Allen C C, Ellinwood E H, Logue P E
Duke University Medical Center, Durham, NC 27710.
J Clin Psychol. 1993 Nov;49(6):874-82. doi: 10.1002/1097-4679(199311)49:6<874::aid-jclp2270490615>3.0.co;2-7.
The construct validity of a computer-assisted battery of neuropsychological tests (CNT) was explored with psychiatric inpatients and normal volunteers. A principal components analysis of inpatient scores revealed simple reaction time, response accuracy, visuomotor skill, and complex processing and memory components. A similar factorial structure was found in normal subjects. However, complex processing and memory measures emerged as separate vigilance and memory components in volunteers. CNT tasks were correlated with nine subtests of the Neurobehavioral Cognitive Status Examination (NCSE). Simple reaction time, and complex processing and memory measures discriminated impaired from nonimpaired inpatients as defined by the NCSE. Recommendations for research on CNT, and computer-assisted tests in general, are made.