A Systematic Review and Delphi Study on Key Factors Influencing the Selection of Accident Analysis Methods Publisher



Pajoohnia M ; Ghasemi F ; Mahmoudi Herris S ; Omidi L
Source: Journal of Safety and Sustainability Published:2026

Abstract

Safety science has developed numerous accident causation models to investigate safety-critical operations. This study implemented a qualitative-quantitative framework to systematically determine optimal criteria for selecting accident analysis methods. The research adopted a two-phase approach, first identifying key selection factors through a comprehensive literature review and systematic categorization of documented criteria, and then validating these findings through expert consultation. The validation phase employed the fuzzy Delphi method, a structured qualitative technique that gathered insights from domain experts, academics, and industry practitioners via systematic questionnaire interviews. To quantify expert agreement during both Delphi rounds, Kendall's W coefficient of concordance was calculated. Initial results showed moderate consensus (W = 0.218), which substantially improved to 0.507 in the second round, demonstrating significantly stronger agreement among participants. The statistically significant p-values (Asymp. Sig < 0.001) confirmed the reliability of these consensus metrics. Analysis of the second-round quantitative data revealed a clear prioritization among the selection criteria for accident analysis methods. The factor “considering safety barriers and controls” emerged as the most critical consideration, followed closely by “analysis level, aim, and scope.” Three additional criteria, including “causal factor identification,” “ease of use and implementation,” and “clear instructions and guidance provided by the method”, shared equal importance as tertiary considerations in the selection of accident analysis methods. These findings provide practitioners with an evidence-based framework for matching analysis techniques to specific accident scenarios and organizational contexts. By addressing previous selection ambiguities through validated, systematically derived criteria, the findings of this research offer a potential pathway toward greater methodological rigor in accident investigation. © 2026 The Authors. Published by Elsevier B.V. on behalf of Central South University. This is an open access article under the CC BY license. http://creativecommons.org/licenses/by/4.0/