Accident Analysis and Prevention 92 (2016) 130138

Contents lists available at ScienceDirect

Accident Analysis and Prevention

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

Leading and lagging indicators of occupational health and safety: The

moderating role of safety leadership
Cathy Sheehan ,1 , Ross Donohue 1 , Tracey Shea 1 , Brian Cooper 1 , Helen De Cieri 1
Department of Management, Monash University, P.O. Box 197, Cauleld East, VIC 3145 Australia

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

Article history: In response to the call for empirical evidence of a connection between leading and lagging indicators of
Received 8 December 2015 occupational health and safety (OHS), the rst aim of the current research is to consider the association
Received in revised form 10 March 2016 between leading and lagging indicators of OHS. Our second aim is to investigate the moderating effect
Accepted 21 March 2016
of safety leadership on the association between leading and lagging indicators. Data were collected from
Available online 6 April 2016
3578 employees nested within 66 workplaces. Multi-level modelling was used to test the two hypotheses.
The results conrm an association between leading and lagging indicators of OHS as well as the moder-
Keywords:
ating impact of middle management safety leadership on the direct association. The association between
Leading indicators
Lagging indicators
leading and lagging indicators provides OHS practitioners with useful information to substantiate efforts
Occupational health and safety within organisations to move away from a traditional focus on lagging indicators towards a preventative
Safety leadership focus on leading indicators. The research also highlights the important role played by middle managers
and the value of OHS leadership development and investment at the middle management level.
2016 Elsevier Ltd. All rights reserved.

1. Introduction
Occupational health and safety (OHS) measurement relies heav-
ily on lagging indicators, such as incidents of workplace injury,
as these measures provide important feedback information about
deciencies and safety incidents that have occurred (Reiman and
Pietikinen, 2012). Lagging OHS indicators are, however, a reac-
tive measurement approach to safety management and measure
events or outcomes that have already happened (Hopkins, 2009;
Laitinen et al., 2013; Reiman and Pietikinen, 2012). As such,
lagging indicators are failure-focused (Sinelnikov et al., 2015).
Recent research emphasises a more proactive evaluation of OHS
activity that emphasises leading indicators, or inputs, that allow
organisations to predict safety concerns and that may reduce the
likelihood of an OHS incident occurring (Grabowski et al., 2007;
Lingard et al., 2011; Reiman and Pietikinen, 2012). Leading indi-
cators can be thought of as eliminating or controlling the precursors
to harm and as such offer organisations the opportunity to detect
Corresponding author.
E-mail addresses: cathy.sheehan@monash.edu (C. Sheehan),
ross.donohue@monash.edu (R. Donohue), tracey.shea@monash.edu (T. Shea),
brian.cooper@monash.edu (B. Cooper), helen.decieri@monash.edu (H.D. Cieri).
1
We acknowledge the support provided for this research project by the WorkSafe
Victoria and the Institute for Safety Compensation and Recovery Research, Australia.
We acknowledge Dr Benjamin Amick III, Institute for Work & Health, Ontario Canada,
for his comments on the development of this research.
and mitigate risks or risk increases before an OHS incident occurs
or a hazardous state is reached (Sinelnikov et al., 2015).
Sinelnikov et al. (2015), in their study of the state of knowledge
and practice on the use of leading indicators of OHS, explained that
although there is increasing interest in leading indicators, there is
a need for further evidence of the link between leading and lag-
ging indicators. These authors also noted the potential enabling
impact of leadership in terms of implementing leading indica-
tors. As recognised in the social information processing perspective
(SIP), leadership behaviour is an important determinant of the
development of employee job attitudes and behaviours (Chen et al.,
2013). The aim of the current research is to address the research
gaps noted above by considering: 1) the association between lead-
ing and lagging indicators of OHS, and 2) the moderating effect of
safety leadership on the association between leading and lagging
indicators.
The structure of the paper is as follows. First, the literature on
leading and lagging indicators of OHS will be reviewed. Hypotheses
are then developed in relation to how leading and lagging indicators
are likely to be associated and whether this association is moder-
ated by safety leadership. Second, the methods used for the study
are described. The third section of the paper presents the results
of the study, while the fourth section provides a discussion of the
studys results and contribution. The nal section of the paper pro-
vides an overview of the studys limitations and avenues for future
research.

