Humans in the wild: A computational ethology approach to the study of social behaviour

Predicting and understanding the how, when, and why of behaviour is a perennial challenge for social psychologists. We argue that the scientific understanding of human behaviour can be accelerated by large-scale computational studies of observable behaviour in everyday contexts. Specifically, we propose a computational ethology approach, the systematic study of behaviour in real-world settings; recent advances in computational methods allow for such studies to be scaled up by using video and audio recordings to map specific behaviours to ecologically relevant stimuli. We illustrate this approach by outlining a recent study using this method and highlight advantages and challenges. In sum, we argue that the synthesis of ethology and computational methods offers unique potential for generating insights into the complexities of human social behaviour in the wild.