Purdue Pegboard as a predictor of the presence and laterality of cerebral lesions.

Abstract

A portion of the Purdue Pegboard Test taking 3 minutes to administer was given to 80 patients on a neurology service of a general hospital. With optimum cutoff scores it was possible to indicate the presence and laterality or diffuseness of brain damage in 70% of the cases. Brain damage without regard to lateralizatio n was correctly determined in 90% of cases with 7.5% false negatives and 2.5% false positives. The test was then cross-validated on 65 consecutive admissions with small loss in predictive efficiency. The data support the use of sensorimotor rather than perceptual cognitive tests in screening for brain damage because they are less dependent upon educational background and because of the lateralizing significance of sensorimotor dysfunction. Attempts of psychologists to diagnose the presence and laterality of cerebral lesions have been critized for failure to use tests which can be quantified reliably, to control for age and education, and to use adequate validating and cross-validating procedures (Meyer, 1961; Yates, 1954). The usefulness of psychodiagnosis of brain damage must also be considered in terms of cost in time and money, the ease with which diagnosis can be made without such psychological study, and the base rate incidence of neurological disorder in the population sampled. Costa and Vaughan (1962) and Vaughan and Costa (1962) in a survey of perceptual cognitive and sensorimotor performance in brain damaged patients found deficits in performance on all tests in brain damaged as compared to control groups. The perceptual cognitive tests, however, correlated highly with education. In a group of patients heterogeneously distributed with regard to education, the predictive efficiency of such tests would be lowered unless separate norms were set up for groups differing in amount of education. An approach which overcomes effects of educational level involves intraindividual comparisons on a large variety of tests. Such techniques reveal specific or general tests