This is the fourth brief in the six-part series titled A Question of Evidence, Part Two. In this brief, Chris Baird explores the research behind many current models, discusses methods commonly used to measure “predictive power,” and outlines what is required to measure the efficacy of various approaches to risk assessment.