Safety culture, Safety behavior and safety performance in ... · Safety culture, Safety behavior and safety performance in Railway companies Abstract Forming a positive safety culture

Challenge H: For an even safer and more secure railway

1

Safety culture, Safety behavior and safety performance in Railway companies

Abstract

Forming a positive safety culture is considered as an essential indicator to prevent thepossible railway accidents. Furthermore, improved safety behavior can reduce the frequencyof work related accident and injuries and safety behavior is often tied to quality ofperformance. Therefore, this research aims at identifying the correlations among safetyculture, safety behavior, and safety performance, as well as explaining the cause-effectfactors affecting the above three dimensions. Most of the existing relevant references wereconducted by literature reviews, few of them have been demonstrated by empirical evidencein railway companies. The factor analysis and structure equation modeling are adopted toexamine the relationships among the safety culture, safety behavior and safety performancefrom Taiwan Railway System drivers’ perspective. A total of 536 effective samples wereobtained.

Previous studies were conducted to investigate the major causes of the rail relevantaccidents and incidents by the analysis of post-accident data. This study contributes to findempirical evidence to investigate the relationship among safety culture, safety behavior, andsafety performance from railway organizational perspectives. Our empirical result indicatesthe positive safety culture has certain effect upon the awareness in safety behavior of railwaydriver. In addition, the positive safety culture was also found to have a positive effect on thepositive perceived safety performance. The safety behavior was found to have a positiveeffect on the perceived safety performance. Several policies suggestions were included andcould be useful for the decision makers of rail system’s operators and the governmentregulators when designing and regulating the safety management system.

Keywords: Safety culture; Safety behavior; Safety performance; Structural equation modeling

1. Introduction

Rail system could often provide fast, reliable and safe transportation service. However, theconsequences of a rail accident are severe due to the kinetic energy caused by the operatingspeed and weight of train as well as the passengers on board (Elms,2001).Human errorcould be attributed to the majority of incidents and accidents of railway system (Baysari et al2008; Krokos and Baker,2007; O’Hare,2000). To reduce the human error in the rail systemoperation seems to be one of the essential issues to be explored.


2

Human factors play a significant role in the safety of rail operations and safety culture wasconsidered as the most difficult topics (Elms, 2001). Turner et al. (1989) defined safetyculture as “the set of beliefs, norms, attitudes, roles, and social and technical practices thatare concerned with minimizing the exposure of employees, managers, customers andmembers of the public to conditions considered dangerous or injurious”. Glennon (1982)found that organizations with poor safety culture scores had higher accident rates than thoseorganizations with better safety culture scores. Safety culture was considered as a leadingindicator of railway system safety (Harper, 1997; Sawacha, 1999; Hudson,1999;Clarke ,2000; Cullen,2001; O’Toole ,2002).

Campbell et al.,(1993) there are only three determinants of individual difference inperformance: knowledge, skill, and motivation mediate the relationship between theantecedents and components of performance. Wong (1999) indicated the development ofsafety systems, safety practice and procedures; monitoring of safety compliance,establishment of safety committees at site level, communication of safety policies to sitepersonnel, participation of safety officers, consultation between site staff and safety officersalso affect the safety performance. Erickson (2000) noted the improvements in organizationalstructure, organizational importance of safety, safety responsibility and accountability,communication, management behavior, employee involvement, and employee responsesand behavior can help improve safety performance.

Previous studies were conducted to investigate the major causes of the rail relevantaccidents and incidents by the analysis of post-accident data (Baysari et al, 2008; Gilroy andGrimes, 2005; Murphy, 2001; Shanahan et al 2005). This study contributes to demonstrateempirical evidence to investigate the relationship among safety culture, safety behavior, andsafety performance from railway organizational perspectives.

Therefore, the rest of this study is organized as follows. Section 2 discusses the relationshipamong the safety culture, safety behavior and perceived safety performance. Section 3presents research hypotheses, questionnaire design, sampling processes and modelingapproach. Section 4 describes the empirical result of the modeling approach. The discussionand conclusions are included in the final section of this study.

2. Methodology

2.1 Research hypotheses

This research aims at identifying the correlations among safety culture, safety behavior, and


3

safety performance, as well as explaining the cause-effect factors affecting the above threedimensions. McDonald et al. (2000) explored the relationships of different aspects of safetyculture and safety management systems and presented a revised model of safetymanagement systems. Improvements in organizational structure, organizational importanceof safety, safety responsibility and accountability, communication, management behavior,employee involvement, and employee responses and behavior can help improve safetyperformance (Erickson, 2000). Edkins and Pollock (1997) indicated useful strategies forimproving driver vigilance should be directed at improving the safety culture of the operatingenvironment.

The lack of motivation in fostering a safety culture at both organizational and project levelshas resulted in a poor safety record in general, with construction being one of the mosthazardous industries globally (Harper,1997; Sawacha ,1999). Cooper (2000) presented areciprocal model of safety culture to understand its dynamic, multi-faceted and holistic nature.The promotion of a positive safety culture is now considered to be a viable way of managingrisk, creating a culture within an organization where everyone is personally involved inensuring safety (Hudson, 1999). Cullen(2001) has encouraged high risk industries to reducetheir reliance on accident and incident data and to direct health and safety systems towardsinvestigating the culture and climate that may contribute to incidents. O’Toole (2002)identified that safety culture as a critical factor that sets the tone for the importance of safetywithin an organization Therefore, the following hypothesis is proposed.

