Clinical versus Actuarial Prediction

摘要: Many statistical models have been developed for making predictions, according to some criterion of optimal fit (e.g., minimizing mean square error). Paul Meehl's influential 1954 book surveyed about 20 studies comparing such predictions—of important human outcomes parole violation)—to predictions clinical experts who had access exactly the same information on which prediction was based. In no case expert superior. Later work examined where base model and judge were not equivalent, or involved models—such as unit weighted ones—not based optimality principles. business medical contexts, judges sometimes make superior but only when they included in models, incorporate models' their own. psychology, are even a subset available clinicians (who might, example, also interview people). Moreover, need be may ‘improper’—as linear combination is ad hoc (but directionally correct) intuitive weights. By year 2000, roughly 150 supported these conclusions.