Risk, threat or vulnerability? An ethogram to qualify concerning behaviour of citizens in police-civilian interactions via bodycam footage

All around the globe, police first responders are confronted with an increasing number of citizens showing concerning behaviour. These citizens are a harm to themselves or others, and to prevent further (physical) escalation from occurring, it is important that police officers quickly make sense of the situation and take follow-up actions (e.g. call for a healthcare or police intervention team). To do this, police should know what behaviour to look out for. Yet, previous research on these interactions predominantly used interview, case study and police database data which do not allow for direct observation of behaviour. In this article, bodycam-footage-based human ethology is considered as a method to overcome this issue. The dataset used for this study consisted of 14 police-citizen interactions in which citizens display concerning behaviour. From this dataset, an ethogram was created, identifying 21 types of concerning behaviours that citizens displayed. It was feasible to reliably code this behaviour on a