1

Safety Climate and the Lone Worker: Review of Research

Marvin J. Dainoff, PhD, CPE Marvin.dainoff@libertymutual.com

15th Annual PRP Symposium University of Cincinnati

2

•  Center for Injury Epidemiology •  Center for Physical Ergonomics •  Center for Behavioral Sciences •  Center for Disability Research

Liberty Mutual Research Institute for Safety generating knowledge to help people live safer and more secure lives

Conduc'ng  non-­‐proprietary  research  –  findings  are  published  in  the  peer-­‐reviewed  scien'fic  literature  

3

Safety Climate Team

Center Director: Marvin Dainoff, Ph.D. Principal Investigators: Emily Huang, Ph.D., Liberty Mutual Research Institute for Safety Dov Zohar, Ph.D., Israel Institute of Technology Co-investigator: Michelle Robertson, Ph.D., CPE

Project Coordinator: Susan Jeffries Harvard/LMRIS Postdoctoral Fellow: Garry Gray, Ph.D., Lauren Murphy, Ph.D., Jin Lee, Ph.D. Research Assistants: Angela Garabet, Jenn Rineer, Peg Rothwell, Anna McFadden, Janelle Cheung Liberty Mutual Consultants: •  Dave Melton - Industry Director, Transportation •  Dave Money - Technical Director, Transportation

4

LMRIS Project Team

5

Standard Interventions Employed to Reduce Risk

Four broad categories: –  Engineering (e.g., redesigning a tool or installing

machine guards) –  Administrative (e.g., changing job procedures or

rotating workers through a particular job) –  Personal Protective Equipment (e.g., protective

glasses or hearing protection) –  Education and Training

Until recently, most intervention efforts have relied on traditional approaches.

6

Beyond the Traditional Approaches

Risk Managers and Safety Directors are now exploring organizational and psychosocial factors in the workplace to complement other approaches in an attempt to make further improvements.

 Safety  Climate  inves/ga/ons  are  a  major  part  of  this  effort.  

7

Definition – Safety Culture

Shared beliefs and values that interact with an organization’s structures and control systems to produce behavior norms (the way of doing things in the organization)

(Reason, 1998; Thompson et al, 1996; Utall, 1983)

8

Organizational Culture vs. Climate

Culture is regarded as an emergent property of group interactions, whereas

Climate refers more to people’s perception of/reactions to those interactions.

Safety Climate is a measurable aspect of Safety Culture

9

Definition – Safety Climate (Zohar, 1980)

Employees’ shared perceptions of their company’s safety policies, procedures, and practices

Focus on: –  Relative priorities

(e.g. safety vs. productivity or efficiency) –  Alignment between espousals and enactments

(e.g. leader’s words vs. actions) –  Internal consistency among policies, procedures

and practices (e.g. discrepancy between supervisor vs. top management)

10

Safety Climate Research Grows…Why?

Huang, Y.H., Chen, P.Y. & Grosch, J.W. (2010). Safety climate: New developments in the conceptualization and industry-focused studies. Editorial Note for Special Issue on Safety Climate. Accident Analysis and Prevention.

11

Meta-Analysis by Christian et al. (2009) Combined Results of 202 Studies

Leading indicator

Lagging indicator

Safety Climate is a leading indicator of future injury

12

Purpose of this Research Program

•  Extend safety climate to the lone/remote worker situation

• Develop industry specific safety climate scales for –  Long haul truckers –  Electric utility workers

•  Develop lone work generic safety climate scale

13

Methodological Overview

–  Literature review

–  Comments from subject matter experts

–  In-depth interviews with employees and managers

–  Field observations

–  Generate initial items

–  Cognitive interviews

–  Pilot tests

Scale Development

Initial Administration

Administration to Remaining

Sample

Outcome Measures

14

Methodological Overview

–  Confirmatory Factor Analysis (full sample)

–  Exploratory Factor Analysis

–  Coefficient Alpha Reliability

Scale Development

Initial Administration

Administration to Remaining

Sample

Outcome Measures

15

Methodological Overview

–  Subjective Behavior Ratings (concurrent)

–  Subjective Injury Data (concurrent)

–  Objective Injury Data (predictive – trucking only)

Scale Development

Initial Administration

Administration to Remaining

Sample

Outcome Measures

16

Multilevel Focus of SC Questions

Company (Top Management) Level

Group (Immediate Supervisor) Level

Employees

Supervisors

17

Lone Worker Safety Climate Research Program

•  Long  Haul  Truckers  – Company Level Safety Climate Scale – Group Level Safety Climate Scale

•  Electric  U'lity  Workers  – Company Level Safety Climate Scale – Group Level Safety Climate Scale

• Generic  Lone  Workers    – Company Level Safety Climate Scale – Group Level Safety Climate Scale

[Generic scales modified for lone workers from Zohar and Luria (2005) and included in survey administration]

