Professional Documents
Culture Documents
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 and mitigate risks or risk increases before an OHS incident occurs
or a hazardous state is reached (Sinelnikov et al., 2015).
Occupational health and safety (OHS) measurement relies heav- Sinelnikov et al. (2015), in their study of the state of knowledge
ily on lagging indicators, such as incidents of workplace injury, and practice on the use of leading indicators of OHS, explained that
as these measures provide important feedback information about although there is increasing interest in leading indicators, there is
deciencies and safety incidents that have occurred (Reiman and a need for further evidence of the link between leading and lag-
Pietikinen, 2012). Lagging OHS indicators are, however, a reac- ging indicators. These authors also noted the potential enabling
tive measurement approach to safety management and measure impact of leadership in terms of implementing leading indica-
events or outcomes that have already happened (Hopkins, 2009; tors. As recognised in the social information processing perspective
Laitinen et al., 2013; Reiman and Pietikinen, 2012). As such, (SIP), leadership behaviour is an important determinant of the
lagging indicators are failure-focused (Sinelnikov et al., 2015). development of employee job attitudes and behaviours (Chen et al.,
Recent research emphasises a more proactive evaluation of OHS 2013). The aim of the current research is to address the research
activity that emphasises leading indicators, or inputs, that allow gaps noted above by considering: 1) the association between lead-
organisations to predict safety concerns and that may reduce the ing and lagging indicators of OHS, and 2) the moderating effect of
likelihood of an OHS incident occurring (Grabowski et al., 2007; safety leadership on the association between leading and lagging
Lingard et al., 2011; Reiman and Pietikinen, 2012). Leading indi- indicators.
cators can be thought of as eliminating or controlling the precursors The structure of the paper is as follows. First, the literature on
to harm and as such offer organisations the opportunity to detect 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
Corresponding author. are described. The third section of the paper presents the results
E-mail addresses: cathy.sheehan@monash.edu (C. Sheehan), of the study, while the fourth section provides a discussion of the
ross.donohue@monash.edu (R. Donohue), tracey.shea@monash.edu (T. Shea), studys results and contribution. The nal section of the paper pro-
brian.cooper@monash.edu (B. Cooper), helen.decieri@monash.edu (H.D. Cieri).
1
vides an overview of the studys limitations and avenues for future
We acknowledge the support provided for this research project by the WorkSafe
Victoria and the Institute for Safety Compensation and Recovery Research, Australia. research.
We acknowledge Dr Benjamin Amick III, Institute for Work & Health, Ontario Canada,
for his comments on the development of this 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 131
tion (ANZSIC, 2006): Arts and Recreation Services; Construction; of lagging indicators (Christian et al., 2009; ONeill et al., 2013).
Electricity, Gas, Water and Waste Services; Health and Community While there is some debate on whether near-misses are a leading
Services; Mining; and Transport, Postal and Warehousing. or lagging indicator (e.g. Hinze et al., 2013), following Probst et al.
The survey was administered to 10,362 employees and 3605 (2013), for the purposes of this study we categorise near-misses as
employees responded, resulting in a response rate of 35 percent. a recordable incident.
Twenty-seven respondents were deleted as they did not identify as
belonging to a specic workplace and could not be included in the 3.2.4. Control variables
multi-level analysis; this reduced the sample size to 3578. Work- We included organisational tenure (years) and employment sta-
place size ranged from 4 to 532 employees with a mean size of 54 tus (coded 1 = continuing, 0 = other) as controls in our analyses
employees. Of the respondents, 24 percent were middle managers as research has indicated that less experienced workers (Breslin
or line managers and the remainder (76 percent) worked in other and Smith, 2006) and lower status employees (Lundberg, 1999)
roles. Most (77 percent) were employed on a continuing basis, 57 are more likely to experience an OHS incident at work. We also
percent were working full-time, 61 percent were male and approx- included role overload as a control variable as it has been shown to
imately half (52 percent) had been working for their organisation be a job demand (Wincent and rtqvist, 2011) related to employee
for ve years or less. outcomes (Andrews and Kacmar, 2014). Role overload was mea-
sured using the 5-item Quantitative Workload Inventory (Spector
3.2. Measures and Jex, 1998), which measures both the volume and pace of
employee workload. The role overload measure had excellent inter-
3.2.1. Leading indicators nal reliability ( = .89) in the current study. To statistically control
Leading indicators were measured using the Organizational Per- for unobserved heterogeneity on OHS outcomes at the organisa-
formance Metric Monash University (OPM-MU: Shea et al., 2016). tional level, we included K-1 dummy variables in the regression
The OPM-MU is a revised version of the Organizational Performance models, where K is the number of organisations (n = 6) in the sample
Metric developed at the Institute for Work & Health, Ontario Canada (Cohen et al., 2003).
