Abstract

This study compares, within the context of parole decision-making, the predictive utility of five statistical methods commonly employed to develop correctional risk-screening devices: (a) two general linear additive models, (b) two configural models, and (c) a model based on a multivariate contingency approach. Devices felt to be operationally useful (at least potentially) were developed and cross-validated using large samples of federal releases. Results suggest that no apparent empirical advantage accrues given use of different models. Even when large amounts of error are randomly added to the predictor item-pool, substantive conclusions do not change. Further, most devices developed are highly intercorrelated. Implications for the practical development of screening instruments are discussed, and further research is suggested.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call