ABSTRACT
Cutting scores on any diagnostic instrument should be evaluated in terms of the relative frequencies and costs/benefits of resulting decision outcomes (Cronbach & Gleser, 1965; Meehl & Rosen, 1955). The decision validity (Paskus, 1993; Paskus & McArdle, 1996) of a diagnostic test can be defined as the discriminative accuracy of the cutting score employed, as evaluated within or across tests for a particular set of testing circumstances. Fig. 14.1 displays two hypothetical populations and the resulting decision outcomes after a cutting score is chosen on some indicator. The criterion-positive population represents those persons experiencing some disease, disorder, or condition of interest, and the criterion-negative population depicts everyone else tested. Typically, any chosen cutting score will result in some correct and incorrect diagnoses for each population. Finding an optimal cutting score involves mathematically reweighting the population distributions by the costs/benefits, locating the cut producing an acceptable distribution of decision outcomes, and comparing the result to the accuracy obtainable using other indicator variables. In the present study, this logic was applied to the detection of brain disorder using various cognitive ability composites. Criterion-based decision analysis representation. https://s3-euw1-ap-pe-df-pch-content-public-p.s3.eu-west-1.amazonaws.com/9781410603234/f4641f35-4cd2-47a4-9296-afddecb689e9/content/fig14_1_B.tif" xmlns:xlink="https://www.w3.org/1999/xlink"/>
