The Short NART: Cross-validation, relationship to IQ and some practical considerations

Abstract

Beardsall & Brayne (1990) have presented a method whereby the full-length National Adult Reading Test (NART) scores of subjects who are of below average reading ability can be predicted from performance on the first half of the test (termed the Short NART). The accuracy with which the Short NART predicted full-length NART scores was examined in a large cross-validation sample (N = 674). A subgroup of this sample (N = 142) was administered the WAIS. The results indicated that the Short NART was only moderately successful in predicting full-length NART scores. However, comparison of the accuracy with which the full-length NART and Short NART predicted WAIS IQs, revealed that the superiority of the former was very minimal. It is concluded that, despite some reservations regarding the Short NART's practical utility, it can be used with reasonable confidence in clinical practice to estimate premorbid IQ.