the work Yes = 1 (42%), no = 0 (9%), do not know/NA = 0 (49%) 0 to 1 0.42 0.49
The likelihood of taking risks getting a job done No = 1 (84%), yes = 0 (16%) 0 to 1 0.84 0.37
Reasons for taking risk (for those who admitted
to the likelihood of taking risks; N ¼ 56)
You feel that is what your supervisors expect (21%), fear of losing
your job (19%), there is no other way to do it (41%), just a
risk taker (19%)
NA NA NA
The feeling of responsibility for coworkers’ safety Strong feeling = 1 (47%), somewhat = 0.5 (46%), no = 0 (7%) 0 to 1 0.70 0.31
Likelihood of addressing peers’ unsafe behavior Very likely = 1 (35%), likely = 0.5 (44%), not likely = 0 (21%) 0 to 1 0.56 0.37
Note: SD = standard deviation; NA = not applicable.
a
For this question, the percentage was calculated by using the number of times each trade being mentioned divided by the number of workers who responded to
the question. Workers were allowed to select more than one trade.
814 / JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / JULY 2013
to continuously improve the safety program, especially for the
detailed safety rules.
Dimension Five: General Safety Culture and Climate
Table 10 shows the analysis results for general safety culture and
climate questions. First, all survey participants were asked to
specify which trade(s) tended to have more safety violations than
others. The researchers believed that an accurate perception can
help executives and site management personnel set up their prior-
ities for safety management. Adequate awareness would also draw
workers’ attention and help create a safer work environment. This
study found consistent rankings for four of the top five trades. Spe-
cifically, drywall, steel, masonry, and electrical were picked by all
three groups. The only difference was that executives and site man-
agement believed the mechanical trade to have the most potential
violations, whereas workers selected roofing. This caused the
mechanical trade to drop out of the workers’ top five. This discrep-
ancy needs to be further studied and understood to address igno-
rance of potential risks.
This study sought the insight of site management personnel and
workers on why some trades had more violations than others. These
insights were based on three predefined risk factors: subs’ safety
programs, employee’s experience and training, and the risk level of
the work. In general, approximately half the site management per-
sonnel saw a connection between the frequent violations and each
of these risk factors, whereas the other half did not see such a con-
nection or simply answered “did not know.” The percentage of
workers who saw such a connection was lower, ranging from
27–42%. The statistical analysis results showed that the perception
of both groups on the connection between frequent violations and
the risk level of the work were relatively consistent. The primary
difference was that, compared to workers, more site management
personnel thought that frequent violations were due to subs’ safety
program (p-value = 0.000) and workers’ experience and training
(p-value = 0.001). It would be beneficial to learn the true rationale
for frequent violations so proper safety management actions can be
taken. However, when the rationale is not clear such as in this case,
the accountability system would be very necessary for reducing
safety violations.
Workers were surveyed for their general safety climate. As
Table 10 shows, 84% said they would not risk getting hurt to get
the job done. Also, 93% expressed either a strong or mild feeling
of responsibility for one another’s safety on the GC’s jobsites. Addi-
tionally, 79% of workers said that they were likely or very likely to
address an unsafe behavior or situation involving their peer(s). These
indicate a rather positive general safety climate in the workplace.
Of the workers who admitted to the likelihood of taking risks
(N ¼ 56), 40% pointed out that they were likely to take safety short-
cuts for higher productivity owing to management pressure. This
implies that a company’s organizational safety culture (whether
workplace safety is a top priority) could impact workers’ safety
attitude and behavior. Another 41% felt that the risks were inherent
in the nature of their job and there was no other way to do it. The
remaining 19% thought that they were simply risk-takers by nature.
Continuing education on safety procedures and the importance of
safety would be necessary to further improve workers’ safety climate.
Feedback for Continuous Improvement
This study solici ted feedback from survey participants for continu-
ous improvement of the safety program. The collected responses
were analyzed and the results are shown in Table 11. All three
groups suggested offering training/continuing education to new and
old employe es and continuously improving the program, including
its accountability system and the 20 non-negotiable behaviors.
