The State of Practice in System Safety Research Evaluation

This paper reports an investigation into the use of evaluation as part of system safety research. Using a simple classification scheme based on the knowledge and evaluation components of research papers, we classify two years of papers at the IET System Safety conference. Our analysis indicates a significant mismatch between a small collection of observational research papers with strong evaluation, and a large body of papers providing guidance which have not been evaluated. Of particular concern is that the majority of these papers do not provide sufficient information to support future evaluation. In response to these findings we suggest a minimum set of properties which guidance research products must have in order to allow the research to be evaluated.