(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- 3.3. Procedure
cators of OHS (Shea et al., 2016). Employees were asked to report
on their perceptions of the workplace they worked in most often, University research ethics approval was obtained before survey
rather than the organisation overall, using a 5-point scale (rang- distribution. Once ethics approval was conrmed, all employees
ing from 1 = strongly disagree to 5 = strongly agree), according to at each participating workplace had the opportunity to complete
the extent to which they agreed or disagreed with the eight state- the survey. The invitation to participate and distribution of the
ments (e.g., Formal OHS audits at regular intervals are a normal part survey depended on both workplace context and whether the
of our workplace). The OPM-MU is shown in Appendix A. This mea- questionnaire was completed using paper-and-pencil or online.
sure has been validated in Shea et al.s (2016) study. The current Paper-and-pencil surveys were distributed at staff meetings or
study uses the same data set and the Cronbach alpha coefcient through the internal mail. In workplaces where employees gath-
was .88. ered for staff meetings, provision was made for the researchers
to attend meetings to distribute information sheets and surveys.
3.2.2. Safety leadership Respondents completed the questionnaires at the staff meeting and
The safety leadership measure used in this study was the oper- returned them in a sealed envelope directly to the researchers. In
ations manager safety leadership scale developed by Wu et al. other organisations, the survey was distributed through the inter-
(2010). This is a 12-item scale that is intended to measure aspects of nal mail where respondents completed the survey in their own time
safety leadership. Sample items are: I objectively analyse the causes and returned the questionnaires in a sealed envelope to a secure
of injuries, I visit the workplace to assess safety, and I encourage central collection point within their workplace. Where the ques-
employees to be safe in their working behaviour. Middle-level and line tionnaire was administered online, the researchers provided the
managers in each workplace were asked to report on their percep- OHS manager with an email containing a link to the online sur-
tions of their own safety leadership, using a 5-point scale (ranging vey; this email was sent by the OHS manager to employees. Privacy
from 1 = strongly disagree to 5 = strongly agree), with respect to the legislation in Australia prevents employers from sharing employee
12 statements. Safety leadership ratings were available for all but email addresses with researchers. Respondents completing the sur-
one workplace. The average number of middle managers and line veys online or those who completed the pencil-and-paper surveys
managers providing leadership ratings was 13 per workplace. at their own pace were sent two reminders, via a global email or a
Our analysis of the safety leadership measure, using exploratory newsletter sent by the OHS manager, two and four weeks after the
factor analysis, revealed a single-factor scale, with all items load- initial invitation. All respondents were assured of anonymity.
ing signicantly on one factor, which explained 52 percent of the
common variance. The Cronbachs alpha coefcient for the safety 3.4. Statistical analysis
leadership measure was excellent ( = .91) and consistent with the
reliability coefcient reported by Wu et al. (2010; = .95). The data were collected from employees nested within work-
places. Multi-level modelling, using the SPSS Mixed Procedure
3.2.3. Lagging indicators (Heck et al., 2013), was employed to test our hypotheses. Multi-
Lagging indicators were operationalised as self-reported OHS level (also called hierarchical linear) modelling is appropriate for
incidents. Respondents were asked to report the number of OHS nested or hierarchical data, such as in the present study, for
incidents they had been personally involved in at work over the which the assumption of independence of observations is violated
past 12 months. These self-report measures (sourced from Probst (Raudenbush and Bryk, 2002). In our two-level regression models,
et al., 2013) were: reported OHS incidents, unreported OHS inci- OHS incidents were the level-1 dependent variables, leading indi-
dents (OHS incidents that were not reported to management) and cators the level-2 independent variable, and safety leadership the
near misses (situations that could have caused an injury/illness, level-2 moderator variable. As recommended by Tabachnick and
but did not). Our inclusion of multiple measures of OHS incidents Fidell (2013), to generate correct standardised regression weights
is consistent with recent research advising investigation of a range 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
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.