Compared with top executives, site management personnel and
workers were more concerned about the implementation of the pro-
gram on jobsites. Consistent advice from site management and
workers included enhancing safety enforcement and consistency
of implementation, improving communication between manage-
ment and workers, and providing safety incentives and awards.
Table 11. Suggestions for Continuous Improvement of the Safety Program
Suggestions for continuous improvement
Frequency count Safety culture/climate dimension
Executives
(N ¼ 71)
Site
management
(N ¼ 229)
Workers
(N ¼ 350)
Awareness
of the
program Acceptance Accountability
Cost/
schedule
impact
General safety
culture/
climate
Seek feedback from subs/employees to
continuously improve the program
10 —— ××
Reevaluate and update 20 non-negotiable
behaviors
9 — 21 ×××
Improve the accountability system — 7 — ×
Provide training/continued education about
the program
8148×××
Enhance enforcement and consistency — 12 23 ×× × ×
Provide safety incentives/rewards to workers 1 7 7 ×× ×
Enhance communication between
management and workers
— 79×× × ×
Enhance safety meeting/evaluation/review — 7 — ×
Keep the program updated and continuously
enforced
— 6 — ×
Simplify reporting — 2 — ××
Lighten up penalties ——12 ×
Make safety rules clear —— 7 ××
Other suggestions
a
4912—— — — —
Total number of suggestions 32 71 99 88 120 165 30 96
Average number of suggestions per
participant in a group
0.45 0.31 0.28 —— — — —
a
Other suggestions represented those that were only mentioned by one participant.
JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / JULY 2013 / 815
In addition, although site management emphasized improving peo-
ple’s accountability, workers wanted to be clear about safety rules
through better communication with management. They believed
that a better understanding could strengthen their accountability
and compli ance.
Table 11 also shows which dimensions of safety culture and
climate could be potentially improved by applying these sugges-
tions and how many suggestions were related to each individual
dimension. The largest number of suggestions (165) were toward
improving accountability, which is the key to making the safety
program work at the application level. This was followed by accep-
tance (120), general safety culture/climate (96), awareness (88), and
cost/schedule impact (30). A statistical analysis of the average num-
bers of suggestions per participant among the three survey groups
showed that top executives gave more suggestions for continuous
improvement than site management (p-value = 0.015), and there
was no difference between site management and workers. To fur-
ther improve the applicability and acceptance of the safety program,
executives should strongly encourage employees to freely comment
on the program while establishing an effective feedback channel.
Conclusions and Recommendations
This paper presented a multilevel survey of safety culture and cli-
mate associated with a new safety program launched by a general
contractor (GC) in the U.S. building construction industry. In this
study, three different multidimensional questionnaires were devel-
oped and used to survey the GC’s top executives, site management
personnel on the GC’s jobsites, and field workers for safety culture,
subculture, and climate, respectively. The completed questionnaires
from the three groups were then analyzed, and any consistencies
and differences were identified. The presented survey instruments
and analysis methods can help practitioners holistically assess their
safety programs and related safety culture and climate.
In this study, a high general awareness of the safety program
was identified among workers. However, their awareness of the de-
tailed safety rules required by the program was inadequate. Several
factors were found to have some positive impact on workers’ learn-
ing outcome related to the 20 non-negotiable behaviors. These in-
cluded the joint use of training and visuals to educate workers,
older age, workers’ past violation experience, and their increased
familiarity with the safety program by repeatedly working on the
GC’s jobsites.
Overall, the safety program and its primary elements were well
accepted by respondents from all three organizational levels. Both
executives and site management perceived that the program had
been introduced and implemented effectively. It had improved the
GC’s safety performance, safety culture, employee involvement in
safety, and the safety performance of subcontractors (subs) to some
degree. Most workers also felt an increase in safety awareness and a
safer work environment.
Ninety-four percent of executives said that they held site man-
agers accountable for enforcing safety, either aggressively or
slightly. This resulted in strong accountability among site manage-
ment personnel in complying with the program’s requirements.