H1: Safety culture has a positive effect on perceived safety performance

Helmreich & Merritt (1998) developed a model of interrelationships of elements that can leadto a safety culture and safety behaviors. This model can also determine the style of trainingand the nature of training delivery, the profession culture may influence the safety culturethrough feelings of responsibility. Clarke (2000) defined the term safety culture, and proposeda theoretical model by which safety culture affects safety behaviors in organizations.Farringto-Darby (2005) also identified forty main factors that influence safe behavior and safeculture. Hofmann and Stetzer (1996) indicated six broad categories of unsafe behaviorincluding improper tool use, improper work strategies at risk to self, failure to wear personalprotective equipment; improper storage of tools; improper storage by others; and improperwork strategies with risk to others. Thus, the following hypothesis is warranted.

H2: Safety culture has a positive effect on safety behavior

Improvements in organizational structure, organizational importance of safety, safety


4

responsibility and accountability, communication, management behavior, employeeinvolvement, and employee responses and behavior can help improve safety performance(Erickson ,2000). Vredenburgh (2002) indicated that safety behavior is often tied to quality ofperformance. Thus, an added benefit of safety behavior maybe productivity improved.Garavan and Obrien (2001) found unsafe acts or behaviors are a major causal factor inworkplace accident/injures. Improved safety behavior reduces the frequency of work relatedaccident and injuries. This phenomenon induces the following hypothesis.

H3: Safety behavior has a positive effect on perceived safety performance

2.2. Questionnaire design

After a review of previous literature and personal interviews with railway safety experts, thestep-by-step stages of the questionnaire design were based on Churchill (1991). The first partof questionnaire, safety behavior, was based on the research of Pousette et al. (2003) andLarsson et al. (2008). The measurement items of safety culture were developed according tothe questionnaires proposed by Fernández-Muñiz (2007) and Díaz-Cabrera et al, and themeasurement items of safety performance were based on Neal’s (2000) study. Respondentswere required to indicate agreement with the importance of each statement on a 5-pointLikert scale ranging from 1 (strongly disagree) to 5 (strongly agree).The questionnaires werepre-tested and revised. Questions in the questionnaire were based on previous studies anddiscussions with a number of railway safety executives and experts. 33 attribute variableswere selected for the questionnaire (see Table 1).

Table 1 could be inserted here

2.3 Sampling processes

This study aims to investigate the relationships among safety culture, safety behavior, andsafety performance in railway industry. To collect primary data in the railway industry, TaiwanRailway Administration (TRA) is considered as a case study. The anonymous questionnaireswere distributed to all TRA drivers in April, 2008. Researchers first sent the anonymousquestionnaires to 7 depots and then asked the supervisors for distributing the questionnairesto drivers during monthly on-the-job training occasion. Eventually, 537 usable samples werecollected (return rate=53.7%).

2.4 Data analysis approach

The research methods include factor analysis, reliability test, and structural equationmodeling (SEM). SEM simultaneously explains the pattern of a series of inter-relateddependence relationships between a set of latent (unobserved) constructs, each measured


5

by one or more observed variables (Hair et al., 2006). SEM includes one or more linearregression equations that describe how the endogenous constructs depend upon theexogenous constructs. Their coefficients are called path coefficients, or sometimesregression weights (Reisinger and Turner, 1999). In addition, SEM is an applicable methodfor analyzing hypotheses as it allows for the analysis of multiple relationships simultaneouslyand provides measures of overall model fit. It also explains the significance of each of therelationships between the variables (Kline, 1998).

3. Results of empirical analyses

The analyses were carried out using the SPSS to conduct descriptive statistics analysis,factor analysis, and reliability analysis, as well as the AMOS statistical packages for SEMmodel analysis to examine the overall fit of the empirical data to the hypothesized model.

3.1 Descriptive statistics result analysis

3.1.1 Profile of respondents

Of all the valid samples, slightly over half of the respondents were aged between 50 to 59years of age (51.4%). With regard to education level , the majority were high school (60.5%).Regarding to working hour, most respondents worked 40 to 49 hours per week. In addition,approximately 54.4% of respondents have a seniority of over 20 years, indicating that manyof TRA drivers were senior drivers ( Table2).


3.1.2 Importance of attribute variables of the dimensions: safety culture, safety behavior andsafety performance

An evaluation of aggregated drivers evaluation on the key factors in safety culture, safetybehavior and safety performance, which has a higher average value is “ Drivers couldactively propose positive suggestions and recommendations to improve organizationalperformance (4.02) in the dimension of safety behavior . This indicates drivers perceive thatcommunication between the staff and the manager level is considered as the most importantservice attribute.

In the dimension of safety culture, “Periodical external evaluations (audits) and preventionmanagement system “ (mean = 3.66) was perceived as the most important factor indicatingindependent safety audit system is essential for forming TRA’s safety culture.


6

With respect to the perceived safety performance, the attribute variable “I help my coworkerswhen they are working under risky or hazardous conditions (mean = 4.30)” is a key issuedetermining safety performance perceived by the TRA’s train drivers. This indicates thatpreventing risk is of importance to improve safety performance.