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 135
incidents. Indeed, some studies (Blegen et al., 2004) have found that
0.2
Low (-1 SD) SL strong safety management systems and strong safety cultures in
0 organisations may actually increase the reporting of OHS incidents.
-0.2 High (+1 SD) SL With respect to the expected moderation effect of safety lead-
-0.4 ership, our ndings indicate that the associations between leading
-0.6 indicators and both unreported incidents and near misses become
-0.8 increasingly negative at higher levels of safety leadership. In other
-1 words, a leader who prioritises safety creates an environment
Low (-1 SD) L High (+1 SD) LI where leading indicators are more likely to reduce near misses
and unreported incidents. Again, there is no signicant association
Fig. 3. Interaction between leading indicators and safety leadership for near misses.
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
indicators and safety leadership for near misses (see Table 2, model
expected reduction in the reported of incidents. That is, the combi-
6). As shown inFig. 3, the association between leading indicators
nation of stronger safety leadership and OHS outcome prevention
and near misses was more negative when safety leadership was
could actually increase willingness to report any incidents.
high one SD above the mean; ( = 0.76, p < 0.01) rather than low
Our nding of a moderating effect for safety leadership on lead-
(one SD below the mean; = .04, p > 0.05). Taken together, these
ing and lagging associations supports Sinelnikov et al.s (2015)
results support the interaction effects specied in hypothesis 2 for
suggestion of leadership as an intensifying factor for leading indi-
unreported OHS incidents and near misses.
cators. In Sinelnikov et al.s research, when respondents were asked
to provide a list of the most important factors that enabled them
5. Discussion to implement leading indicators of OHS, leadership was nominated
most frequently. The moderating effect for safety leadership iden-
Lagging indicators of OHS, such as near misses as well as tied in the current research is also consistent with arguments
reported and unreported incidents, provide important feedback from the SIP perspective and the social construction of work, which
to organisations about deciencies in the management of OHS state that employees rely on social cues to develop attitudes and
(Reiman and Pietikinen, 2012). Although such data are valuable, behaviours (Salancik and Pfeffer, 1978). Middle managers, consis-
there has been a call for research that provides an additional focus tent with their selection in the current research, provide a powerful
on OHS leading indicators as these eliminate or control the pre- and credible source of information for employees. They are a group
cursors to harm, as opposed to lagging indicators, which measure of leaders who have more credibility, relative to more junior super-
events or outcomes that have already occurred (Grabowski et al., visory staff, and greater involvement in policy design (Chen et al.,
2007; Hopkins, 2009; Lingard et al., 2011; Reiman and Pietikainen, 2013; Meyer, 1994). They also focus attention on key aspects of the
2012; Shea et al., 2016). It has been further suggested that lead- work role and, if these managers value safety, this, in turn, will be
ing indicators research can be enhanced by examining the factors prioritised by employees (Grant et al., 2010). Finally, as role mod-
associated with, or that enhance the impact of, leading indicators. els who adhere to safe work procedures, these managers allow
Suggestions include consideration of the impact of factors such employees to learn vicariously appropriate behaviours (Cree and
as leadership commitment, engagement, understanding, and sup- Kelloway, 1997).
