Abstract
Experiments have an important part to play in developing evidence-based policy and practice in criminal justice. They need, however, to be construed in realist terms if valid conclusions are to be drawn. The precise design of an experiment needs also to be tailored to the level of program theory development. In only very particular circumstances will random controlled trials (RCTs) be appropriate. An illustration of successful use of non-RCT realist experimental evaluations in repeat victimization research and policy development in Britain is given. The achievement of this fruitful relationship between research and policy development itself depended on a (perhaps unusually) favorable context.
Get full access to this article
View all access options for this article.