3.2 Exploratory factor analysis

Factor analysis is a technique used to reduce a large set of variables to a smaller set ofunderlying factors, helping to detect the presence of meaningful patterns among the originalvariables and extracting the crucial dimensions. Principal components analysis withVARIMAX rotation was employed to identify key dimensions, and factor loadings greater than0.50 are statistically significant (Hair et al., 2006). This study uses 33 attribute variables toconduct factor analysis and 6 factors are extracted.

3.2.1 Factor analysis result and factor labeling

We have three main dimensions: safety culture, safety behavior and safety performance forexamining their relationships. According to the exploratory factor analysis, safety culturedimension includes that communication and emergency response and safety management.Safety behavior dimension includes structural safety behavior and personal safety behavior.Perceived safety performance includes safety compliance and participation and perceivedaccident rate. The Cronbach’s α values of all factors are greater than 0.6.

3.3 Structural equation modeling analysis

Some indices are used to evaluate the proposed SEM model such as 2x value, thegoodness of-fit index (GFI) value, the adjusted goodness-of-fit index (AGFI), and the rootmean square error of approximation (RMSEA) (Hair et al., 2006). The value of the Chi-square

( 2x value = 10.33) is not significant at the 1% level, which implies the differences betweenpredicted and actual matrices are not significant and demonstrates the model’s fitness to thedata collected. The goodness of-fit index (GFI) value is 0.994. After adjustment for thedegrees of freedom relative to the number of variables, the adjusted goodness-of-fit index(AGFI) is 0.977, implying that 97% of the variances and covariance in the data observed arepredicted by the estimated model. The overall modified model fit is found adequate afterprocessing a normalized residual analysis. The p value is significant (P = 0.11 > 0.05), andtherefore the model is credited. The RMR = 0.109 and RMSEA = 0.04, implying that themodel has a model-fit evaluation with an appropriate goodness-of-fit (see Figure 1).


7

AMOS program could also provide the assessment of the unidimensionality of the model.Unidimensionality could be verified through the overall goodness of fit of the confirmatoryfactor model, and through the convergence and discrimination of items (Anderson andGerbing, 1988). The fitness of measurement model of the items on the constructs is almost inthe appropriate direction and statistically significant at the 0.05 level. The overall model fit isfound to be appropriate, which supports unidimensionality (see Figure 1).

Figure 1 could be inserted here

Convergent validity could be assessed by t values that are statistically significant on thefactor loadings. Our result indicates almost all the t values are greater than 1.96 or smallerthan -1.96, which implies statistical significance (Segar 1997; Byrne, 2001). The larger thefactor loadings or coefficients are, as compared with their standard errors, the stronger is theevidence that there is a relationship between the measured variables and their constructs(Koufteros, 1999). Table 3 reveals that most of the variables exceed the t value at the 0.05level of significance implying that the majority of variables are significantly related to theirspecified constructs.


3.3.1 The influence of safety culture on safety behaviorThe analysis results show that safety culture would influence safety behavior. That is,improving the internal safety culture is able to improve employees’ cognition and attitudetoward safety behavior. Through structural equation modeling (SEM), this study discoversthat the path coefficient from safety culture to safety behavior is 0.10 (t-value=5.14>1.96),indicating that safety culture is positively related to safety behavior of TRA. H1 is thussupported, and this result is consistent with previous literature. Particularly, “communicationand responsiveness” and “safety management” are the most significant factors to evaluatesafety culture and both positively affect safety culture.

3.3.2 The influence of safety culture on safety performance.According to Cooper (2000) and Williams (2003), safety culture has a significant and positiveinfluence on organizational safety performance. In this study, the path coefficient from safetyculture to safety performance is 0.11 (t-value=3.58>1.96), indicating that safety culture ispositively and significantly related to safety performance of TRA. Namely, with a maturesafety culture, TRA is capable of improving the safety performance, and that is consistentwith previous research. H2 in this study is supported as well. In addition, “safety compliance


8

and participation” affects safety performance most, so the coefficient is assigned as 1, andthe path coefficient of “accident rate” is hence 0.98. The t-values of both factors are greaterthan 1.96, indicating these two factors significantly and positively influence safetyperformance.

3.3.3 The influence of safety behavior on safety performancePrevious research has proven that improving safety behavior is helpful for safetyperformance (Grindle, Dickinson, and Boettcher, 2000; McAfee and Winn,1989;Sulzer-Azaroff, Harris, and Blake-McCann, 1994). This empirical results show that the pathcoefficient from safety behavior to safety performance is 0.70 (t-value=3.03>1.96), whichindicates a significant and positive relationship between safety behavior and safetyperformance of TRA. Moreover, H3 in this study is supported. In this study, “organizationalsafety behavior” has the greatest impact on safety behavior, and the coefficient is assignedas 1. Therefore, the coefficient of “personal safety behavior” is 2.79. The t-values of bothfactors are greater than 1.96, indicating these two factors have a significant and positiveinfluence on safety behavior.