http://dx.doi.org/10.1016/j.aap.2016.03.018
0001-4575/ 2016 Elsevier Ltd. All rights reserved.
 C. Sheehan et al. / Accident Analysis and Prevention 92 (2016) 130138
2. Theory
The following sections review the literature on leading and lag-
ging indicators of OHS and, drawing from a social information
processing (SIP) perspective, an argument is made for the moder-
ating effect of safety leadership on the association between leading
and lagging indicators.
2.1. Leading and lagging indicators of OHS
Injuries and accidents in the workplace can engender potentially
damaging consequences for individual employees and have serious
work performance outcomes at the organisational level (Battaglia
et al., 2015; LeBeau et al., 2014). Lagging indicator data provide nec-
essary information on safety performance related to injuries and
accidents that can motivate people to work on improving safety
performance (Reiman and Pietikinen, 2012). Relevant lagging data
include workplace injuries and near misses (Goldenhar et al., 2003;
Li et al., 2013; Probst et al., 2013), often based on self-reported
incidents (Wachter and Yorio, 2014).
Sinelnikov et al. (2015) have explained that leading indicators,
in contrast, can eliminate or control the precursors to harm and
provide early warning signals of potential failure. Leading indica-
tors are associated with active, positive steps that organisations
can take to avoid an OHS incident (Baker et al., 2007; Blair and
OToole, 2010; Grabowski et al., 2007; Lingard et al., 2011). Lag-
ging indicators on the other hand are failure focussed and measure
OHS incidents that have already happened (Hopkins, 2009). Lead-
ing indicators are valuable therefore as they enable organisations
to identify and correct deciencies to prevent or mitigate the worst
effects of injuries or damage.
Despite the potential worth associated with leading indica-
tors of OHS, denitions of the construct are unclear (Reiman and
Pietikinen, 2012). Writers, for example, have used a variety of
terms to depict leading OHS activity including upstream, head-
ing, positive, and predictive indicators (Hinze et al., 2013). One
approach to dening leading indicators is to focus on how they
can be differentiated from lagging or trailing indicators (Dyreborg,
2009; Hopkins, 2009; Kjelln, 2009). The distinction is not, how-
ever, without its complications. It is possible that a lagging indicator
may also act as a leading indicator if, for example, it is able to predict
another OHS outcome or event (Dyreborg, 2009).
Hopkins (2009) provided a considered discussion of leading
and lagging terminology within the context of personal and pro-
cess safety. Personal safety problems refer to problems that affect
individuals and the term lagging typically relates to injuries and
fatalities. Process safety hazards, on the other hand, are those aris-
ing from the processing activity in which a plant may be engaged
and result in damage to the plant and have the potential to gener-
ate multiple fatalities. For such events, a lagging indicator relates
to harm and failure and may include a major catastrophic event
such as an explosion or a re. These, however, are rare events and
it becomes difcult to create a meaningful measure over time. In
the case of personal safety, the distinction between leading and lag-
ging indicators is somewhat less problematic. In this context, the
term lagging indicator generally refers only to measures of OHS
incidents, such as reported OHS incidents, unreported OHS inci-
dents, and near misses. In contrast, leading indicators are those that
directly measure aspects of the OHS management system, such as
the frequency or timeliness of audits.
Having established the importance of leading indicators, we
draw attention to ongoing discussion about the content of leading
indicator domains. The following list of leading indicators repre-
sents a synthesis of the leading indicator literature that highlights
specic domains (also see Shea et al., 2016).
131
Accountability for OHS that involves a proactive OHS work-
place culture and emphasises a sense of shared responsibility
and accountability for OHS is important. Such a culture promotes
active scrutiny and transparency in reporting and is likely to posi-
tively inuence safety behaviour in the workplace (Dyreborg, 2009;
Fernndez-Muniz et al., 2009).
Audits and workplace OHS inspections, designed to provide
appropriate and comprehensive information, are seen to be of value
with the proviso that appropriate and timely corrective action is
taken to address identied issues (Carson and Snowden, 2010;
Hallowell et al., 2013; Sinelnikov et al., 2015).
Consultation and communication about OHS is considered a pri-
ority, including regular, formal and informal communication and
consultation about OHS (Dejoy et al., 2004; Grabowski et al., 2007;
Health and Safety Executive, 2005).
Empowerment and employee involvement in decision mak-
ing about OHS encourages employees to take responsibility for
their behaviour and leads to positive safety behaviour outcomes
(Nahrgang et al., 2011; Wurzelbacher and Jin, 2011).
Management commitment and leadership is valuable and is
demonstrated in active engagement in areas such as OHS informa-
tion gathering, behaviour as OHS role models and support for OHS
as a high priority across the organisation (Choudhry et al., 2007;
Frazier et al., 2013; Health and Safety Executive, 2005; Lingard et al.,
2011; Zohar, 2010).
Positive feedback and recognition for OHS is considered to be
a leading indicator but not including rewards that might lead
to under-reporting of incidents or injuries (Daniels and Marlow,
2005).
Prioritisation of OHS, embedded in the organisation having pri-
macy alongside efciency and productivity, has emerged as an
important leading indicator (Glendon and Clarke, 2016; Health and
Safety Executive, 2005; Van Dyck et al., 2013; Zanko and Dawson,
2012).
Risk management of OHS, including risk assessment, control,
inspection and maintenance of psychosocial, physical and/or phys-
iological dimensions of OHS, has emerged as a valuable priority
(Fernndez-Muniz et al., 2009; Hopkins, 2009; Kjelln, 2009;
Pidgeon, 1991).
Systems for OHS are important and are typically implemented
and maintained by managers and in work groups. Such systems
include workplace policies, processes and practices designed to
control and monitor OHS (Frazier et al., 2013; Payne et al., 2009;
Pidgeon, 1991; Wachter and Yorio, 2014; Wurzelbacher and Jin,
2011).
The provision of OHS training, information, tools, and resources
that promote preparedness to act and provide relevant response
plans are key leading indicators of OHS (Health and Safety
Executive, 2005; Lingard et al., 2011).
Based on the above review, we propose that the construct of
leading indicators of OHS, that eliminate or control the precursors
to harm, offers organisations the opportunity to detect and mitigate
risks, or risk increases, before an OHS incident occurs or a haz-
ardous state is reached (Grabowski et al., 2007; Lingard et al., 2011;
Reiman and Pietikinen, 2012; Sinelnikov et al., 2015). The follow-
ing hypothesis is formed to reect the impact of leading indicators
of OHS on mitigating OHS incidents:
H1 Leading indicators of OHS will be negatively associated with
lagging indicators of OHS
2.2. Safety leadership as a moderator
Safety leadership has emerged within the OHS literature as a
key construct. Wong et al., 2016, for example, in their review of
the safety leadership literature concluded that workplace supervi-
sors have substantial inuence on the safety performance of their
132 C. Sheehan et al. / Accident Analysis and Prevention 92 (2016) 130138
employees. Much of the research has considered safety leadership
as an antecedent to lagging indicators such as injury rates (Barling
et al., 2002; Vredenburgh, 2002; Zohar, 2002) and employee expo-
sure rates to hazards (Gershon et al., 2000). Wong et al. (2016)
extend the discussion of the direct relationship between safety
leadership and outcomes to consider the moderators of trust and
safety culture. There has also been evidence in the literature of
safety leadership as a moderator of antecedent-OHS outcome asso-
ciations.
For example, Huang et al. (2004) established that the positive
effect of the quality of the execution of organisational safety policies
on safety outcomes was strengthened under conditions of strong
supervisor support for safety. More recently, Grifn and Hu (2013)
found that the degree to which a leader encouraged and promoted
safety-related learning moderated (i.e., intensied) the positive
association between safety monitoring and employee safety par-
ticipation. In a study examining young workers safety, Teed et al.,
2008 found that the association between transformational leader-
ship and safety outcomes was moderated by passive leadership (i.e.,
management by exception and laisse-faire leadership behaviours),
such that the positive association was weaker under conditions
when leaders demonstrated more passive leadership behaviours.
Similarly, Simard and Marchand (1997) provided evidence which
indicated that when supervisors encouraged safety participation,
employees were more likely to comply with safety rules.
With respect to the leading indicators research, Sinelnikov et al.
(2015) demonstrated that safety leadership potentially enables
leading indicators of OHS. Their analysis showed that when leading
indicators receive more attention in management communication
and when managers have greater expertise with respect to lead-
ing indicators, it is easier to take preventive action through leading
indicators. Theoretical support for the proposition that safety lead-
ership moderates the association between leading and lagging
indicators of OHS can be found in the social information process-
ing perspective (SIP; Salancik and Pfeffer, 1978). A key tenet of
the SIP perspective is that job attitudes are not formed exclusively
as a result of the objective aspects of the work itself, such as the
tasks or practices, but rather they also develop from how the work
is socially constructed. Thus, when confronted with new policies,
procedures or practices, the information gathering and processing
capacity of employees is typically circumscribed (Groth et al., 2002).
As a consequence, they must rely on social cues (i.e., information)
within their work environment in order to construct perceptions
and attitudes which, in turn, inuence their behaviour (Boekhorst,
2014).
Leaders are powerful actors in the social interaction processes
that develop within organisations and therefore they are usually
important sources of such cues through their explicit statements
and behaviours (Gonzalez-Roma and Peiro, 2014). They provide
followers with essential information about organisational practices
and policies as well as other work related issues (Gonzalez-Roma
et al., 2002). According to the SIP perspective there are three pri-
mary processes by which a leader can inuence a followers job
attitudes and work behaviour.
First, leaders can do so by guiding the subordinates job and
role perceptions and this inuence may be particularly acute for
middle-level managers (Chen et al., 2013). With regard to OHS, the
information provided to new incumbents about the OHS aspects
