Predictive policing is about using large data bases (Big Data) to predict crime. Whereas traditional methods such as hot spot analyses and crime mapping only use historical data to identify high risk situations, these predictive models try to forecast future situations. As such the outcomes may support practitioners in allocating their resources to specific places, times, potential (groups of) offenders or victims.
The definition of Big Data is not unequivocal, but it is mostly characterized by three V’s: Volume (amount of data), Velocity (speed at which the data is added and processed) and Variety (the fact that data may come from multiple sources using different formats and structures). As is also reflected in this definition, current discussions mostly focus on technical features of Big Data as related to the accuracy of predictions. The present project however, deals with its psychological impact. Predictive policing is still in its infancy and evaluations indicate that it is unclear how predictions should be used in practice (Saunders, Hunt & Hollywood, 2016). One interesting question, for example, is how outcomes are interpreted by practitioners, as predictions are only based on correlations in the available data, possibly unrelated to theoretical insights (Chan & Moses, 2016).
Possible questions to be addresses are:
- Do practitioners accept outcome of models that are unrelated to criminological theories or that conflict with their gut feeling? How do they combine this information with their notions of criminological theory and their own gut feeling concerning high risk places or individuals?
- How do practitioners interpret and use probabilistic information? What does an increased risk of, for example, 10% mean? Are there specific biases related to using predictive models?
Chan, J., & Moses, L. B. (2016). Is Big Data challenging criminology? Theoretical criminology, 20(1), 21-39.
Saunders, J., Hunt, P., & Hollywood, J. S. (2016). Predictions put into practice: a quasi-experimental evaluation of Chicago’s predictive policing pilot. Journal of Experimental Criminology, 12(3), 347-371.
Immediately. Please contact firstname.lastname@example.org if you are interested in this master thesis topic.