4. Conclusions and discussion

The present study provides the empirical evidence to support the relationship among thesafety culture, safety behavior and perceived performance of railway drivers. Our empiricalresult indicates the positive safety culture can influence on the safety behavior of railwaydriver. In addition, the positive safety culture was also found to have a positive effect on thepositive safety perceived performance. The safety behavior was found to have a positiveeffect on the perceived safety performance, which corresponds well the finding of Grindle,Dickinson and Boettcher (2000); McAfee and Winn (1989); Sulzer-Azaroff, Harris, andBlake-McCann(1994).

Of the latent construct “safety culture”, communication and emergency and safetymanagement were considered as two essential variables affecting safety culture, which isconsistent with Brown et al., 2000; Demichela and Piccinini, 2006; Ferna’ndez-Mun˜ iz,Manuel Montes-Peo’n, and Va’ zquez-Orda’s (2007) ;Hofmann and Stetzer, (1996) ; Mearnset al.,(2003); Zohar, (1980) and the relationship between safety culture and communicationand emergency was justified, which is consistent with Farringto-Darby, 2005; D´ıaz-Cabreraet al, (2007);Fernandez-Muniz (2007).


9

Our finding also shows structural safety behavior and personal safety behavior are two mainvariables affecting the construct of safety behavior (Pousette et al., 2003; Larsson et al,2008).In regard to the perceived safety performance, safety compliance and participation andperceived accident rate were deemed as the two essential factors.

5. Managerial implications

These findings contribute several implications. First, the executives of TRA should set up aclear goal for the entire organization to protect passengers’ safety, such as no injuries areallowed to happen in the short run. In the middle and the long term, because the mostimportant factors affecting safety behavior are “organizational safety behavior” and “personalsafety behavior”, the regulated safety behavior should be propagated to the staff. Employees’personal safety behavior should also be inspected in order to establish relative reward andpunishment regulations. Finally, “communication and responsiveness” and “safetymanagement” are the most important factors affecting safety culture. TRA is thus suggestedto establish a seamless communication channel between staff and executives. To ensure theeffectiveness of the communication channel, each case should be documented. Furthermore,TRA can set up a specific supervising department to periodically inspect the performanceimprovement and modify the inappropriate regulations and unfinished objectives.

5.1 Limitations and future research

Due to the limited time and budget, this study addressed the train drivers to investigate thecause-effect relationship among safety culture, safety behavior and perceived safetyperformance. The future study could investigate the multi-level effect of safety culture, safetybehavior on the safety performance to examine the hierarchical perceived difference. Inaddition, the future study could investigate various types’ railway companies’ safety cultureand safety behavior on safety performance such as the difference in the transit system,conventional railway system and high-speed rail system’s operators.


10

Reference

1. Adie, W., Cairns, J., Macdiarmid, J., Ross, J., Watt, S., and Taylor, C. L., (2005). “Safetyculture and accident risk control: Perceptions of professional divers and offshoreworkers”. Safety Science, Vol. 43, NO.2, pp.131-145.

2. Baysari, McIntosh and Wilson ., (2008).“Classification of errors contributing to railincidents and accidents: A comparison of two human error identification techniques”.Safety Science. Article in Press.

3. Berends, J.J., (1996).“On the Measurement of Safety Culture (Unpublished graduationreport)”. Eindhoven University of Technology, Eindhoven.

4. Bott, K.G., (1996). “Reports C1, C2, C3a and C4a: driver behaviour”. Rail Safety andStandards Board Research Catalogue.

5. Brown, K.A. Brown, P.G. ,Willis and G.E. Prussia .,(2000).“Predicting safe employeebehavior in the steel industry: development and test of a sociotechnical model”, Journalof Operations Managemen .Vol.18, NO.4, pp. 445–465.

6. Buck, L., (1963).“Errors in the perception of railway signals”. Ergonomics, VOL .6,pp.181–192.

7. Cacciabue, P.C., (2005a).“Human error risk management methodology for safety auditof a large railway organization”. In: Wilson, J., Norris, B., Clarke, S., Mills, A. (Eds.), RailHuman Factors: Supporting the Integrated Railway. Ashgate Publishing Limited,Cornwall.

8. Cacciabue, P.C., (2005b). “Human error risk management methodology for safety auditof a large railway organization”. Applied Ergonomics, Vol. 36, No. 6, and pp.709–718.

9. Campbell, J.P., McCloy, R.A., Oppler, S.H., and Sager, C.E., (1993). “A theory ofperformance”. In: Schmitt, J., Borman, W.C. Associates. Personnel Selection inOrganizations. Jossey-Bass, San Francisco, CA, pp. 35–69.

10. Choudhry, Fang., and Mohamed., (2007).“The nature of safety culture: A survey of thestate-of-the-art”. Safety Science.Vol.45, NO.10, pp. 993-1012.

11. Clarke, S., (1999). “Perceptions of organizational safety: implications for thedevelopment of safety culture”. Journal of Organizational Behaviour. Vol.20, NO.2,pp.185–198.

12. Confederation of British Industry (CBI) (1991).“Survey of Students ‘Attitudes: 17 and 18Year Olds Going on to Higher Education”. London: CBI.

13. Cooper, C.L., and Sutherland, V.J., (1987). “Job stress, mental health, and accidentsamong offshore workers in the oil and gas extraction industries”. J. Occup. Vol.29, NO.2,pp.119–125.