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 5.1. Research contribution and implications
factor in terms of inuencing safety outcomes (Martnez-Crcoles
et al., 2011; Wong et al., 2016) and with the theoretical arguments First, the ndings of our research address Sinelnikov et al.s
presented in the SIP perspective (Salancik and Pfeffer, 1978). In (2015) call for ongoing research that provides an empirical link
response to these developments within the literature, the aim of between leading and lagging OHS indicators. The established link
the current research has been to consider the direct association strengthens information available to managers about what can be
between leading and lagging indicators of OHS and the moderating 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 ings of this study would need to be replicated with a longitudinal
the impact of leading indicator activity. As noted by Sinelnikov et al. design before stronger inferences could be drawn in relation to the
(2015), some top-level executives still exhibit a tendency to solely associations between leading and lagging indicators of OHS. For
embrace lagging indicators as they are more familiar with tradi- example, it may take some time for leading indicators to affect lag-
tional OHS performance data. These authors have suggested that ging indicators and these temporal dynamics are best captured with
some of the responsibility for re-education rests with OHS practi- longitudinal data.
tioners; however, to be able to make a case to senior management, Finally, our research focused on the moderating impact of safety
these practitioners need access to metrics that clearly show the leadership of middle managers on leading and lagging indicator
value of addressing and measuring leading indicators of OHS. The associations and there is scope to investigate variation in impact
current research assists by providing such evidence of the impact across all leadership levels. Future work could also examine the
of leading indicator activity on lagging indicators that can be com- implementation of leading indicators and their ongoing measure-
municated to senior management. ment in workplaces. Specically, research could investigate the
Third, our research highlights the important role played by importance of planning, education, connection with existing pro-
middle management. As noted above, senior management com- cesses and practices, and implementation of a regular evaluation
mitment to OHS leading indicator activity is critical, but there is process (Hallowell et al., 2013; Sinelnikov et al., 2015).
a need for sustained safety leadership training programs for man-
agers at all levels in order for them to understand their OHS roles 6. Conclusions
and to learn to be safety leaders. Elsewhere Wu et al. (2010) have
The overall aim of the current research was to examine the
similarly noted the importance of safety leadership throughout the
association between leading and lagging indicators of OHS and
organisation and the allocation of time and resources necessary to
the moderating effect of safety leadership on this association. The
develop safety leadership.
research makes a number of contributions. The research provides
5.2. Limitations and future directions evidence of an association between leading indicators and lagging
indicators of OHS and, as such, contributes to leading indicators
First, although the research surveyed 3578 employees from research by addressing Sinelnikov et al.s (2015) call for empirical
rms drawn from six different industries, there is scope for enhanc- evidence of a connection between leading and lagging indicators.
ing the generalisability of the ndings by examining the association Evidence of this association also provides OHS practitioners with
between leading and lagging indicators across a wider range of useful information to substantiate efforts within organisations to
industries. Second, given the variations in risk exposure across move away from a primary focus on lagging indicators, towards a
industries, there is scope in future research to explore system- preventative focus on leading indicator activity. Additionally the
atic differences in the impact of leadership and industry risk levels conrmation of the moderating impact of safety leadership of mid-
on the association between leading and lagging indicators. Wong dle managers in the direct association, specically at the middle
et al. (2016), for example, noted the possible boundary conditions of management level, highlights the important role played by this
industry risk levels and explained that most of the studies on lead- management group and the value of OHS leadership development
ership and safety outcomes were conducted in blue-collar safety and investment at the middle management level.
critical populations. Third, to reduce the risk of common method We acknowledge the support provided for this research project
variance (CMV) manager ratings of safety leadership were aggre- by the WorkSafe Victoria and the Institute for Safety Compensation
gated to the workplace level. A meta-analysis by Christian and and Recovery Research, Australia. We acknowledge Dr Benjamin
colleagues (Christian et al., 2009) showed that self-reported safety Amick III, Institute for Work & Health, Ontario Canada, for his com-
outcome data are relatively immune from response biases. Nev- ments on the development of this research.
ertheless, we acknowledge that OHS reporting practices are likely
to vary between industry sectors and within workplaces in each Appendix A.
sector. Investigation of such differences may be an area of future
research. Fourth, our study was limited to medium-to-large organ- OPM- Monash University#
isations; future research could investigate whether the effects are Please read each statement carefully and select the number that
evident in smaller businesses. Fifth, the cross-sectional nature of best shows your views about health and safety at this workplace.
the study precludes making inferences about causality. The nd-
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
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
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
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
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
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 Pidgeon, N.F., 1991. Safety culture and risk management in organizations. J.
and safety participation: the role of monitoring, inspiring, and learning. Saf. Cross-Cult. Psychol. 22, 129140.
