Defining, Diagnosing and Predicting Safety Culture

Health and safety

Chernobyl. Piper Alpha. BP Oil Disaster. These names may spark memories of news images relating to horrendous scenes of destruction, loss of life and vast environmental damage. Along with the shared element of carnage, the common thread throughout the post incident investigations was the focus on the contribution that Safety Culture (or lack thereof) played.

‘Safety Culture’ (or climate as it is commonly referred to, but let’s ignore the Culture vs Climate debate for now) has seen many definitions however the one I personally like is “(the) shared perceptions of work environment characteristics as they pertain to safety matters that affect a group of individuals” (1). Safety culture has received a lot of attention from different organisations, professionals and communities alike, which is justified due to the attributed influence of the aforementioned disasters.

Upon completing my Masters degree, I was employed to help manage, develop and evaluate an existing initiative aiming to influence and evolve the safety culture of a large coal fired power station. Since then I have been hired by a Mechanical Engineering contracting company to develop a similar initiative working on sites which range from steelworks and cement factories to power stations. Part of the work within this project has seen me use Safety Culture surveys to help evaluate sites and companies safety culture.

Many models of Safety Culture tend to follow either the HSE’s (Health and Safety Executive) Safety Culture Maturity model (2), DuPonts Bradley Curve (3) or heavily research based models such as Parker, Lawrie and Hudson’s (4) five-factor model of Safety Culture. All of these models see safety culture as a progressive maturation in a company, placing a company definitively at a certain level. For example, a company couldn’t partly be in the Proactive or Generative (the Holy Grail of Safety Culture) stage and partly in the lower level Reactive culture.

In reality I have found these models somewhat limiting in terms of diagnosing Safety Culture because in some areas I find a site may be proactive and in others reactive, limiting how I can accurately feedback what the actual state of a site’s culture might be. Nevertheless, they have been great for use in training or action groups to start discussions, find differences in views, and formulate ideas on how to improve.

Alongside the models there are certain trait-like theories examing which areas create more mature cultures. These theories break down Safety Culture into numerous areas which can vary between five and twelve areas with common elements centring on trust, communication, management commitment and using proactive indicators. A popular theory purported by Dr Tim Marsh (5), a leader in behavioural safety and linking Psychology and safety, states that proactive safety cultures are effective in managing six key areas, namely: communication, workforce involvement, modelling, challenging, analysis and use of lead (proactive) indicators.

I have found these better for guiding my individual actions compared with the maturity models. They do require greater care in explanation within training or communication setting and I have had to invest more time in considering the most appropriate language to use for different groups.

For all the interest in this area, there would seem to be little agreement any one theory of Safety Culture. For example what the Big Five (whether you agree or disagree with its claims) has done for personality, Safety Culture research lacks (though attempts have been made)6 and Guldenmung (7) warned we were not ready yet for such a theory. However over a decade later, Zohar (8) suggests that we may be ready now to go back and look at the theoretical underpinning of Safety Culture, given the vast amount of research done over the past thirty years.

Whilst I hope for a development in Safety Culture theory (‘there is nothing as practical as a good theory’) I still believe safety professionals need to better understand how safety culture actually influences work practice, how to effectively conduct a Safety Culture diagnosis, and what this information should be used for.

Currently, there seems to be mixed evidence that safety culture categorically predicts safety outcomes. Often, it is difficult to pin point the exact factors at play, as in reality achieving a safe culture is often dependant on many other factors such as safety motivation, safety knowledge, transformational leadership, the work context and production pressures (9). I would argue that these other areas are better predictors of performance and safety culture helps us to emphasise their importance.

Furthermore, different methods for defining Safety Culture could have a large impact on the way culture is defined. Antonson (10) looked at survey methods verses more qualitative methods of defining safety culture and how they produced vastly different views of culture. Within this was also a suggestion that the different methods may actually be picking up on different concepts within culture. This is an interesting observation and one that doesn’t instil me with confidence in predicting safety outcomes from safety culture surveys alone.

Along with this I suspect that the effective use of safety culture surveys could be improved upon. A report by the Keil Institute (11) showing that of 142 companies that took part in the review and used the HSE Safety Climate survey tool, only 13% used the benchmarking data to inform the results. From personal experience when I’ve enquired how certain sites or HSE departments have conducted their surveys, many have not used benchmarking nor have they repeated them over multiple periods of time. Both of these problems have lead to a restricted view of safety culture thus lacking reference points for effective interpretation.

These three issues together – predicting safety outcomes, methods of diagnosing safety culture and effective interpretation of survey data –makes me question how the energy and manufacturing industries I’ve worked in, use information relating to safety culture. Many safety professionals (including myself) I suspect, are tempted to use such information as ‘predictive measures’ of safety performance; however I question whether, because of the outlined state of Safety Culture knowledge and practice, this is appropriate.

I truly believe though that the concept and diagnosis of Safety Culture is a useful tool in the safety professional’s toolbox. It can help guide improvement plans, highlight areas of concern before issues arise and can safely start discussions surrounding controversial or sensitive subjects.

Whether it took the disasters mentioned to make high risk industries focus their interest in safety culture matters little now. Psychologists should be happy that safety culture research and diagnosis is another area where their skill-set can be put to effective and beneficial use.


1 – Christian, M, S., Bradley, J, C., Wallace, J, C., Burke, M (2009). Workplace Safety: A Meta-Analysis of the Roles of Person and Situational Factors. Journal of Applied Psychology. 94,5, 1103-1127.

2 – Keil Institute Report (2000) Safety Culture Maturity Model. HSE Report 2000/049

3 – DuPont Website accessed Jan 2014. DuPont Bradley Curve. –

4 – Parker, D., Lawrie, M., Hudson, P (2006). A Framework for Understanding the Development of Organisational Safety Culture. Safety Science. 44,6, 551-562.

5 – Marsh, D (2009) ASM: Affective Safety Management. IIRSM.

6 – Flin, R., Mearns, K., O’Connor, P., & Bryden, R. (2000). Measuring Safety Climate: Identifying the Common Features. Safety Science, 34, 177–193.

7 – Guldenmund, F. (2000). The Nature of Safety Culture: a Review of Theory and Research. Safety Science, 34, 215– 257

8 – Zohar, D (2010) Thirty Years of Safety Climate Research: Reflections and Future Directions. Accident Analysis & Prevention. 42, 1517-1522.

9 – Cooper, M, D., Phillips, R, A (2004) Exploratory Analysis of the Safety Climate and Safety Behaviour Relationship. Journal of Safety Research. 35, 497-512.

10 – Antonsen, S (2009). Safety Culture Assessment: Mission Impossible? Journal of Crisis Management. 17,4, 242-254.

11 – Keil Institute Report (2002). Evaluating the Effectiveness of Health and Safety Executive’s Health and Safety Climate Survey Tool. HSE Report 2002/042

Aaron Percival Bio

With an MSc in Occupational Psychology from Nottingham University, Aaron joined Power Industrial Group Ltd, an independent Mechanical Engineering Contractor. He has been responsible for developing a companywide behavioural safety initiative with additional work relating to apprentice wellbeing and advising on human factors. Main interests include Safety culture, decision making, human factors and accident causation.