Tim Mullett
Assistant Professor, Behavioural Science
University of Warwick
Using health and policing data to predict impulsive and harmful behaviours
The advent of digital databases and the ever growing accessibility of big data means that it is increasingly possible to take behavioural science out of the laboratory and examine effects in the real world. This research uses policing and healthcare data from England to examine patterns in harmful behaviours. For example, that incident rates are higher amongst individuals who live in neighbourhoods with easier access to shops and amenities, demonstrating the role of impulsivity in a real-world setting. We also examine specific issues, such as the oft held belief within UK police forces that there is a spike in alcohol related violence and other incidents around common pay days of monthly salaried employees. We find no evidence in either police incidents, or hospital admissions that such a payday effect exists. However, we do find a substantial Christmas effect in police incidents and it appears that previous police analysis failed to properly control for this, resulting in the apparent finding of a payday effect. We also examine rates of domestic violence around England football (soccer) team matches. We find a significant increase in rates of domestic violence beginning during matches, and then decaying over the following hours. However, contrary to previous studies using American football (NFL) matches, we find that this effect is largest following England wins. Our detailed data allows us to show that the increase in domestic violence is mediated by alcohol consumption, which is greater following victories.
The presentation will begin at 12:00pm. Food and drinks will be provided.