A comparison of indices of premorbid intelligence in clinical populations.

Estimation of premorbid intelligence using the regression formula developed by Barona et al. (1984) was compared with prorated IQ's based on the higher of the Vocabulary or Picture Completion subtests from the WAIS-R. Actual and predicted premorbid IQ's of 50 brain-damaged patients were compared to those of 35 non-brain-damaged patient controls equated on age and education. For both groups, the Barona et al. (1984) formula produced mean IQ's close to 100, though greatly restricted the range. The prorated IQ's tended to be over-estimated in the patient control group, but not the brain-damaged group. Compared with the regression based formula, the distribution of these prorated IQ's was more consistent with the WAIS-R standardization sample.