Abstract
20 brain-damaged and 20 non-damaged retardates, matched for sex, IQ (57 to 83), and CA (20 to 51 yr.), were administered the Memory-for-Designs test. Two raters independently scored each protocol, with reliability between .93 and .97. Analysis of variance of pooled raw scores showed no significant variances for diagnostic category, sex, or their interaction. Diagnostic classifications by both raw and difference scores were nondiscriminating of brain-damaged from non-damaged Ss, reflecting 70% false positive rate for the non-damaged. Analysis of 3-point error scores showed significant variance for diagnostic category in the predicted direction (p < .05). Classification by 3-point errors had accuracy levels of 80% for male and female non-damaged Ss, 70% for female brain-damaged Ss, but only 50% for male brain-damaged Ss.