14. Cooper, M.D., (2000). “Towards a model of safety culture”, Safety Science, Vol. 36, No.2,pp.111-136.


11

15. Cooper, M. D., and Phillip, R.A... (2004). “Exploratory analysis of the safety climate and

safety behavior relationship＂.Journal of Safety Research, Vol.35, No.5, pp.497-512.

16. Cox, S., and Cox, T. (1991), “The structure of employee attitudes to safety: A Europeanexample”. Work and Stress, Vol.5, No.2, pp.93–106.

17. Crick, J., (2004). “Driver error data collection project”. Rail Safety and Standards BoardResearch Catalogue.

18. Cullen, W.D., (1990). “The Public Inquiry into the Piper Alpha Disaster”, HMSO, London.19. Davies, L.J., (1988). “Understanding organizational culture: a soft systems perspective”.

Systems Practice .Vol.1, NO.1, pp.11–30.20. Demichela and Piccinini., (2006).“How the management aspects can affect the results of

the QRA ”. Journal of Loss Prevention in the Process Industries. Vol 19, NO.1, pp.70-77.21. D´ıaz-Cabrera, D., Hern´andez-Fernaud, E and Isla-D´ıaz, R.,(2007).“An evaluation of a

new instrument to measure organizational safety culture values and practices”. AccidentAnalysis and Prevention. Vol 39, NO.6.pp.1202–1211.

22. Donald, I., and Canter, D., (1993). “Psychological factors and the accident plateau”.Health and Safety Information Bulletin. Vol.215, NO., pp.5–8.

23. Donald, I., and Canter, D., (1994). "Employee attitudes and safety in the chemicalindustry", Journal of Loss Prevention in the Process Industries. Vol.7, NO.3,pp.203–208.

24. Edkins, G.D., and Pollock, C.M., (1996). “Pro-active safety management: applicationand evaluation within a rail context”. Safety Science. Vol. 24, No. 2, pp.83–93.

25. Edkins, G. D. (1998). “The indicate safety program: evaluation of a method to proactivelyimprove airline safety performance”. Safety Science, Vol. 30, No.3, pp.275-295.

26. Elms, D. (2001). “Rail Safety”. Reliability Engineering and System Safety .Vol. 74, No.3,pp.291-297.

27. Erickson, J.A. (2000). “Corporate culture: The key to safety performance”. OccupationalHazards .Vol.62, NO.4, pp. 45–50.

28. Farrington-Darby, T., Pickup, L. and Wilson, J. R. (2005). “Safety culture in railwaymaintenance”. Safety Science. Vol. 43, No.1, pp.39–60.

29. Ferna’ n dez-Mun iz ,B. Montes-Peo’n , J. M , and Va’ zquez-Orda’ s, C. J.(2007).“Safety culture: Analysis of the causal relationships between its key dimensions”.Journal of Safety Research.Vol 38, NO. 6, pp. 627-641.

30. Ferna’ n dez-Mun iz ,B. Montes-Peo’n , J. M ., and Va’ zquez-Orda’ s, C. J.(2007),“Safety management system: Development and validation of a multidimensional scale”.Journal of Loss Prevention in the Process Industries. Vol. 20, No.1, pp.52–68.

31. Frost. and McKelvie, S. (2004).“Self-esteem and body satisfaction in male and femaleelementary school, high school, and university students”. Sex Roles, Vol.51, NO.1-2,


12

pp.45-54.32. Garavan, T.N. and O'Brien, F., (2001). “An investigation into the relationship between

safety climate and safety behaviors in Irish organizations”. Irish Journal ofManagement .Vol.22, NO.1, pp. 141–170.

33. Geller, E.S. (1994). “Ten principles for achieving a total safety culture”. ProfessionalSafety, Vol.39, N0.9, pp. 18–24.

34. Gilroy, J., and Grimes, E., (2005). “The development and application of a rail humanreliability assessment tool”. In: Paper Presented at the Second European Conference onRail Human Factors.

35. Glennon, D.P., (1982a).“Measuring organizational safety climate”. Australian SafetyNews. pp.23-28.

36. Glennon, D.P., (1982b).“Safety climate in organizations”. Proceedings of the 19thAnnual Conference of the Ergonomics Society of Australia and New Zealand, pp.17-31.

37. Grote, G., and Ku¨nzler, C. (2000).“Diagnosis of safety culture in safety managementaudits”. Safety Science, Vol.34, NO.1-3, pp.131–150.

38. Grindle et al., . A.C. Grindle, A.M. Dickinson and W. Boettcher.,(2000). “Behavioralsafety research in manufacturing settings: a review of the literature”. Journal ofOrganizational Behavior Management .Vol.20, NO.1, pp. 29–68.

39. Guastello, S.J., (1989).“Catastrophe modeling of the accident process: evaluation of anaccident reduction program using the Occupational Hazards Survey”. Accident Analysis& Prevention. Vol.21, NO.1, pp. 61–77.

40. Guastello, S.J., (1991).“Psychosocial variables related to transit accidents: acatastrophe model”. Work Stress .Vol.5, NO.1, pp. 17–28.

41. Harper, R.S., and Koehn, E. (1997). “Managing industrial construction safety insoutheast Texas.” Journal of Construction Engineering and Management, Vol.124, NO.6, pp.452–457.

42. Havold, J.I., and Nesset, E., (2008). “From safety culture to safety orientation: Validationand simplification of a safety orientation scale using a sample of seafarers working forNorwegian ship owners”. Safety Science.