However, nearly one-fourth of them still did not feel comfortable
enforcing safety. The majority of workers knew that they would be
held accountable for safety violations. Unfortunately, a higher
percentage of them did not know the right consequences for non-
negotiable violations. This indicates a need for continuing educa-
tion about the accountability system.
Workers and site management personnel who were directly in-
volved in the execution or supervision of field operations felt a
greater negative cost/schedule impact from the program. Regard-
less of these feelings, some top executives and management staff
still explicated that the program was a worthwhile effort to enforce
safety, prevent accidents, and save the company money in the
long run.
All three survey groups had consistent perceptions on trade(s)
that might commit more safety violations than others. However,
there was not much consensus in terms of what factors might lead
to higher violations. A generally positive safety climate was found
on the GC’s jobsites. Workers were less likely to take risks to get
the job done and more likely to address unsafe behavior in their
peers. This showed that the safety program was effective in build-
ing positive safety culture and climate.
The research findings summarized in this paper, coupled with
employee feedback, can be used to help the GC further improve
its safety program. To benefit other contractors or the industry
as a whole in improving field safety programs, this study made
the following generalized recommendations:
• An effective safety program needs to provide employees with
clear expectations on jobsite safety rules, a strong sense of ac-
countability for the safety of themselves and others, and ways to
improve their safety awareness and attitude. Feedback from em-
ployees and subs needs to be solicited to continuously improve
the program and keep the safety rules updated and well received.
• To maintain and increase employees’ awareness of a newly
launched or existing safety program, training/continued educa-
tion in multiple formats (e.g., classroom training, videos, and
visuals) will be needed to keep the program fresh, well under-
stood, and constantly promoted.
• Using hars h penalties to hold workers responsible for safety
violations may cause fear in the workplace. Integrating positive
reinforcement (e.g., safety bonuses, incentives, or awards) into a
safety program can help maintain high morale and loyalty
among workers, which will increase the acceptance of a safety
program.
• Management commitment and worker involvement are the keys
to successful safety programs. It is important to enforce safety
consistently and treat all workers equally. In addition, effective
communication between site management and workers needs
to be established to help workers understand safety rules for
better compliance.
• It is necessary not only to educate workers about taking personal
responsibility for their safety but also to promote feelings of re-
sponsibility for coworkers’ safety in the workplace. To create
friendly, trusting, and safer work environments, workers should
be strongly encouraged to remind their peers of safety violations
or unsafe acts. This will improve general safety climate among
workers.
• Finally, it is important to understand that differences may exist
between a GC and subs’ safety culture and safety programs
(e.g., in specific safety rules). It will take some time to famil-
iarize subs’ workers with the GC’s program and minimize thei r
reluctance to accept new safety rules. Therefore, the GC’s safety
program needs to be continuously promoted and enforced.
References
Abdelhamid, T. S., and Everett, J. G. (2000). “Identifying root cause of
construction accidents.” J. Constr. Eng. Manage., 126(1), 52–60.
Aksorn, T., and Hadikusumo, B. H. W. (2008). “Critical success factors
influencing safety program performance in Thai construction projects.”
Saf. Sci., 46(4), 709–727.
Anton, T. J. (1989). Occupational safety and health management, 2nd Ed.,
McGraw-Hill, New York.
816 / JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / JULY 2013
Berends, J. J. (1996). “On the measurement of safety culture.” Master
thesis, Eindhoven Univ. of Technology, Eindhoven, Netherlands.
Brown, R. L., and Holmes, H. (1986). “The use of a factor analytic pro-
cedure for assessing the validity of an employee safety climate model.”
Accid. Anal. Prev., 18(6), 455–470.
Bureau of Labor Statistics (BLS). (2010). “Table A-9. Fatal occupational
injuries by event or exposure for all fatalities and major private industry
sector, All United States, 2009.” 〈http://www.bls.gov/iif/oshwc/cfoi/
cftb0249.pdf〉 (Apr. 28, 2010).