Sci. 60, 196202. Probst, T.M., Graso, M., Estrada, A.X., Greer, S., 2013. Consideration of future safety
Groth, M., Goldman, B.M., Gilliland, S.W., Bies, R.L., 2002. Commitment to legal consequences: a new predictor of employee safety. Accid. Anal. Prev. 55,
claiming: inuences of attributions, social guidance, and organizational tenure. 123134.
J. Appl. Psychol. 87, 781788. Raudenbush, S.W., Bryk, A.S., 2002. Hierarchical Linear Models, 2nd ed. Sage
Hallowell, M., Hinze, J., Baud, K., Wehle, A., 2013. Proactive construction safety Publications, Thousand Oaks.
control: measuring, monitoring, and responding to safety leading indicators. J. Reiman, T., Pietikinen, E., 2012. Leading indicators of system safetymonitoring
Constr. Eng. Manage. 139 (10). and driving the organizational safety potential. Safety Sci. 50, 19932000.
Health and Safety Executive, (2005). A review of safety culture and safety climate Salancik, G.R., Pfeffer, J., 1978. A social information processing approach to job
literature for the development of the safety culture inspection toolkit. HSE attitudes and task design. Adm. Sci. Q. 23, 224253.
Books: Bristol. Shea, T., De Cieri, H., Donohue, R., Cooper, B., Sheehan, C., 2016. Leading indicators
Heck, R.K., Thomas, S.L., Tabata, L.N., 2013. Multilevel and Longitudinal Modeling of occupational health and safety: an employee and workplace level validation
with IBM SPSS, 2nd ed. Routledge, New York. study. Safety Sci. 85, 293304.
Hinze, J., Thurman, S., Wehle, A., 2013. Leading indicators of construction safety Simard, M., Marchand, A., 1997. Workgroups propensity to comply with safety
performance. Saf. Sci. 51, 2328. rules: the inuence of micro-macro organisational factors. Ergonomics 40,
Hopkins, A., 2009. Thinking about process safety indicators. Saf. Sci. 47, 460465. 172188.
Huang, Y.-H., Chen, P.Y., Krauss, A.D., Rogers, D.A., 2004. Quality of the execution of Sinelnikov, S., Inouye, J., Kerper, S., 2015. Using leading indicators to measure
corporate safety policies and employee safety outcomes: assessing the occupational health and safety performance. Safety Sci. 72, 240248.
moderating role of supervisor safety support and the mediating role of Spector, Paul E., Jex, Steve M., 1998. Development of four self-report measures of
employee safety control. J. Bus. Psychol. 18, 483506. job stressors and strain: Interpersonal Conict at Work Scale, Organizational
IWH, 2011. Benchmarking Organizational Leading Indicators for the Prevention Constraints Scale, Quantitative Workload Inventory, and Physical Symptoms
and Management of Injuries and Illnesses: Final Report. Institute for Work & Inventory. J. Occup. Health Psychol. 3 (4), 356367.
Health, Toronto, Ontario. Tabachnick, B.G., Fidell, L.S., 2013. Using Multivariate Statistics, 6th ed. Pearson,
IWH, 2013. Developing Leading Indicators of Work Injury and Illness. Institute for Sydney, NSW.
Work & Health, Toronto, Ontario. Taggar, S., Ellis, R., 2007. The role of leaders in shaping formal team norms.
Kjelln, U., 2009. The safety measurement problem revisited. Saf. Sci. 47, 486489. Leadership Q. 18, 105120.
Laitinen, H., Vuorinen, M., Simola, A., Yrjnheikki, E., 2013. Observation-based Teed, M., Kelloway, E.K., Mullen, J.E., 2008. Young workers safety. The Impact of
proactive OHS outcome indicatorsvalidity of the Elmeri+ method. Saf. Sci. 54, Inconsistent Leadership Paper Presented at the Work, Stress and Health
6979. Conference, Washington, D.C.
LeBeau, M., Duguay, P., Boucher, A., 2014. Costs of occupational injuries and Van Dyck, C., Dimitrova, N.G., de Korne, D.F., Hiddema, F., 2013. Walk the talk:
diseases in Quebec. J. Safety Res. 50, 8998. leaders enacted priority of safety, incident reporting, and error management.