of their job by middle-level managers may be particularly credible,
given their higher status, relative to junior-level managers. By the
same token, while senior managers are typically the architects of
the organisations OHS policy, middle-level managers are likely to
enjoy much closer social proximity to their followers and there-
fore their cues about safe work practices are likely to be seen as
highly salient (Meyer, 1994). This is consistent with the ndings
of Wu et al. (2010), which indicated that the middle manage-
Fig. 1. Model of the proposed moderating effect of safety leadership on the associ-
ation between leading and lagging OHS indicators.
ment group has a primary responsibility for safety interaction or
directions, guidance and advice; safety informing or the reinforce-
ment and communication of the organisations safety policy; and
safety decision-making or the implementation of safety strategies
through planning, resource allocation, and safety improvement.
Indeed, Wu et al. (2010) found that, relative to their junior and
senior counterparts, the role played by middle-level managers was
the strongest predictor of safety climate.
Second, according to the SIP perspective, leaders can have the
effect of shifting members attention and shaping their interpreta-
tions (Grant et al., 2010). For example, leaders who are focused on
safety can, by virtue of the cues they provide, make the OHS aspects
of the job, which may have been ignored or minimised by employ-
ees, far more prominent and focal. Similarly, the social cues given by
leaders who place a strong emphasis on safety can provide employ-
ees with a frame for assessing the level of risk associated with their
role, potentially shaping their interpretation of their job as one in
which OHS incidents are not seen as an inherent and acceptable
risk.
Finally, proponents of the SIP perspective argue that leaders can
affect subordinates attitudes and behaviours by providing a role
modelling function (Chen et al., 2013). It is often the case that the
behaviour of unit leaders and line managers is easily observable by
their followers and it communicates to employees the behaviours
that are expected and supported by the organisation. Thus, when a
leader reliably and repeatedly models behaviours that are consis-
tent with the OHS policy and practices, employees observing this
behaviour are likely to learn vicariously what are deemed appropri-
ate actions and what is valued by the organisation. Indeed, there is
empirical evidence that supports this as Cree and Kelloway (1997)
found that the level of safe work practices modelled by leaders had
consequences for the safety behaviour of their followers.
In view of the impact that characteristics of middle managers
can have on the enactment of leading indicators, our second
hypothesis becomes:
H2 Safety leadership of middle managers will moderate the associ-
ation between leading indicators and lagging indicators of OHS, such
that the association between leading indicators and lagging indicators
of OHS will be increasingly negative at lower levels of safety leadership.
Fig. 1 presents a summary of our conceptual model.
3. Method
3.1. Sample
The sample for this study was recruited via a national multi-
industry survey conducted in Australia between September 2013
and November 2014. We targeted large employers and six organ-
isations were recruited into the study with OHS managers being
approached through professional networks. Sixty-six workplaces
(each a single worksite of an organisation) agreed to participate
in the survey and were representative of six industries identied
in the Australian and New Zealand Standard Industrial Classica-
 C. Sheehan et al. / Accident Analysis and Prevention 92 (2016) 130138
tion (ANZSIC, 2006): Arts and Recreation Services; Construction;
Electricity, Gas, Water and Waste Services; Health and Community
Services; Mining; and Transport, Postal and Warehousing.
The survey was administered to 10,362 employees and 3605
employees responded, resulting in a response rate of 35 percent.
Twenty-seven respondents were deleted as they did not identify as
belonging to a specic workplace and could not be included in the
multi-level analysis; this reduced the sample size to 3578. Work-
place size ranged from 4 to 532 employees with a mean size of 54
employees. Of the respondents, 24 percent were middle managers
or line managers and the remainder (76 percent) worked in other
roles. Most (77 percent) were employed on a continuing basis, 57
percent were working full-time, 61 percent were male and approx-
imately half (52 percent) had been working for their organisation
for ve years or less.
3.2. Measures
3.2.1. Leading indicators
Leading indicators were measured using the Organizational Per-
formance Metric Monash University (OPM-MU: Shea et al., 2016).
The OPM-MU is a revised version of the Organizational Performance
Metric developed at the Institute for Work & Health, Ontario Canada
(IWH-OPM: IWH, 2011,2013). The OPM-MU is an 8-item scale that
has been reported to be a reliable and valid measure of leading indi-
cators of OHS (Shea et al., 2016). Employees were asked to report
on their perceptions of the workplace they worked in most often,
rather than the organisation overall, using a 5-point scale (rang-
ing from 1 = strongly disagree to 5 = strongly agree), according to
the extent to which they agreed or disagreed with the eight state-
ments (e.g., Formal OHS audits at regular intervals are a normal part
of our workplace). The OPM-MU is shown in Appendix A. This mea-
sure has been validated in Shea et al.s (2016) study. The current
study uses the same data set and the Cronbach alpha coefcient
was .88.
3.2.2. Safety leadership
The safety leadership measure used in this study was the oper-
ations manager safety leadership scale developed by Wu et al.
(2010). This is a 12-item scale that is intended to measure aspects of
safety leadership. Sample items are: I objectively analyse the causes
of injuries, I visit the workplace to assess safety, and I encourage
employees to be safe in their working behaviour. Middle-level and line
managers in each workplace were asked to report on their percep-
tions of their own safety leadership, using a 5-point scale (ranging
from 1 = strongly disagree to 5 = strongly agree), with respect to the
12 statements. Safety leadership ratings were available for all but
one workplace. The average number of middle managers and line
managers providing leadership ratings was 13 per workplace.
Our analysis of the safety leadership measure, using exploratory
factor analysis, revealed a single-factor scale, with all items load-
ing signicantly on one factor, which explained 52 percent of the
common variance. The Cronbachs alpha coefcient for the safety
leadership measure was excellent ( = .91) and consistent with the
reliability coefcient reported by Wu et al. (2010; = .95).
3.2.3. Lagging indicators
Lagging indicators were operationalised as self-reported OHS
incidents. Respondents were asked to report the number of OHS
incidents they had been personally involved in at work over the
past 12 months. These self-report measures (sourced from Probst
et al., 2013) were: reported OHS incidents, unreported OHS inci-
dents (OHS incidents that were not reported to management) and
near misses (situations that could have caused an injury/illness,
but did not). Our inclusion of multiple measures of OHS incidents
is consistent with recent research advising investigation of a range
133
of lagging indicators (Christian et al., 2009; ONeill et al., 2013).
While there is some debate on whether near-misses are a leading
or lagging indicator (e.g. Hinze et al., 2013), following Probst et al.
(2013), for the purposes of this study we categorise near-misses as
a recordable incident.
3.2.4. Control variables
We included organisational tenure (years) and employment sta-
tus (coded 1 = continuing, 0 = other) as controls in our analyses
as research has indicated that less experienced workers (Breslin
and Smith, 2006) and lower status employees (Lundberg, 1999)
are more likely to experience an OHS incident at work. We also
included role overload as a control variable as it has been shown to
be a job demand (Wincent and rtqvist, 2011) related to employee
outcomes (Andrews and Kacmar, 2014). Role overload was mea-
sured using the 5-item Quantitative Workload Inventory (Spector
and Jex, 1998), which measures both the volume and pace of
employee workload. The role overload measure had excellent inter-
nal reliability ( = .89) in the current study. To statistically control
for unobserved heterogeneity on OHS outcomes at the organisa-
tional level, we included K-1 dummy variables in the regression
models, where K is the number of organisations (n = 6) in the sample
(Cohen et al., 2003).
3.3. Procedure
University research ethics approval was obtained before survey
distribution. Once ethics approval was conrmed, all employees
at each participating workplace had the opportunity to complete
the survey. The invitation to participate and distribution of the
survey depended on both workplace context and whether the
questionnaire was completed using paper-and-pencil or online.
Paper-and-pencil surveys were distributed at staff meetings or
through the internal mail. In workplaces where employees gath-
ered for staff meetings, provision was made for the researchers
to attend meetings to distribute information sheets and surveys.
Respondents completed the questionnaires at the staff meeting and
returned them in a sealed envelope directly to the researchers. In
other organisations, the survey was distributed through the inter-
nal mail where respondents completed the survey in their own time
and returned the questionnaires in a sealed envelope to a secure
central collection point within their workplace. Where the ques-
tionnaire was administered online, the researchers provided the
OHS manager with an email containing a link to the online sur-
vey; this email was sent by the OHS manager to employees. Privacy
legislation in Australia prevents employers from sharing employee
email addresses with researchers. Respondents completing the sur-
veys online or those who completed the pencil-and-paper surveys
at their own pace were sent two reminders, via a global email or a
newsletter sent by the OHS manager, two and four weeks after the
initial invitation. All respondents were assured of anonymity.
3.4. Statistical analysis
The data were collected from employees nested within work-
places. Multi-level modelling, using the SPSS Mixed Procedure
(Heck et al., 2013), was employed to test our hypotheses. Multi-
level (also called hierarchical linear) modelling is appropriate for
nested or hierarchical data, such as in the present study, for
which the assumption of independence of observations is violated
(Raudenbush and Bryk, 2002). In our two-level regression models,
OHS incidents were the level-1 dependent variables, leading indi-
cators the level-2 independent variable, and safety leadership the
level-2 moderator variable. As recommended by Tabachnick and
Fidell (2013), to generate correct standardised regression weights
in moderated regression, all variables were z-standardised prior to
134 C. Sheehan et al. / Accident Analysis and Prevention 92 (2016) 130138

