Faculty Proceedings, Presentations, Speeches and Lectures


Prediction of Premorbid IQ with the Wechsler Test of Adult Reading Demographic Estimates

Event Location / Date(s)

Miami, FL / October 9-12, 2002

Document Type

Conference Proceeding

Presentation Date


Conference Name / Publication Title

22nd Annual Meeting of the National Academy of Neuropsychology



Estimates of premorbid intelligence are typically necessary to determine if a patient has sustained intellectual reduction as a consequence of traumatic brain injury. Estimates of premorbid IQ based upon patient demographic characteristics can be made by means of the Barona, Reynolds, and Chastain (1984) prediction equation (derived using the WAIS-R standardization sample) or from tables published with the Wechsler Test of Adult Reading (2001), derived using the WAIS-3 standardization sample. The accuracy of WTAR demographic predictions has not been examined. The current study compared the accuracy of Barona and WTAR predictions in a sample of 81 head injured patients and 68 psychiatric controls. Head trauma patients were an average of 10.3 months postinjuries characterized by a mean GCS of 9.5 (S.D. = 4.1), had abnormal CT scan findings in 86% of cases, and obtained a mean WAIS-3 FSIQ of 85.6 (S.D. = 14.1). Barona and WTAR demographic FSIQ predictions were significantly higher than obtained IQs (M = 98.0, S.D. = 7.6 and M = 97.0, S.D. = 6.7, respectively). Length of posttraumatic amnesia correlated with extent of intellectual decline using the WTAR (r = .29) and Barona (r = .25) estimates. Neurologically normal psychiatric patients with diagnoses of mood or anxiety disorders obtained a mean WAIS-3 FSIQ of 99.0 (S.D. = 13.4). Barona (M = 103.2, S.D. = 8.3) and WTAR (M = 100.6, S.D. = 7.7) predictions each correlated r = .59 with obtained FSIQ. Obtained and WTAR predicted IQs did not differ significantly. Barona predictions were significantly higher than obtained FSIQ. Extent of intellectual decline using WTAR or Barona estimates discriminated head trauma and psychiatric patients with 72.5 and 63.8% accuracy, respectively.

This document is currently not available here.