WISC-R IQ estimates from the Luria-Nebraska Neuropsychological Battery

Multiple regression equations predicting WISC-R IQs from Luria-Nebraska T scores and subjects' chronological age were developed and cross-validated on a total sample of 152 adolescent psychiatric patients. Data from a randomly drawn subsample of 100 subjects were entered into three stepwise multiple-regression analyses to derive equations predicting WISC-R IQs. These equations were applied to the remaining subsample (n = 52) as were the equations developed by McKay, et al. (1981) for estimating adult WAIS IQs. Mean differences, correlations, and classification hit rates between the two sets of predicted IQs and the actual WISC-R IQs were examined. Findings were similar to those of previous studies examining the predictive relationship between the Luria-Nebraska and adult intelligence. As expected, chronological age entered significantly into all three stepwise multiple-regression analyses. The newly derived equations had significantly higher correlations with the WISC-R for Verbal and Full Scale IQs and significantly better predictive accuracy for Full Scale IQ. Rates of IQ misclassification warrant caution in the use of these methods, but it is concluded that the equations derived here predict about as well as do short forms of the WISC-R.