What’s the relationship between safety culture surveys and accident investigation findings?
Should we expect some similarities and alignment between the two – or even predictive validity of surveys to real-world performance?
Lots to be said here (and research), but tomorrow I’ll be posting a new study from Syd Dekker as co-author on how, in their sample, a safety culture maturity survey wasn’t predictive or aligned to a real-world accident that occurred mere weeks after the survey.
Another study from Antonsen in 2009 is relevant (see attached image; link to summary in comments).
Antonsen’s study found that both the safety culture assessment and resultant accident investigation provided two “dramatically different” descriptions of the same culture.
Further, the survey was unable to sensitise the organisation to impending failure. Therefore, this safety culture survey “may have little predictive value”.
1) Antonsen’s study: https://doi.org/10.1111/j.1468-5973.2009.00585.x
2) My summary of his study: https://www.linkedin.com/pulse/safety-culture-assessment-mission-impossible-ben-hutchinson
SafetyInsights.org 
One thought on “The (non)relationship between safety culture maturity and accidents?”