A Selection Artifact in Delinquency Data Revisited

Studies of correctional programs have observed that delinquents exhibit a sharp rise in their arrest rates up to the time of intervention, and that they continue their activity after their release from the programs, but at a lower rate. The nature of this phenomenon has precipitated controversy; one school of thought attributes it to the effectiveness of the correctional programs, another claims that it may be due largely to a selection artifact. An analysis proposed in a recent article by Maltz and Pollock supported the artifact hypothesis. In this paper, we point out an error in Maltz and Pollock's analysis, and show analytically and by simulation that the correction of this error reduces the size of the selection artifact. However, it is still plausible to the attribute the observed phenomenon to a selection artifact.