Table 1
Means, standard deviations, and inter-correlations among the study variables.

Variable M SD 1 2 3 4 5 6 7

1. Continuing role 0.77 0.42

2. Organisational tenure 2.74 1.30 .11**
3. Role overload 3.54 1.09 .08** .02
4. Leading indicators (aggregated) 3.62 0.73 .08** 0.12** 0.09**
5. Safety leadership (aggregated) 4.09 0.58 .02 0.01 .03 .55**
6. Reported incidents 0.57 3.13 .06** .04* .07** 0.02 .11*
7. Unreported incidents 0.9 5.00 0.01 .01 .07** 0.13** 0.01 .58**
8. Near misses 1.09 7.18 .01 .02 .07** 0.11** 0.00 .33** .53**
* **
Note: p < 0.05, p < 0.01.

analysis. Standardisation is a form of grand mean centering com- 4.2. Multi-level modelling
monly performed on predictors in multi-level analysis (Mathieu
and Taylor, 2007). OHS outcome data were log transformed to Table 1 presents the means, standard deviations, and inter-
improve normality and linearity. correlations among the study variables.
Results for the regression analysis are shown in Table 2 below.
The rst hypothesis, that leading indicators would be nega-
tively associated with lagging indicators of OHS, was tested with
4. Results
respect to three lagging indicator measures: reported OHS inci-
dents, unreported OHS incidents, and near misses. With regard
4.1. Aggregation
to the reported OHS incidents, after controlling for type of role,
organisational tenure, and role overload, leading indicators were
In line with the majority of leadership studies, we dened safety
not signicantly related to this lagging indicator (see Table 2, Model
leadership as a group-level construct (Taggar and Ellis, 2007). Lead-
1). In terms of unreported OHS incidents, leading indicators of OHS
ing indicators has been conceived in the literature as both an
was a statistically signicant predictor, with higher levels of lead-
individual- and as a group-level variable. Indeed, as discussed ear-
ing indicators being associated with fewer unreported incidents
lier in this paper, Hopkins (2009) referred to personal and process
(see Table 2, model 3). Similarly, leading indicators were negatively
OHS leading indicators and, in the climate literature, safety climate
related to near misses (see Table 2, model 5). Therefore, hypothesis
has been treated as an individual-level construct (psychological
1 was supported for unreported OHS incidents and near misses.
safety climate) and as a group-level phenomenon (Christian et al.,
The second hypothesis, that safety leadership will moderate the
2009). Here we construe leading indicators as a group-level con-
association between leading indicators and lagging indicators of
struct. We accept that individual workers will have perceptions
OHS, such that the association will be increasingly negative at lower
regarding worksite OHS policies, practices and procedures, how-
levels of safety leadership, was also tested with respect to three
ever, we argue that when employees at a worksite share these
lagging indicator measures: reported OHS incidents, unreported
perceptions, a group-level phenomenon emerges. We calculated
OHS incidents, and near misses. No statistically signicant interac-
the within-group correlation (rwg ) and the intra-class correlation
tion was observed between leading indicators and safety leadership
ICC(2) values to ascertain whether ratings of these constructs were
for reported OHS incidents (see Table 2, model 2). The interaction
adequate for aggregation to the workplace level. The mean rwg
between leading indicators and safety leadership was statistically
values (calculated for a multi-item scale assuming a uniform null
signicant for unreported OHS incidents (see Table 2, model 4). To
distribution) were .97 for leading indicators and. 97 for safety lead-
aid in the interpretation of this interaction effect, we plotted the
ership, indicating a high level of within-group agreement (LeBreton
simple slopes of leading indicators and OHS outcomes using the
and Senter, 2008). The ICC(2) values, which measure the reliabil-
conventional low (1 SD below the mean) and high (1 SD above the
ity of the group means, were .90 for leading indicators and .58
mean) values of the moderator, safety leadership (Aiken and West,
for safety leadership. The ICC(2) value for safety leadership was
1991).
slightly below the commonly used lower bound of .60 (cf. Glick,
As shown in Fig. 2 the association between leading indicators
1985), which was possibly related to the relatively small number
and unreported OHS incidents was strongly negative when safety
of middle managers and line managers in each workplace. Taken
leadership was high (one SD above the mean; = 0.92, p < 0.01)
together, these results provide support for aggregation of OHS lead-
rather than low (one SD below the mean; = .08, p > 0.05). A sta-
ing indicators and safety leadership to the workplace level (Bliese,
tistically signicant interaction was also observed between leading
2000).

