An evaluation of criteria used by clinicians to infer pathology from figure drawings.

6 psychologists and 8 nonpsychologists were given a mixture of figure drawings made by 30 normal and 30 psychiatric adolescents and instructed to infer which drawings were by patients and to state the criteria used. All criteria were cross-validated on a new sample (N = 86). Valid criteria of pathology were "bizarre," "distorted," "incomplete," and "transparent." Valid criteria of normality were: "happy expression" and "nothing pathological." Many criteria used by clinicians led to inaccurate inferences. Nonpsychologists discriminated as well as psychologists (65% and 64% accuracy, respectively). An empirically based prediction formula discriminated with 79% accuracy on cross-validation. It is concluded that elimination of invalid criteria can lead to a sharpening of clinical judgment and more accuracy in evaluation of figure drawings. (59 ref.) (PsycINFO Database Record (c) 2010 APA, all rights reserved)