The preceding analysis suggests that reconviction rates and the frequency of reconviction among Drug Court cases were very similar to those among offenders given DTTOs before the Drug Courts were introduced and among offenders nationally sentenced under Summary proceedings to DTTOs. However it must also be acknowledged that the number of Drug Court cases and pre-Drug Court DTTOs on which this analysis was based was relatively low and that other factors whose effects cannot easily be quantified - such as the more efficient rolling up of cases in the Drug Courts - may have served to distort comparisons. Research published by the Home Office suggests that where relatively small differences in reconviction are anticipated - as would be the case when comparing Drug Court cases and DTTOs which are similar in many respects - relatively large sample sizes are required to demonstrate a significant reduction in recidivism (Harper and Chitty, 2005). On the basis that a 2% difference in reconviction rates between cases within the Drug Court and cases outwith the Drug Courts was anticipated and tested, a sample size of 6,000 cases in each sample would have been required to obtain statistical proof of this difference 9. Unfortunately, the number of cases seen in the Drug Courts was only 470 and in this respect, the Drug Court sample size falls far short of what would be required for a robust reconviction analysis and the findings must therefore be treated with appropriate caution, especially since international research has demonstrated with some consistency reduced levels of drug use and offending among offenders who participate in Drug Court programmes (McIvor, forthcoming).