Table 2
Results of multilevel regression analyses.

Reported incidents Unreported incidents Near misses

Model 1 Model 2 Model 3 Model 4 Model 5 Model 6

Within level
Continuing role .07** .07** 0.02 0.02 .04* 0.01
Organisational tenure .09** .09** .05** .05** .11** .05*
Role overload .09** .09** .12** .12** 0.13** .12**
Between level
Leading indicators .06 .05 0.32** 0.42** 0.32** 0.36**
Safety leadership .05 0.25** .10
Leading indicators*Safety leadership 0.10 0.50** 0.40**
R2 (OLS) .08 .08 .03 .04 .05 .06

Note: * p < .05, ** p < .01. Standardised coefcients reported. Fixed effects for organisations were included, but not reported for ease of presentation.
 C. Sheehan et al. / Accident Analysis and Prevention 92 (2016) 130138
1.5
1 The results indicate that leading indicators are negatively asso-
Unreported OHS incidents
0.5
Low (-1 SD) SL
0
High (+1 SD) SL
-0.5
-1
Low (- 1 SD) LI High (+1 SD) LI
-1.5
Fig. 2. Interaction between leading indicators and safety leadership for unreported
OHS incidents.
1 Muniz et al., 2009; Grabowski et al., 2007). Such activity could
0.8
0.6
0.4
Near misses
0.2
Low (-1 SD) SL
0 organisations may actually increase the reporting of OHS incidents.
-0.2 High (+1 SD) SL
-0.4
-0.6
-0.8
-1
Low (-1 SD) L High (+1 SD) LI
Fig. 3. Interaction between leading indicators and safety leadership for near misses.
indicators and safety leadership for near misses (see Table 2, model
6). As shown inFig. 3, the association between leading indicators
and near misses was more negative when safety leadership was
high one SD above the mean; ( = 0.76, p < 0.01) rather than low
(one SD below the mean; = .04, p > 0.05). Taken together, these
results support the interaction effects specied in hypothesis 2 for
unreported OHS incidents and near misses.
5. Discussion
Lagging indicators of OHS, such as near misses as well as
reported and unreported incidents, provide important feedback
to organisations about deciencies in the management of OHS
(Reiman and Pietikinen, 2012). Although such data are valuable,
there has been a call for research that provides an additional focus
on OHS leading indicators as these eliminate or control the pre-
cursors to harm, as opposed to lagging indicators, which measure
events or outcomes that have already occurred (Grabowski et al.,
2007; Hopkins, 2009; Lingard et al., 2011; Reiman and Pietikainen,
2012; Shea et al., 2016). It has been further suggested that lead-
ing indicators research can be enhanced by examining the factors
associated with, or that enhance the impact of, leading indicators.
Suggestions include consideration of the impact of factors such
as leadership commitment, engagement, understanding, and sup-
port (Sinelnikov et al., 2015). Our focus on leadership as a factor
that moderates leading-lagging indicator associations is consistent
with the priority given to leadership elsewhere as an important
factor in terms of inuencing safety outcomes (Martnez-Crcoles
et al., 2011; Wong et al., 2016) and with the theoretical arguments
presented in the SIP perspective (Salancik and Pfeffer, 1978). In
response to these developments within the literature, the aim of
the current research has been to consider the direct association
between leading and lagging indicators of OHS and the moderating
135
effect of safety leadership of middle managers on these associa-
tions.
ciated with the lagging indicators of unreported incidents and
near misses, but are not predictive of reported OHS incidents. The
negative associations between OHS leading indicators and both
unreported incidents and near misses supports the impact of the
former as eliminating or controlling precursors to harm (Grabowski
et al., 2007; Lingard et al., 2011; Reiman and Pietikinen, 2012;
Sinelnikov et al., 2015). The results do not provide evidence, how-
ever, of an association between leading indicators and reported
incidents as an outcome measure. The lack of a signicant asso-
ciation may be a result of the increase in communication ows
that characterise leading indicator activity. Specic leading indi-
cator areas, for example, include increased OHS consultation and
communication (Dejoy et al., 2004; Dyreborg, 2009; Fernndez-
actually encourage high incident reporting as a form of feedback
and therefore nullify the expected negative association between
leading indicator activity and the lagging indicator of reported OHS
incidents. Indeed, some studies (Blegen et al., 2004) have found that
strong safety management systems and strong safety cultures in
With respect to the expected moderation effect of safety lead-
ership, our ndings indicate that the associations between leading
indicators and both unreported incidents and near misses become
increasingly negative at higher levels of safety leadership. In other
words, a leader who prioritises safety creates an environment
where leading indicators are more likely to reduce near misses
and unreported incidents. Again, there is no signicant association
with reported incidents but, consistent with the argument made
previously, it is possible that stronger safety leadership, combined
with activity to prevent OHS outcomes, could actually nullify any
expected reduction in the reported of incidents. That is, the combi-
nation of stronger safety leadership and OHS outcome prevention
could actually increase willingness to report any incidents.
Our nding of a moderating effect for safety leadership on lead-
ing and lagging associations supports Sinelnikov et al.s (2015)
suggestion of leadership as an intensifying factor for leading indi-
cators. In Sinelnikov et al.s research, when respondents were asked
to provide a list of the most important factors that enabled them
to implement leading indicators of OHS, leadership was nominated
most frequently. The moderating effect for safety leadership iden-
tied in the current research is also consistent with arguments
from the SIP perspective and the social construction of work, which
state that employees rely on social cues to develop attitudes and
behaviours (Salancik and Pfeffer, 1978). Middle managers, consis-
tent with their selection in the current research, provide a powerful
and credible source of information for employees. They are a group
of leaders who have more credibility, relative to more junior super-
visory staff, and greater involvement in policy design (Chen et al.,
2013; Meyer, 1994). They also focus attention on key aspects of the
work role and, if these managers value safety, this, in turn, will be
prioritised by employees (Grant et al., 2010). Finally, as role mod-
els who adhere to safe work procedures, these managers allow
employees to learn vicariously appropriate behaviours (Cree and
Kelloway, 1997).
5.1. Research contribution and implications
First, the ndings of our research address Sinelnikov et al.s
(2015) call for ongoing research that provides an empirical link
between leading and lagging OHS indicators. The established link
strengthens information available to managers about what can be
done to prevent OHS incidents from occurring.
136 C. Sheehan et al. / Accident Analysis and Prevention 92 (2016) 130138
Second, our study provides OHS practitioners with evidence of
the impact of leading indicator activity. As noted by Sinelnikov et al.
(2015), some top-level executives still exhibit a tendency to solely
embrace lagging indicators as they are more familiar with tradi-
tional OHS performance data. These authors have suggested that
some of the responsibility for re-education rests with OHS practi-
tioners; however, to be able to make a case to senior management,
these practitioners need access to metrics that clearly show the
value of addressing and measuring leading indicators of OHS. The
current research assists by providing such evidence of the impact
of leading indicator activity on lagging indicators that can be com-
municated to senior management.
Third, our research highlights the important role played by
middle management. As noted above, senior management com-
mitment to OHS leading indicator activity is critical, but there is
a need for sustained safety leadership training programs for man-
agers at all levels in order for them to understand their OHS roles
and to learn to be safety leaders. Elsewhere Wu et al. (2010) have
similarly noted the importance of safety leadership throughout the
organisation and the allocation of time and resources necessary to
develop safety leadership.
5.2. Limitations and future directions
First, although the research surveyed 3578 employees from
rms drawn from six different industries, there is scope for enhanc-
ing the generalisability of the ndings by examining the association
between leading and lagging indicators across a wider range of
industries. Second, given the variations in risk exposure across
industries, there is scope in future research to explore system-
atic differences in the impact of leadership and industry risk levels
on the association between leading and lagging indicators. Wong
et al. (2016), for example, noted the possible boundary conditions of
industry risk levels and explained that most of the studies on lead-
ership and safety outcomes were conducted in blue-collar safety
critical populations. Third, to reduce the risk of common method
variance (CMV) manager ratings of safety leadership were aggre-
gated to the workplace level. A meta-analysis by Christian and
colleagues (Christian et al., 2009) showed that self-reported safety
outcome data are relatively immune from response biases. Nev-
ertheless, we acknowledge that OHS reporting practices are likely
to vary between industry sectors and within workplaces in each
sector. Investigation of such differences may be an area of future
research. Fourth, our study was limited to medium-to-large organ-
isations; future research could investigate whether the effects are
evident in smaller businesses. Fifth, the cross-sectional nature of
the study precludes making inferences about causality. The nd-
ings of this study would need to be replicated with a longitudinal
design before stronger inferences could be drawn in relation to the
associations between leading and lagging indicators of OHS. For
example, it may take some time for leading indicators to affect lag-
ging indicators and these temporal dynamics are best captured with
longitudinal data.
Finally, our research focused on the moderating impact of safety
leadership of middle managers on leading and lagging indicator
associations and there is scope to investigate variation in impact
across all leadership levels. Future work could also examine the
implementation of leading indicators and their ongoing measure-
ment in workplaces. Specically, research could investigate the
importance of planning, education, connection with existing pro-
cesses and practices, and implementation of a regular evaluation
process (Hallowell et al., 2013; Sinelnikov et al., 2015).
6. Conclusions
The overall aim of the current research was to examine the
association between leading and lagging indicators of OHS and
the moderating effect of safety leadership on this association. The
research makes a number of contributions. The research provides
evidence of an association between leading indicators and lagging
indicators of OHS and, as such, contributes to leading indicators
research by addressing Sinelnikov et al.s (2015) call for empirical
evidence of a connection between leading and lagging indicators.
Evidence of this association also provides OHS practitioners with
useful information to substantiate efforts within organisations to
move away from a primary focus on lagging indicators, towards a
preventative focus on leading indicator activity. Additionally the
conrmation of the moderating impact of safety leadership of mid-
dle managers in the direct association, specically at the middle
management level, highlights the important role played by this
management group and the value of OHS leadership development
and investment at the middle management level.
We acknowledge the support provided for this research project
by the WorkSafe Victoria and the Institute for Safety Compensation
and Recovery Research, Australia. We acknowledge Dr Benjamin
Amick III, Institute for Work & Health, Ontario Canada, for his com-
ments on the development of this research.
Appendix A.
OPM- Monash University#
Please read each statement carefully and select the number that
best shows your views about health and safety at this workplace.
 C. Sheehan et al. / Accident Analysis and Prevention 92 (2016) 130138 137