43. Health and Safety Commission, (1993). Organizing for Safety. Third Report, HumanFactors Study Group, Advisory Committee on the Safety of Nuclear Installations. HMSO,London.

44. Hofmann, D.A and Stetzer, A., (1996).“A cross-level investigation of factors influencingunsafe behaviors and accidents”. Personnel Psychology .Vol.49, NO.2, pp. 307–339.

45. Holcom, M.L., Lehman, W.E.K., and Simpson, D.D., (1993).“Employee accidents:influences of personal characteristics and substance use in jobs differing in accidentpotential”. Journal of Safety Research. Vol.24, NO.4, pp.205–221.

46. Hurst, N. W., Young, S., Donald, I., Gibson, H., and Muyselaar, A. (1996). “Measures of


13

safety management performance and attitudes to safety at major hazard sites”. Journalof Loss Prevention in the Process Industries. Vol.9, NO.2, pp.161–172.

47. Hudson, P., (1999). “Safety culture–theory and practice. The human factor in systemreliability—is human performance predictable?” Siena RTO MP-032, pp.1-12.

48. International Atomic Energy Authority, Viennahttp://www.iaea.org/NewsCenter/Focus/Chernobyl/index.shtml

49. Johnson, G., (1992). “Managing strategic change — strategy, culture and action”. LongRange Planning. Vol.25, NO.1, pp. 9–19.

50. Crooks, K.J., Baker, D.P., 2007. Preface to the special section on classifying andunderstanding human error. Human Factors 49 (2), 175–177.

51. Larsson, S., Pousette, A., and Torner, M., (2008). “Psychological climate and safety inthe construction industry-mediated influence on safety behaviour”. Safety Science,Vol.46, NO.3, pp.405–412.

52. Lantz, B.M., Blevins, M.W., (2002). “An analysis of commercial vehicle driver trafficconviction data to identify high safety risk motor carriers. In: Zach Zacharia(Ed.)”.Proceedings of International Truck and Bus Safety Research and PolicySymposium, Tennessee, USA. pp. 549–556.

53. Little, A.D., (1996), “Assessment of driver–signalman communication risk”. Rail Safetyand Standards Board Research Catalogue.

54. Marottoli, R.A., Cooney, L.M. and Tinetti, M.E., (1997). “Self-report versus state recordsfor identifying crashes among older drivers”. Journals of Gerontology Series A:Biological Sciences and Medical Sciences .Vol.52, NO.3, pp. 184–187.

55. Mattila, M., Rantanen, E. and Hyttinen, M. (1994),“The quality of work environment,supervision and safety in building construction”, Safety Science, Vol.17 , NO.4 ,pp.257-268.

56. McAfee, R.B.and Winn, A.R. (1989), “The use of incentives/feedback to enhance workplace safety: A critique of the literature”, Journal of safety Research, Vol.20, NO.1,pp.7-19.

57. McDonald et al. (2000). “Safety management systems and safety culture in aircraftmaintenance organizations”. Safety Science. Vol.34, NO.1-3, pp.151-176.

58. Mearns et al. (2003). “Safety climate, safety management practice and safetyperformance in offshore environments”. Safety Science, Vol.41, NO.8, pp.641–680.

59. Mitchison, N., and Papadakis, G. A. (1999). “Safety management system under SevesoII: Implementation and assessment”. Journal of Loss Prevention in the ProcessIndustries, Vol.12, NO.1, pp. 43–51.

60. Murphy, P., (2001), “The role of communications in accidents and incidents during railpossessions”, Rail Safety and Standards Board Research Catalogue.

61. Neal, A., Griffin, M.A., and Hart, P.M., (2000). “The impact of organizational climate on

http://www.iaea.org/NewsCenter/Focus/Chernobyl/index.shtml


14

safety climate and individual behavior”, Safety Science, Vol. 34, NO.1-3, pp.99-109.62. Neal, A., and Griffin, M. A. (2002), “Safety climate and safety behavior”, Australian

Journal of Management, Vol.27, NO.1-3, pp.67-77.63. Oliver et al., A. Oliver, A. Cheyne, J.M. Tomas and S. Cox., (2002). “The effects of

organizational and individual factors on occupational accidents”. Journal of Occupationaland Organizational Psychology. Vol. 75, NO.4, pp. 473–488.

64. O’Hare, D., 2000. The ‘Wheel of Misfortune’: a taxonomic approach to human factors inaccident investigation and analysis in aviation and other complex systems. Ergonomics43 (12), 2001–2019.

65. O'Reilly III, Chatman and Caldwell (1991). “People and organizational culture: A profilecomparison approach to assessing person-organization Fit”.Academy of ManagementJournal, Vol.34, NO.3, pp.487-516.

66. Ostrom, L., Wilhelmsen, and C., Kaplan, B., (1993).“Assessing safety culture”. NuclearSafety. Vol.34, NO.2, pp.163-172.

67. Petersen, D., (2000). “Safety Management 2000: Our Strengths and Weaknesses”.Professional Safety. Vol.45, NO.1, pp.16-19.

68. Porter, D., (1992). “A systematic human error analysis of the train driving task”. RailSafety and Standards Board Research Catalogue.