LeBreton, J.M., Senter, J.L., 2008. Answers to 20 questions about interrater Adv. Health Care Manage. 14, 95117.
reliability and interrater agreement. Org. Res. Methods 11, 815852. Vredenburgh, A., 2002. Organizational safety: which management practices are
Li, F., Jiang, L., Yao, X., Li, Y., 2013. Job demands, job resources and safety outcomes: most effective in reducing employee injury rates? J. Safety Res. 33, 259276.
the roles of emotional exhaustion and safety compliance. Accid. Anal. Prev. 51, Wachter, J.K., Yorio, P.L., 2014. A system of safety management practices and
243251. worker engagement for reducing and preventing accidents: an empirical and
Lingard, H., Wakeeld, R., Cashin, P., 2011. The development and testing of a theoretical investigation. Accid. Anal. Prev. 68, 117130.
hierarchical measure of project OHS performance. Eng. Constr. Arch. Manage. Wincent, J., rtqvist, D., 2011. Examining positive performance implications of role
18, 3049. stressors by the indirect inuence of positive affect: a study of new business
Lundberg, U., 1999. Stress responses in low-status jobs and their relationship to managers. J. Appl. Soc. Psychol. 41 (3), 699727.
health risks: musculoskeletal disorders. Ann. N. Y. Acad. Sci. 896, 162172. Wong, J.H.K., Kelloway, E.K., Makhan, D.W., 2016. Safety leadership. In: Clarke, S.,
Martnez-Crcoles, M., Gracia, F.J., Toms, I., Peir, J.M., Schbel, M., 2011. Probst, T.M., Guldenmund, F., Passmore, J. (Eds.), The Wiley-Blackwell
Leadership and employees perceived safety behaviours in a nuclear power Handbook of the Psychology of Occupational Safety and Workplace Health.
plant: a structural equation model. Saf. Sci. 49, 11181129. Chichester, UK, Wiley-Blackwell, pp. 83110.
Mathieu, J.E., Taylor, S.E., 2007. A framework for testing meso-mediational Wu, T.-C., Lin, C.-H., Shiau, S.-Y., 2010. Predicting safety culture: the roles of
relationships in organizational behavior. J. Org. Behav. 28 (2), 141172. employer, operations manager and safety professional. J. Safety Res. 41,
Meyer, G.W., 1994. Social information processing and social networks: a test of 423431.
social inuence mechanisms. Hum. Relat. 47, 10131047. Wurzelbacher, S., Jin, Y., 2011. A framework for evaluating OSH program
Nahrgang, J.D., Morgeson, F.P., Hofmann, D.A., 2011. Safety at work: a effectiveness using leading and trailing metrics. J. Safety Res. 42, 199207.
meta-analytic investigation of the link between job demands, job resources, Zanko, M., Dawson, P., 2012. Occupational health and safety management in
burnout, engagement, and safety outcomes. J. Appl. Psychol. 96 (1), 7194. organizations: a review. Int. J. Manage. Rev. 14 (3), 328344.
ONeill, S., Martinov-Bennie, N., Cheung, A., 2013. Issues in the Measurement and Zohar, D., 2002. The effects of leadership dimensions, safety climate, and assigned
Reporting of Work Health and Safety Performance: A Review. Macquarie priorities on minor injuries in work groups. J. Org. Behav. 23, 7592.
Lighthouse Press, Sydney, NSW, Australia. Zohar, D., 2010. Thirty years of safety climate research: reections and future
Payne, S.C., Bergman, M.E., Beus, J.M., Rodrguez, J.M., Henning, J.B., 2009. Safety directions. Accid. Anal. Prev. 42, 15171522.
climate: leading or lagging indicator of safety outcomes? J. Loss Prev. Process
Ind. 22, 735739.