18

SC in the Trucking Industry

Company A B C D E F G H

# Respondents 558 248 2,030 461 290 4,003 235 270

Response Rate 55% 73% 34% 37% 58% 51% 40% N/A

•  8  Large  trucking  firms  in  the  US  •  9,095  respondents  (8095  employees,  1,000  supervisors)  •  Final  SC  survey  includes  40  items  (20  Group-­‐level,  20  Company-­‐level  Safety  Climate)  

Data  collected:  •  Subjec've:  SC  and  self-­‐reported  behaviors    •  Objec've:  Accident/injury  data  (6  months  post  survey)  

19

Trucking Safety Climate Scale

• Company-level safety climate (upper management) 20 items Three dimensions (Proactive Practices, Driver Safety Priority, and Supervisory Care Promotion) “My company allows drivers to change their schedules when they are getting too tired.”

• Group-level safety climate (immediate supervisors) 20 items Three dimensions (Safety Promotion, Delivery Limits, and Cell Phone Disapproval) “My dispatcher stops talking to me on the phone if he/she hears I’m driving.”

20

Injury Rate versus Safety Climate Score for Participating Carriers

R² = 0.5905

21

Outcome Measures-Trucking

• Pooled safety climate scores

Safety Climate (company-level and

group-level)

Safety Behavior

Lost Days

Hard Braking

-

+

-

22

Scientific Publication

Transportation Research Part F: Traffic Psychology and Behavior (Huang, Zohar, et al., 2013) Development and Validation of Safety Climate Scales for Lone Workers using Truck Drivers as Exemplar

23

SC in the Utility Industry

Company A B

# Respondents 1,560 869

Response Rate 46% 74%

•  Two  large  electric  u'lity  firms  •  2,421  respondents  •  48  item  survey  (19  Group-­‐level,  29  Company-­‐level)  

Data  collected:  •  Subjec've:  SC,  self-­‐reported  behaviors,  self-­‐reported  accident/injuries  

24

Utility/Electric Safety Climate Scale

• Company-level safety climate: 29 items

Six dimensions (Safety Pro-activity, General Training, Trucks and Equipment, Field Orientation, Financial Investment, and Schedule flexibility) “My company invests in the best safety equipment available (e.g., power tools, meter pullers, lighting)”

• Group-level safety climate: 19 items

Three dimensions (Supervisory Care, Participation Encouragement, and Safety Straight Talk)

“My supervisor takes the time to check on me, especially when I’m stressed or tired.”

25

Outcome Measures-Utility

• Pooled safety climate scores

Safety Climate (company-level and

group-level)

Safety Behavior

Lost Days

Recordable Incidents

-

+

-

26

Scientific Publication

A manuscript was published in the journal Accident Analysis and Prevention (Huang, Zohar, et al., 2013) Development and Validation of Safety Climate Scales for Mobile Remote Workers using Utility/Electrical Workers as Exemplar

27

Generic Items for Lone Workers adopted from Zohar and Luria (2005)

• Company-level safety climate: 6 items

• Group-level safety climate: 6 items

Company-level safety climate

Top management at this company – 1. Reacts quickly to solve the problem when told about safety conce

rns 2. Is strict about working safely when delivery falls behind schedule

3. Uses any available information to improve existing safety rules

4. Invests a lot in safety training for workers

5. Listens carefully to our ideas about improving safety

6. Tries to continually improve safety levels in each department Group-level safety climate

My direct supervisor – 1. Discusses with us how to improve safety

2. Compliments employees who pay special attention to safety

3. Is strict about working safely even when we are tired or stressed

4. Frequently talks about safety issues throughout the work week

5. Refuses to ignore safety rules when work falls behind schedule

6. Uses explanations (not just compliance) to get us to act safely

28

Benefit of Industry-Specific Scale

0

5

10

15

20

25

Generic Trucking

Variance (%)

Scale Type

Company Group

29

Scientific Publication

A manuscript was published in the journal Accident Analysis and Prevention (Lee, Huang, et al., 2014). External Validity of a Generic Safety Climate Scale for Lone Workers across Different Industries and Companies. Accident Analysis and Prevention, 63, 138-145.

30

Key Findings

•  Generic scales and industry-specific scales are reliable and valid instruments for measuring SC in lone workers (trucking and utility workers).

•  Both generic and industry-specific scales predicted driving safety behavior (self-reported) and road injury outcomes (accident data).

•  The industry-specific safety climate scale demonstrated stronger predictive value than the generic scale (data only available for trucking).

•  Both industry-specific scales as well as the generic scales show measurement equivalence in that scale items and measurement constructs have consistent meaning across different companies.