69. Pousette, A., Torner, M., and Larsson, S., (2003).“Associations between safety climate,psychosocial work environment and self-reported safety behaviour”. In: The NordicErgonomic Society 35th Annual Conference, Reykjavik, Iceland. pp.10–13.

70. Reason .J.(1994). “REVIEW”, Management Overview British Railway Board, Vol.1,London.

71. Reason .J.(1995). “A System Approach to Organizational error”. Ergonomics, Vol 38,NO.8, pp.1708-1721.

72. Relf, A.P.(2004).“Achieving Software Quality through Source CodeReadability”.Qualcon .pp.1-18.

73. Richter, A and Koch, C., (2004) .“Integration, differentiation and ambiguity in safetycultures”. Safety Science. Vol.42, NO.8, pp. 703–722.

74. Rundmo .T., Hestad.H and Ulleberg. P.(1998). “Organizational factors, safety attitudesand workload among offshore oil personnel”, Safety Science, Vol.29, NO.2, pp.75-87.

75. Sawacha, E, Naoum, S., and Fong ,D .(1999).“Factors affecting safety performance onconstruction sites”. International Journal of Project Management Vol. 17, No.5, pp.309-315.

76. Schein, E., (1990). “Organizational culture”, American Psychologist, Vol. 45, NO.2,pp.109-119.

77. Shanahan, P., Gregory, D., Shannon, M., and Gibson, H., (2005).“The role ofcommunication errors in railway incident causation”, In: Paper Presented at the Second


15

European Conference on Rail Human Factors.78. Seo,Dong-Chul .(2004). “A cross-validation of safety climate scale using confirmatory

factor analytic approach”, Journal of Safety Research , Vol. 35 , NO.4, pp.427– 445.79. Sulzer-Azaroff .,Austin,. B. Sulzer-Azaroff and Austin, J., (2000).“Does BBS work?

behavior-based safety & injury reduction: a survey of the evidence”. Professional Safety .Vol...45, NO.7 , pp.19–24.

80. Tomas, J.M. Tomas, J.L. Melia and Oliver, A ., (1999). A cross-validation of a structuralequation model of accidents: organizational and psychological variables as predictors ofwork safety. Work and Stress. Vol.13, NO.1, pp. 49–58.

81. Toole, M., (2002).“The relationship between employees' perceptions of safety andorganizational culture ”. Journal of Safety Research .Vol 33, NO. 2, pp. 231-243 .

82. Turner, B.A., (1988).“Connoisseurship in the study of organizational cultures”. InBryman, A. (Ed.), Doing Research in Organizations. Routledge, London, pp. 108–122.

83. Turnerb, A., Pidgeonn, F., Blockledy, L., and Tom, B., (1989). “Safety culture: itsimportance in future risk management”. Position paper for the Second World BankWorkshop on Safety Control and Risk Management, Karlstad, Sweden, November.

84. Uttal, W.R., (1983). “Visual form detection in 3-dimensional space”. Hillsdale, NewJersey: Lawrence Erlbaum.

85. Vanderhaegen, F., (2001).“A non-probabilistic prospective and retrospective humanreliability analysis method—application to railway system”. Reliability Engineering andSystem Safety ,Vol. 71 , NO.1, pp.1-13.

86. Vredenburgh,A.G., (2002).“Organizational safety: Which management practices aremost effective in reducing employee injury rates ”. Journal of Safety Research.Vol 33,NO. 2, pp. 259-276 .

87. Williams, J.H., (2003).“People-based safety”.Professional Safety. Vol. 48, NO. 2,pp.1-32.

88. Wills,A.R., Watson, B., and Biggs, H.C.,(2006).“Comparing safety climate factors aspredictors of work-related driving behavior”. Journal of Safety Research.Vol 37, NO. 4,pp. 375-383 .

89. Wong, K.W., Chan, P.C., and Lo, K.K., (1999).“Factors affecting the safety performanceof contractors and construction sites. Proceedings of the second internationalconference of CIB working commission W99”. Honolulu/Hawaii . pp.19–23.

90. Zohar, D. (1980), “Safety Climate in Industrial Organization : Theoretical and AppliedImplications” , Journal of Applied Psychology.Vol.65, NO.1 , pp.96-102.


16

2.79(5.48)

0.37(15.29)

1.00(---)

0.98(8.61)

1.00(---)

1.00(---)

0.70*(3.03)

0.11*(3.58)

0.10*(5.14)

SafetyCulture

SafetyBehavior

Personal SafetyBehavior

Structural SafetyBehavior

SafetyPerformance

SafetyCompliance and

Participation

Accident Rate

CommunicationAnd emergency

SafetyManagement

χ2 ＝10.33 df=6, P value ＝0.11 GFI ＝0.994, AGFI＝0.977,RMSR＝0.109, CFI＝0.995

Note: Path coefficients with * are critical ratios exceeding 1.96 at the 0.05 level of significance

(t statistics).

