"Old" and "new" safety thinking: Perspectives of aviation safety investigators. (May 2020)

• Record Type: Journal Article
• Authors: Karanikas, Nektarios; Chionis, Dimitrios; Plioutsias, Anastasios
• Is Part Of: Safety science. Volume 125(2020)
• Publication Date: 2020-05
• Subjects: Old safety thinking -- New safety thinking -- Human error -- Safety investigations; Industrial accidents -- Periodicals; Accident Prevention -- Periodicals; Safety -- Periodicals; Travail -- Accidents -- Périodiques; 363.11
• DOI: 10.1016/j.ssci.2020.104632
• Languages: English
• ISSNs: 0925-7535

Abstract

Highlights: Literature tends to highlight a dichotomy between "old" and "new" safety thinking. A survey across safety investigators showed that this duality is invalid in practice. Investigators accept and apply partially aspects of both new and old safety thinking. Various factors drive investigation practices more than safety paradigms and models. Abstract: The development of various safety paradigms over time has led to public discussions that tend to highlight a dichotomy between the so-called "old" and "new" safety thinking. Although these two approaches might be based on opposite views that can feed debates and discussions, the degree to which they are binary in practice and respective explanations have not been adequately researched. Following a review of literature, we developed a framework that refers to nine aspects that denote new safety thinking practices pertinent to safety investigations and includes the three basic safety model categories: sequential, epidemiological and systemic. We administered a survey to examine the extent of agreement of safety investigators with statements reflecting the old and new safety thinking practices as well as the familiarity with and degree of application of the latter and the three safety model types above, and we collected respective comments. The 41 safety investigators who participated in the study were quite familiar and agreeable with the new safety thinking aspects. Overall, they had applied these aspects with a moderate frequency during investigations, without though abolishing practices related to the old paradigm due to time, resource, data and training limitations and cultural or managerial influences. Epidemiological models were the most frequently applied due to their optimum efficiency-thoroughness balance. In general, our findings suggested that the sample was not unanimously against or in favour of each of the old and new investigation practices included in the survey, this indicating that the duality between these two paradigms might not be valid in real-world settings. Although the results of this study cannot be generalised, this paper communicates insightful messages as well as recommendations and could function as an impetus for further research on this topic.