Abstract
Arithmetic algorithms for predicting premorbid Full Scale IQ were evaluated in a sample of 125 brain injured adults who had been evaluated within two months of their injury, and then again one year later. FSIQ at post-test was used as the criterion variable. The prediction equation, based on demographic information only (Barona, Chastain, & Reynolds, 1984), proved to be a modest predictor of premorbid FSIQ. The BEST-3 (Vanderploeg & Schinka, 1995), a prediction equation that incorporates demographic information and performance data, was sensitive to the severity of the brain injury. Subsequent analyses found the BEST-3 to be a better predictor of recovery of function than it was an estimate of premorbid functioning. This study demonstrates the need for flexibility in the interpretation of results, as what was thought to be a prediction equation for premorbid functioning was better viewed as an estimate of recovery.