For the purpose of this survey an audit means a formal process of evaluating and reporting on how the workplace manages health and safety in accordance with a recognised standard. Regular means that an audit is repeated at regular
References

Strongly Agree
Australian and New Zealand Standard Industrial Classication (ANZSIC), 2006.
Revision 1.0) serial number 1292.0 (accessed 18 February 2016) http://www.
abs.gov.au/ausstats//Previousproducts/
AF04F89CEE4E54D6CA25711F00146D76?opendocument.
Aiken, L.S., West, S.G., 1991. Multiple Regression: Testing and Interpreting

5
5
5
5
5
5
5
5
Interactions. Sage Publications, Newbury Park, CA.
Andrews, M.C., Kacmar, K.M., 2014. Easing employee strain: the interactive effects
Agree

of empowerment and justice on the role overload-strain relationship. J. Behav.

Appl. Manage. 15 (2), 4358.
4
4
4
4
4
4
4
4
Baker, J., Bowman, F., Erwin, G., Gorton, S., Hendershot, D., Leveson, N., Priest, S.,
Rosenthal, I., Tebo, P., Weigmann, D., Wilson, L.D., 2007. The Report of the BP
Neither agree nor disagree

US Reneries Independent Safety Review Panel, Retrieved April 2012 from:

www.bp.com/bakerpanelreport.
Barling, J., Loughlin, C., Kelloway, E.K., 2002. Development and test of a model
linking safety-specic transformational leadership and occupational safety. J.
Appl. Psychol. 87, 488496.
Battaglia, M., Passetti, E., Frey, M., 2015. Occupational health and safety
management in municipal waste companies: a note on the Italian sector. Saf.
Sci. 72, 5565.
Blair, E., OToole, M., 2010. Leading measures: enhancing safety climate & driving
safety performance. Prof. Saf. J. 8 (1), 2934.
3
3
3
3
3
3
3
3

Blegen, M.A., Vaughn, T., Pepper, G., Vijir, C., Stratton, K., Boyd, M., Armstrong, G.,
2004. Patient and staff safety: voluntary reporting. Am. J. Med. Qual. 19, 6774.
Bliese, P.D., 2000. Within-grou agreement, non-indeendence and reliability:
Disagree

Imlications for data aggregation and analysis. In: Klein, K.J., Kozlowski, S.W.J.
(Eds.), Multilevel Theory, Research, and Organizational Theory: Foundations,
2
2
2
2
2
2
2
2

Extensions, and New Directions. Jossey-Bass, San Francisco, pp. 349381.

