Criterion validation of premorbid intelligence estimation in persons with traumatic brain injury: 'Hold/don't hold' versus 'best performance' procedures

The goal of the present study was to validate previously suggested regression equations for the estimation of premorbid ability against a real premorbid intellectual criterion. Fifty-four patients with traumatic brain injuries, for whom a premorbid military Primary Psychometric Rating (PPR) was available, participated in the study. Two prediction procedures were validated: (a) 'BEST-10', which generates a predicted score from the highest observed score of 10 WAIS-R subtests, according to the 'best performance' estimation principle. (b) 'BEST-2', which generates the predicted score from the higher of two subtests considered a priori resistant to neurological damage according to the 'hold/don't hold' principle. The two procedures showed similar correlation with the premorbid criterion. However, the BEST-10 method provided a more accurate estimation, generating a non-significant 2- point underestimation. The results support the application of previously proposed equations for estimating premorbid intelligence, and suggest that the use of the best performance principle is preferable as compared to the hold/don't hold principle.