ABSTRACT
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. This logic was applied to the detection of brain disorder using various cognitive ability composites. The verbal-performance difference has been widely used but often criticized, this chapter examines its potential efficacy to several other Wechsler subtest composites. It also explains that brain disorder was defined as a dichotomous variable; type of brain damage was not considered. Optimum cutting scores on each predictor variable for a given set of utilities were assessed using similar methods to those employed by McArdle and Hamagami. Using an iterative procedure, successive cutting scores were defined across the full range of data, and costs/benefits were applied as weights to the resulting decision outcome frequencies at each cutting score.
