The Need for Systematic Replication and Tests of Validity in Simulation

Abstract

This chapter outlines how simulation methods might be used to make valid causal inferences in the social sciences, specifically the study of crime. We argue that significant threats to validity exist for simulation studies and that, if researchers do not actively take measures to minimize these, much of the promise of simulation will not come to pass. Further, we nominate replication as a general method to facilitate the generation of valid findings. It is anticipated, with the burgeoning interest in simulation methods in criminology, that simulation studies will be published in sufficient detail that allows researcher scrutiny and replication, with a view to developing a cumulative body of scientific knowledge.