Study title

The project consists of two main parts: Part 1 is a survey of methodological assumptions for the investigation of accidents that provides valid knowledge about causal relationships. It will be focused on the importance of using a structured and standardized method of analysis and aggregation of results from multiple events . The study is based on the analysis method "Cognitive Reliability and Error Analysis Method" CREAM ( Hollnagel , 1998). This is a generic method for causality analysis of critical incidents in systems that involve interaction between operators, equipment and organization. The method will be tested through case studies of railroad incidents and road traffic accidents. The project builds on a previous project where CREAM was tested on pre-collected data, and where no collections were made of supplementary data collection in connection with the analyzes. One conclusion was that the data in many cases was inadequate and that the method should therefore be integrated into the investigation work so that the method helps to influence which data is recorded about the events. The analyzes of railroad incidents will be conducted in collaboration with the NSB and analysis of road traffic accidents in cooperation with the Norwegian Public Roads accident analysis groups. Implications for other modes of transport will be discussed. Project Part 2 is an examination of how knowledge is created and accumulated through the government's Accident Investigation Commission is understood and used among relevant actors with security responsibilities . The goal is to find to what extent and in what form the Commission's reports have consequences and is in use , but also the factors that may explain the extent and type of application . The study will be based on qualitative interviews with key stakeholders inside and outside the Commission. Data collection in Part 2 is coordinated with the project " Accident investigation and learning effects within transport modes...