Key points are not available for this paper at this time.
• 1/ Police reports provide a wealth of data on the causes of road traffic collisions. • 2/ Discord between report procedures and users’ perceptions can cause reporting errors. • 3/ Hierarchical clustering minimised report procedure and user perception discord. • 4/ New insights are revealed by police officers’ perceptions of crash causation. Worldwide, road crashes are a major course of death and serious injury. Police reports provide a rich source of data on the proximal causes (e.g., impairment by alcohol, failure to look properly) of road traffic collisions. Yet, road safety research has raised concerns about the quality and reliability of police reported data. In the UK crash report form, contributory factors are categorised (e.g., vehicle defects, driver error or reaction) to aid police officers in identifying appropriate factors. However, discord between the classification of contributory factors in crash reports and police officers’ own categorical perceptions may lead to misunderstanding, and in turn, misreporting of contributory factors. The current investigation recruited 162 police officers to report their perceptions of the relations among contributory factors in the UK crash report form. Hierarchical clustering analysis was used to identify an optimal category structure based on police officers’ perceptions. The clustering analysis identified a classification system with seven or eleven categories of contributory factors, maximising the internal coherence of categories and minimising discord with police officers’ perceptions. The findings also yield new insights into police officers’ perceptions of crash causation and demonstrate how statistical techniques can be used to inform the design of road traffic collision report forms.
Jonathan J. Rolison (Mon,) studied this question.