We had a great meeting on Tuesday with three interesting and very different presentations.
Pam Murphy told us about her role as a member of the Independent Prisons Monitoring Board at HMP Wandsworth. I wasn’t aware of this organisation which was set up by statute to prevent the abuse of prisoners in custody. They provide a national monitoring framework and publish annual reports based on the work of their people based in prisons.
Aidan Mews and Mark Purver from the MoJ presented a summary of the work they (and Laura Di Bella) had done to evaluate the impact of the Sex Offender Treatment Programme that is run in prisons. The programme was originally accredited in 1992 and, although some studies had been done on its effectiveness, some of the findings were ambiguous or contradictory. This study is the most robust that has been carried out and used a technique called Propensity Score Matching to come up with a treatment sample of prisoners and a control group. This was the best option given the inability to run a Randomised Control Trial. Results suggest that while the Core SOTP is generally associated with little or no changes in reoffending, there were some statistically significant differences. The findings were published last year and were, unsurprisingly, picked up by the National press.
Our final speaker was Professor Stewart Robinson from Loughborough University who ran a tutorial on the Verification and Validation of Models. He explained that verification tests whether the conceptual model design has been satisfactorily converted into a computer model. Validation tests the extent to which the model is sufficiently accurate for its stated purpose. He described a range of ways in which both of these could be conducted and emphasised that V&V needs to be done continuously and iteratively throughout the modelling process.
The slides are available on the ORS website.
Our next meeting is on 5th June 2018. Contact firstname.lastname@example.org for details.