Boekhorst, J.A., 2014. The role of authentic leadership in fostering workplace
inclusion: a social information processing perspective. Hum. Resour. Manage.
Strongly Disagree

54, 241264.
Breslin, F.C., Smith, P., 2006. Trial by re: a multivariate examination of the relation
between job tenure and work injuries. Occup. Environ. Med. 63, 2732.
Carson, P.A., Snowden, D., 2010. Health, safety and environment metrics in loss
preventionpart 1. Loss Prev. Bull. 212, 1115.
Chen, Z., Takeuchi, R., Shum, C., 2013. A social information processing perspective
1
1
1
1
1
1
1
1

of co-worker inuence on a focal employee. Org. Sci. 24, 16181639.

Choudhry, R.M., Fang, D., Mohamed, S., 2007. The nature of safety culture: a survey
of the state-of-the-art. Saf. Sci. 45, 9931012.
This workplace considers health and safety at least as important as production and quality in the way work is done

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Christian, M.S., Bradley, J.C., Wallace, J.C., Burke, M.J., 2009. Workplace safety: a
meta-analysis of the roles of person and situation factors. J. Appl. Psychol. 94,
11031127.
Cohen, J., Cohen, P., West, S.G., Aiken, L.S., 2003. Applied Regression/Correlation
analysis for the behavioral sciences. Mahwah, NJ: Lawrence Erlbaum
Associates.
Those in charge of OHS have the authority to make the changes they have identied as necessary

Cree, T., Kelloway, K., 1997. Responses to occupational hazards: exit and
participation. J. Occup. Health Psychol. 2, 304311.
Daniels, C., Marlow, P., 2005. Literature review on the reporting of workplace
injury trends. Retrieved June 2012, from www.hse.gov.uk/research/hsl pdf/
Everyone has the resources and/or equipment they need to complete their work safely

2005/hsl0536.pdf.
Dejoy, D.M., Schaffer, B.S., Wilson, M.G., Vandenberg, R.J., Butts, M.M., 2004.
Creating safer workplaces: assessing the determinants and role of safety
Everyone at this workplace values ongoing OHS improvement in this workplace

climate. J. Safety Res. 35, 8190.

Employees are always involved in decisions affecting their health and safety

Dyreborg, J., 2009. The causal relation between lead and lag indicators. Saf. Sci. 47,
Formal OHS audits at regular intervals are a normal part of our workplace*

474475.
Workers and supervisors have the information they need to work safely

Fernndez-Muniz, B., Montes-Pen, J.M., Vzquez-Ords, C.J., 2009. Relation

between occupational safety management and rm performance. Saf. Sci. 47,
980991.
Frazier, C.B., Ludwig, T.D., Whitaker, B., Roberts, D.S., 2013. A hierarchical factor
analysis of a safety culture survey. J. Safety Res. 45, 1528.
Gershon, R.R., Karkashian, C.D., Grosch, J.W., Murphy, L.R., Escamilla-Cejudo, A.,
Flanagan, P.A., Bernacki, E., Kasting, C., Martin, L., 2000. Hospital safety climate
and its relationship with safe work practices and workplace exposure
incidents. Am. J. Infect. Control 28, 211221.
Those who act safely receive positive recognition

Glendon, A.I., Clarke, S.G., 2016. Human Safety and Risk Management: A
Psychological Perspective, 3rd edn. CRC Press/Taylor & Francis, Boca Raton, FL.
Glick, W.H., 1985. Conceptualizing and measuring organizational and psychological
climate: pitfalls in multilevel research. Acad. Manage. Rev. 10, 601616.
Goldenhar, L., Williams, L.J., Swanson, N., 2003. Modelling relationships between
job stressors and injury and near-miss outcomes for construction labourers.
Work Stress 17 (3), 218240.
intervals, for example, once every year.

Gonzalez-Roma, V., Peiro, J.M., 2014. Climate and culture strength. In: Schneider,
B., Barbera, K.M. (Eds.), The Oxford Handbook of Organizational Climate and
Culture. New York, Oxford University Press, pp. 496531.
Gonzalez-Roma, V., Peiro, J.M., Tordera, N., 2002. An examination of the
antecedents and moderator inuences of climate strength. J. Appl. Psychol. 87,
465473.
Grabowski, M., Ayyalasomayajula, P., Merrick, J., Harrald, J.R., Roberts, K., 2007.
Leading indicators of safety in virtual organizations. Saf. Sci. 45, 10131043.
Grant, A.M., Fried, Y., Juillerat, T., 2010. Work matters: job design in classic and
contemporary perspectives. In: Zedeck, S. (Ed.), The APA Handbook of
Industrial and Organizational Psychology, Vol 1. American Psychological
Association, Washington DC, pp. 417453.
1.
2.
3.
4.
5.
6.
7.
8.

