| Popis: |
This research makes a significant and original contribution to emerging debates within criminology and the social sciences more broadly, concerning the academic merit of using Bayesian statistics to analyse complex social problems, such as crime, with a view to promoting progressive and evidence-based policy reform agendas. It uses the risk assessment process in youth justice as a case study to demonstrate the utility of adding Bayesian approaches in the standard analytical tool box used to investigate the aetiology of offending behaviours, particularly when dealing with relatively small data-sets.The findings presented reinforce that it is possible using a Bayesian approach to ‘do more with less’ in terms of the number of cases analysed, and model the impact on the likelihood of further offending of individual characteristics, offending history, different types of offending and contact with the youth justice system. In considering the implications of its findings, the thesis considers how adopting a post-positivist stance - as called for by critics of the risk assessment process used within youth justice in England and Wales – enables new insights to be offered concerning the complex relationship between the framework of risk and protective factors and offending behaviours.It is concluded that they are distinct advantages associated with the adoption of novel statistical techniques within criminology, especially at a time where there is an increased emphasis on making greater use of administrative data to develop robust evidence-based policy. |