Figure 1 SEM analytical result


17

Table1 Attributes variables affecting the safety culture, safety behavior and perceived safetyperformance

Safety behavior attribute variables Mean SDDrivers should actively propose positive suggestions andrecommendations to improve organizational performance 4.02 0.64

Drivers should seek for mistakes and errors to report tosupervisors 3.92 0.67

We have written rules and procedures that guide creativeproblem solving 3.57 0.76

Our workers have the authority to correct problems when theyoccur 3.44 0.90

Drivers should meet safety quality demand, regardless ofreasonability 3.39 1.05

We believe that when making decisions, the overall effects of adecision should be considered 3.21 0.87

We believe that employees from one department should workwith employees from other departments 2.91 0.91

Safety culture attribute variables Mean SDExternal evaluations (audits) periodically conducted of validityand reliability of prevention management system 3.66 0.80

Written circulars elaborated and meetings organized to informworkers about risks associated with their work and how toprevent accidents.

3.65 0.70

Management has established in writing the functions ofcommitment and participation and the responsibilities in safetyfor all organization members

3.64 0.83

All workers informed about emergency plan. 3.58 0.76At the introduction of new system, organization will negotiatewith system providers in conducting educational training coursesfor handling instructions

3.58 0.88

When accidents occur, current emergency handling plan shouldbe able to solve such incidents 3.53 0.78

Written declaration is available to all workers reflectingmanagement’s concern for safety, principles of action andobjectives to achieve.

3.53 0.77

Organizational safety policy can focus on safety problem issues 3.52 0.82All workers informed about emergency plan 3.51 0.78There is a fluent communication embodied in periodic andfrequent meetings, campaigns or oral presentations to transmitprinciples and rules of action.

3.50 0.90

Accidents and incidents reported, investigated, analyzed andrecorded. 3.49 0.79

When starting in new job position worker provided writteninformation about procedures and correct way of doing tasks. 3.49 0.81

Firm coordinates its health and safety policies with other HRpolicies to ensure commitment and well-being of workers. 3.48 0.87

Station staff and drivers maintain a good communication 3.44 0.82When performing regular safety inspections, organization canimmediately handle appropriately to solve safety problems 3.43 0.78

Frequent use of teams made up of workers from different partsof organization to resolve specific problems relating to workingconditions.

3.26 0.93

Resolutions frequently adopted that originated fromconsultations with or suggestions from workers. 3.20 0.92

When organization adopt staffs’ safety proposals, rewards are 3.00 1.01


18

givenSafety performance attribute variables Mean SD

I help my coworkers when they are working under risky orhazardous conditions. 4.30 0.62

I ensure the highest levels of safety when I carry out my job. 4.21 0.58I use the correct safety procedures for carrying out my job. 4.12 0.59I voluntarily carry out tasks or activities that help to improveworkplace safety. 3.98 0.59

Over the past year, I strengthened my emergency handlingcapability 3.77 0.63

Over the past year, the occurrences of railway relatedresponsible or none responsible incidents increased significantly 3.50 0.77

Over the past year, death and injuries of railway staff during dutydecreased significantly 3.48 0.72

Over the past year, break-downs of railway machineryequipment decreased significantly 3.24 0.85


19

Table 2. Profile of respondentsCharacteristics of the respondents Number of

respondentsPercentage ofrespondents(%)

Age30-39 65 12.1040-49 184 34.3050-59 276 51.40More than 60 12 2.20Education backgroundJunior high school 14 2.60Senior high school 325 60.50University and above 198 36.90Train driving experienceLess than 5 years 34 6.306-10 years 66 12.3011-20 years 145 27.00More than 20 years 292 54.40PositionDriver 496 92.40Chief driver 28 5.20Driver assistant 13 2.40Working hours per week40-49 413 76.9050-59 113 21.00More than 60 11 2.00


20

Table 3 Parameter estimates, standard errors and critical ratios for the revised model

Latentvariable Item

Unstandardizedfactor loading

Completelystandardizedfactorloading

Standarderror(a)

CriticalRatio(b)

Safetyculture

Communicationand emergencyresponse

1.000 0.946 -- --

Safetymanagement 0.367 0.699 0.024 15.288

Safetybehavior

Structuralsafety behavior 1.000 0.316 -- --

PersonalSafety behavior 2.794 0.925 0.509 5.487

Perceivedsafetyperformance

Safetycomplianceandparticipation

1.000 0.528 -- --

Perceivedaccident rate 0.984 0.585 0.114 8.605

aS.E is an estimate of the standard error of the covariance.bC.R.is the critical ratio obtained by dividing the estimate of the covariance by its standarderror. A value exceeding 1.96 represents a level of significance of 0.05.cIndicates a parameter fixed at 1.0 in the original solution.


21

Table 4. The estimates of the modified SEM model

Variables Estimate

S.E.(a) C.R.(b) P

Regression

Safety culture → Safety behavior 0.10 0.02 5.140.001

Safety culture → Perceived safetyperformance

0.11 0.03 3.580.039

Safety behavior →Perceived safetyperformance

0.70 0.23 3.030.001

aS.E. is an estimate of the standard error of the covariance.bC.R.is the critical ratio obtained by dividing the covariance estimate by its standard error;cUnderlined values are critical ratios exceeding 1.96 at the 0.05 level of significance.

χ2 value＝10.33 (P＝0.11), df＝6 ,χ2/df＝1.667, GFI＝0.994, AGFI＝0.977, RMSEA＝0.04

Documents

Safety culture, Safety behavior and safety performance in ... · Safety culture, Safety behavior and safety performance in Railway companies Abstract Forming a positive safety culture