Cabrera, D. D., Isla, R., and Vilela, L. D. (1997). “An evaluation of safety
climate in ground handling activities.” Aviation Safety, Proc., IASC-97
Int. Aviation Safety Conf., H. M. Soekkha, ed., Zeist, Netherlands,
255–268.
Center to Protect Workers’ Rights. (1993). An agenda for change: Report
of the National Conference on Ergonomics, Safety, and Health in
Construction, Center to Protect Workers’ Rights, Washington, DC.
Chen, Q., and Jin, R. (2012). “Safety4Site commitment to enhance jobsite
safety management and performance.” J. Constr. Eng. Manage., 138(4),
509–519.
Choudhry, R. M., Fang, D., and Mohamed, S. (2007). “The nature of safety
culture: A survey of the state-of-the-art.” Saf. Sci., 45(10), 993–1012.
Cooper, M. D. (2000). “Towards a model of safety culture.” Saf. Sci., 36(2),
111–136.
Cooper, M. D., and Phillips, R. A. (1994). “Validation of a safety climate
measure.” Proc., BPS Occupational Psychology Conf., Leicester, UK.
Cox, S., and Flin, R. (1998). “Safety culture: Philosopher’s stone or man of
straw?” Work Stress, 12(3), 189–201.
Coyle, I. R., Sleeman, S. D., and Adams, N. (1995). “Safety climate.”
J. Saf. Res., 26(4), 247–254.
Dedobbeleer, N., and Béland, F. (1991). “
A safety climate measure for
construction sites.” J. Saf. Res., 22(2), 97–103.
DeJoy, D. (1985). “Attributional processes and hazard control management
in industry.” J. Saf. Res., 16(2), 61–71.
Diaz, R. I., and Cabrera, D. D. (1997). “Safety climate and attitude as
evaluation measures of organizational safety.” Accid. Anal. Prev.,
29(5), 643–650.
Ekvall, G. (1983). “Climate, structure and innovativeness of organizations:
A theoretical framework and an experiment.” Rep. 1, FArådet, the
Swedish Council for Management and Organizational Behavior,
Stockholm, Sweden.
Findley, M., Smith, S. M., Kress, T., Petty, G., and Enoch, K. (2004).
“Safety program elements in construction.” Prof. Saf., 49(2), 14–21.
Glendon, A. I., and McKenna, E. F. (1995). Human safety and risk man-
agement, Chapman and Hall, London.
Grote, G., and Kunzler, C. (2000). “Diagnosis of safety culture in safety
management audits.” Saf. Sci., 34(1–3), 131–150.
Guldenmund, F. W. (2000). “The nature of safety culture: A review of
theory and research.” Saf. Sci., 34(1–3), 215–257.
Hakkinen, K. (1995). “A learning-by-doing strategy to improve top man-
agement involvement in safety.” Saf. Sci., 20(2–3), 299–304.
Hofmann, D. A., Jacobs, R., and Landy, F. (1995). “High reliability process
industries: Individual, micro, and macro organizational influences on
safety performance.” J. Saf. Res., 26(3), 131–149.
Johnson, R. A. (2005). Miller and Freund’s probability and statistics for
engineers, 7th Ed., Pearson Prentice Hall, NJ.
Loushine, W. T., Hoonakker, P. L. T., Carayon, P., and Smith, M. J. (2006).
“Quality safety management in construction.” Total Qual. Manage.,
17(9), 1171–1212.
Martin, K. H. (1995). “Improving staff safety through an aggression
management program.” Arch. Psychiatr. Nurs., 9(4), 211–215.
Mattila, M., Hyttinen, M., and Rantanen, E. (1994a). “Effective supervisory
behavior and safety at the building site.” Int. J. Ind. Ergon., 13(2),
85–93.
Mattila, M., Rantanen, E., and Hyttinen, M. (1994b). “The quality of work
environment, supervision and safety in building construction.” Saf. Sci.,
17(4), 257–268.
McDonald, N., Corrigan, S., Daly, C., and Cromie, S. (2000). “Safety man-
agement systems and safety culture in aircraft maintenance organiza-
tions.” Saf. Sci., 34(1–3), 151–176.