#

138 C. Sheehan et al. / Accident Analysis and Prevention 92 (2016) 130138
Grifn, M.A., Hu, X., 2013. How leaders differentially motivate safety compliance
and safety participation: the role of monitoring, inspiring, and learning. Saf.
Sci. 60, 196202.
Groth, M., Goldman, B.M., Gilliland, S.W., Bies, R.L., 2002. Commitment to legal
claiming: inuences of attributions, social guidance, and organizational tenure.
J. Appl. Psychol. 87, 781788.
Hallowell, M., Hinze, J., Baud, K., Wehle, A., 2013. Proactive construction safety
control: measuring, monitoring, and responding to safety leading indicators. J.
Constr. Eng. Manage. 139 (10).
Health and Safety Executive, (2005). A review of safety culture and safety climate
literature for the development of the safety culture inspection toolkit. HSE
Books: Bristol.
Heck, R.K., Thomas, S.L., Tabata, L.N., 2013. Multilevel and Longitudinal Modeling
with IBM SPSS, 2nd ed. Routledge, New York.
Hinze, J., Thurman, S., Wehle, A., 2013. Leading indicators of construction safety
performance. Saf. Sci. 51, 2328.
Hopkins, A., 2009. Thinking about process safety indicators. Saf. Sci. 47, 460465.
Huang, Y.-H., Chen, P.Y., Krauss, A.D., Rogers, D.A., 2004. Quality of the execution of
corporate safety policies and employee safety outcomes: assessing the
moderating role of supervisor safety support and the mediating role of
employee safety control. J. Bus. Psychol. 18, 483506.
IWH, 2011. Benchmarking Organizational Leading Indicators for the Prevention
and Management of Injuries and Illnesses: Final Report. Institute for Work &
Health, Toronto, Ontario.
IWH, 2013. Developing Leading Indicators of Work Injury and Illness. Institute for
Work & Health, Toronto, Ontario.
Kjelln, U., 2009. The safety measurement problem revisited. Saf. Sci. 47, 486489.
Laitinen, H., Vuorinen, M., Simola, A., Yrjnheikki, E., 2013. Observation-based
proactive OHS outcome indicatorsvalidity of the Elmeri+ method. Saf. Sci. 54,
6979.
LeBeau, M., Duguay, P., Boucher, A., 2014. Costs of occupational injuries and
diseases in Quebec. J. Safety Res. 50, 8998.
LeBreton, J.M., Senter, J.L., 2008. Answers to 20 questions about interrater
reliability and interrater agreement. Org. Res. Methods 11, 815852.
Li, F., Jiang, L., Yao, X., Li, Y., 2013. Job demands, job resources and safety outcomes:
the roles of emotional exhaustion and safety compliance. Accid. Anal. Prev. 51,
243251.
Lingard, H., Wakeeld, R., Cashin, P., 2011. The development and testing of a
hierarchical measure of project OHS performance. Eng. Constr. Arch. Manage.
18, 3049.
Lundberg, U., 1999. Stress responses in low-status jobs and their relationship to
health risks: musculoskeletal disorders. Ann. N. Y. Acad. Sci. 896, 162172.
Martnez-Crcoles, M., Gracia, F.J., Toms, I., Peir, J.M., Schbel, M., 2011.
Leadership and employees perceived safety behaviours in a nuclear power
plant: a structural equation model. Saf. Sci. 49, 11181129.
Mathieu, J.E., Taylor, S.E., 2007. A framework for testing meso-mediational
relationships in organizational behavior. J. Org. Behav. 28 (2), 141172.
Meyer, G.W., 1994. Social information processing and social networks: a test of
social inuence mechanisms. Hum. Relat. 47, 10131047.
Nahrgang, J.D., Morgeson, F.P., Hofmann, D.A., 2011. Safety at work: a
meta-analytic investigation of the link between job demands, job resources,
burnout, engagement, and safety outcomes. J. Appl. Psychol. 96 (1), 7194.
ONeill, S., Martinov-Bennie, N., Cheung, A., 2013. Issues in the Measurement and
Reporting of Work Health and Safety Performance: A Review. Macquarie
Lighthouse Press, Sydney, NSW, Australia.
Payne, S.C., Bergman, M.E., Beus, J.M., Rodrguez, J.M., Henning, J.B., 2009. Safety
climate: leading or lagging indicator of safety outcomes? J. Loss Prev. Process
Ind. 22, 735739.
Pidgeon, N.F., 1991. Safety culture and risk management in organizations. J.
Cross-Cult. Psychol. 22, 129140.
Probst, T.M., Graso, M., Estrada, A.X., Greer, S., 2013. Consideration of future safety
consequences: a new predictor of employee safety. Accid. Anal. Prev. 55,
123134.
Raudenbush, S.W., Bryk, A.S., 2002. Hierarchical Linear Models, 2nd ed. Sage
Publications, Thousand Oaks.
Reiman, T., Pietikinen, E., 2012. Leading indicators of system safetymonitoring
and driving the organizational safety potential. Safety Sci. 50, 19932000.
Salancik, G.R., Pfeffer, J., 1978. A social information processing approach to job
attitudes and task design. Adm. Sci. Q. 23, 224253.
Shea, T., De Cieri, H., Donohue, R., Cooper, B., Sheehan, C., 2016. Leading indicators
of occupational health and safety: an employee and workplace level validation
study. Safety Sci. 85, 293304.
Simard, M., Marchand, A., 1997. Workgroups propensity to comply with safety
rules: the inuence of micro-macro organisational factors. Ergonomics 40,
172188.
Sinelnikov, S., Inouye, J., Kerper, S., 2015. Using leading indicators to measure
occupational health and safety performance. Safety Sci. 72, 240248.
Spector, Paul E., Jex, Steve M., 1998. Development of four self-report measures of
job stressors and strain: Interpersonal Conict at Work Scale, Organizational
Constraints Scale, Quantitative Workload Inventory, and Physical Symptoms
Inventory. J. Occup. Health Psychol. 3 (4), 356367.
Tabachnick, B.G., Fidell, L.S., 2013. Using Multivariate Statistics, 6th ed. Pearson,
Sydney, NSW.
Taggar, S., Ellis, R., 2007. The role of leaders in shaping formal team norms.
Leadership Q. 18, 105120.
Teed, M., Kelloway, E.K., Mullen, J.E., 2008. Young workers safety. The Impact of
Inconsistent Leadership Paper Presented at the Work, Stress and Health
Conference, Washington, D.C.
Van Dyck, C., Dimitrova, N.G., de Korne, D.F., Hiddema, F., 2013. Walk the talk:
leaders enacted priority of safety, incident reporting, and error management.
Adv. Health Care Manage. 14, 95117.
Vredenburgh, A., 2002. Organizational safety: which management practices are
most effective in reducing employee injury rates? J. Safety Res. 33, 259276.
Wachter, J.K., Yorio, P.L., 2014. A system of safety management practices and
worker engagement for reducing and preventing accidents: an empirical and
theoretical investigation. Accid. Anal. Prev. 68, 117130.
Wincent, J., rtqvist, D., 2011. Examining positive performance implications of role
stressors by the indirect inuence of positive affect: a study of new business
managers. J. Appl. Soc. Psychol. 41 (3), 699727.
Wong, J.H.K., Kelloway, E.K., Makhan, D.W., 2016. Safety leadership. In: Clarke, S.,
Probst, T.M., Guldenmund, F., Passmore, J. (Eds.), The Wiley-Blackwell
Handbook of the Psychology of Occupational Safety and Workplace Health.
Chichester, UK, Wiley-Blackwell, pp. 83110.
Wu, T.-C., Lin, C.-H., Shiau, S.-Y., 2010. Predicting safety culture: the roles of
employer, operations manager and safety professional. J. Safety Res. 41,
423431.
Wurzelbacher, S., Jin, Y., 2011. A framework for evaluating OSH program
effectiveness using leading and trailing metrics. J. Safety Res. 42, 199207.
Zanko, M., Dawson, P., 2012. Occupational health and safety management in
organizations: a review. Int. J. Manage. Rev. 14 (3), 328344.
Zohar, D., 2002. The effects of leadership dimensions, safety climate, and assigned
priorities on minor injuries in work groups. J. Org. Behav. 23, 7592.
Zohar, D., 2010. Thirty years of safety climate research: reections and future
directions. Accid. Anal. Prev. 42, 15171522.