31

Additional Findings

32

Leaders Create Culture

Leader-Member Exchange (LMX) and Work Ownership both promote safety climate

Testing the role of contextual attributes of lone work known from the management science literature

33

Employee vs. Supervisors’ Perceptions (Trucking Data Illustrated: parallel finding for Utilities)

•  For both company- and group-level safety climate, employee and supervisor perceptions of safety climate were significantly different – supervisors consistently reported higher levels of safety

climate

•  Only employee perceptions of safety climate significantly predicted safety behavior (directly) and injury outcomes (indirectly) – supervisor perceptions had no predictive value

Emp_SC

Sup_SC

driving safety days out

.33**

.03 -1.24**

.03

34

Additional Published Scientific Papers

•  Zohar, D., Huang, Y.H., Robertson, M.M. & Lee, J. (2014). A Mediation Model Linking Supervisory Leadership and Work Ownership with Safety Climate as Predictors of Truck Driver Safety Performance. Accident Analysis and Prevention, 62, 17-25.

•  Huang, Y., Zohar, D., Robertson, M. M., Lee, J., Rineer, J., Murphy, L., Garabet, A. & Dainoff, M.J. (2014). Supervisor vs. Employee Safety Climate Perceptions: Association with Safety Behavior and Outcomes for Lone Workers. Transportation Research Part: Traffic Psychology and Behavior.

35

Interaction Between Group and Company Safety Climate

•  Employees’ CSC and GSC perceptions are highly correlated (r = .78), but notable gaps between the two

•  Both scores were equally predictive of safety behavior

•  The two interacted with one another to predict safety behavior such that if either the CSC or GSC scores were low, the other’s contribution to safety behavior was complimentary.

•  These findings suggested that CSC and GSC function in a compensatory manner and promote safe driving behavior even when either CSC or GSC scores are low

Safety Behavior

Company SC

36

Safety Climate Affects Work Quality

• Company-level and Group-level safety climate perceptions directly influence job satisfaction, employee engagement, and objective turnover rate (one year after survey).

Safety Climate (company-level and

group-level)

Job satisfaction

Employee engagement

Objective turn-over rate

+

+

-

37

Conference Presentations/Proceedings

•  Huang, Y.H., Lee, J., Rineer, J., McFadden, A., & Robertson, M.M. (APA). The Interaction between Organization- and Group-level Safety Climate Perceptions: Association with Safety Behavior for Lone Workers.

•  Huang, Y.H., Zohar, D, et al. (ODAM), The Impact of Safety Climate beyond Safety Outcomes: Job Satisfaction, Employee Engagement and Objective Turnover Rate.

38

Conclusion: Towards Evidence-Based Interventions

•  Safety climate is a valid thermometer across diverse settings and applications.

•  Current work being extended by a second wave of data collection -> longitudinal study

•  While predictive of outcomes, and able to discriminate good from bad organizations, SC is not diagnostic

•  Intervention requires systematic evaluation to identify system weak points: currently ongoing

39

The Future

•  Exploring the intersection between safety climate, systems safety and management science

•  Developing interventions based on SC screening followed by comprehensive analysis guided by sociotechnical systems theory

40

Generating knowledge to help people live safer, more secure lives

www.libertymutualgroup.com/researchinstitute www.libertymutualgroup.com/researchinstitute

41

Safety Climate as Shared Perceptions

•  Original Conception: basis for shared perceptions --- workers within the same work group interact with each other and supervisor

•  Statistical homogeneity criteria for determining if perceptions are shared – Utility: shared perceptions – Trucking: not shared perceptions

•  Implications: – Psychological Safety Climate is operative for trucking, but

is still predictive of safety outcomes – Work group level Safety Climate predictive of safety

outcomes for utilities. Potential focus of interventions

42

Multilevel Focus of SC Questions

Company (Top Management) Level

Group (Immediate Supervisor) Level

Supervisors

EE

EE E E

E

Employees

43

Shared Safety Climate Perception Homogeneity tests

• Utility/electric workers’ safety climate perceptions were shared within the same supervisor group!

Criteria: rwgj > .70 (James et al., 1993) ICC(1) > .10 (Murphy & Myors, 1998) ICC(2) > .70 (LeBreton & Senter, 2008)

Pilot company (1560

participants, 115 work-units, average 12.4 employees)

2nd company (861 participants, 68 work-units,

average 9.1 employees)

rwgj ICC(1) ICC(2) rwgj ICC(1) ICC(2) Total CSC  

.95 .21 .73 .96 .25 .75

  Total GSC  

.93 .19 .70 .93 .27 .77

44

Safety Climate and Employee Engagement as Predictors of Safety Behavior and Road Injury

Engagement

TSC Driving Safety Road Injury -.04** .29** -.38**

TSC × Egmt

.20**

-.09** -.11**

.15**

Safety climate and employee engagement predicted safety behavior, which mediated their effect on subsequently measured road injury outcomes. There is a moderation effect of engagement on the strength of relationship between climate perceptions and safety behavior.