Mearns, K., Whitaker, S. M., and Flin, R. (2003). “Safety climate, safety
management practice and safety performance in offshore environ-
ments.” Saf. Sci., 41(8), 641–680.
Melia, J., Mearns, K., Silvia, A., Silvia, M., and Luisa Lima, M. (2008).
“Safety climate responses and the perceived risk of accidents in the con-
struction industry.” Saf. Sci., 46(6), 949–958.
Menckel, E., and Carter, N. (1985). “The development and evaluation of
accident prevention routines: A case study.” J. Saf. Res., 16(2), 73–82.
Michaud, P. A. (1995). Accident prevention and OSHA compliance,
1st Ed., CRC Press, Boca Raton, FL.
Mohamed, S. (2003). “Scorecard approach to benchmarking organizational
safety culture in construction.” J. Constr. Eng. Manage., 129(1), 80–88.
National Occupational Research Agenda (NORA). (2008). “National
construction agenda for occupational safety and health research and
practice in the U.S. construction sector.” N ORA Construction
Sector Council. 〈http://www.cdc.gov/niosh/nora/comment/agendas/
construction/pdfs/ConstOct2008.pdf 〉 (Oct. 11, 2011).
Neal, A., Griffin, M. A., and Hart, P. M. (2000). “
The impact of organi-
zational climate on safety climate and individual behavior.” Saf. Sci.,
34(1–3), 99–109.
Occupational Safety and Health Administration (OSHA). (2011). “Con-
struction Focus Four training.” U.S. Dept. of Labor, Washington, DC,
〈http://www.osha.gov/dte/outreach/construction/focus_four/index.html〉
(Apr. 2, 2011).
Oh, J. I. H., and Sol, V. M. (2008). “The policy program improving occupa-
tional safety in the Netherlands: An innovative view on occupational
safety.” Saf. Sci. 46(2), 155 –163.
Peyton, R. X., and Rubio, T. C. (1991). Construction safety practices and
principles, Van Nostrand Reinhold, New York.
Rowlinson, S. (2003). Hong Kong construction: Safety management and
law, 2nd Ed., Sweet and Maxwell Asia, Hong Kong.
Schein, E. H. (1992). Organizational culture and leadership, 2nd Ed.,
Jossey-Bass, San Francisco.
Schein, E. H. (1996). “Three cultures management: The key to organiza-
tional learning.” Sloan Manage. Rev., 38(1), 9–20.
Thompson, R. C., Hilton, T. F., and Witt, L. A. (1998). “Where the safety
rubber meets the shop floor: A confirmatory model of management
influence on workplace safety.” J. Saf. Res., 29(1), 15–24.
Wiegmann, D. A., Zhang, H., Thaden, T. V., Sharma, G., and Mitchell, A.
(2002). “A synthesis of safety culture and safety climate research.”
Technical Rep. No. ARL-02-3/FAA-02-2, Univ. of Illinois at Urbana-
Champaign, Savoy, IL.
Wreathall, J. (1995). “Organizational culture, behavior norms, and safety.”
Proc., Int. Topical Meeting on Safety Culture in Nuclear Installations,
IAEA, Vienna, Austria, 24–28.
Zohar, D. (1980). “Safety climate in industrial organizations: Theoretical
and applied implications.” J. Appl. Psychol., 65(1), 96–102.
Zohar, D. (2000). “A group-level model of safety climate: Testing the effect
of group climate on microaccidents in manufacturing jobs.” J. Appl.
Psychol., 85(4), 587–596.
Zohar, D. (2002). “The effects of leadership dimensions, safety climate, and
assigned priorities on minor injuries in work groups.”
J. Organ. Behav.,
23(1), 75–92.
JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / JULY 2013 / 817
CopyrightofJournalofConstructionEngineering&ManagementisthepropertyofAmerican
SocietyofCivilEngineersanditscontentmaynotbecopiedoremailedtomultiplesitesor
postedtoalistservwithoutthecopyrightholder'sexpresswrittenpermission.However,users
mayprint,download,oremailarticlesforindividualuse.
