Comparison of different versions of the Boston naming test

Abstract

Abstract This study compared a series of 49 clinical patients' performance on the different versions of the Boston Naming Test (BNT). Specifically, items were scored according to the criteria for the 85-item experimental version, for the 60-item published version, and for two 42-item short forms developed for test-retest purposes. Pearson correlations revealed high correlations among the different versions, and between BNT scores and other language measures such as WAIS-R Verbal IQ and the WAIS-R Vocabulary subtest score. Education correlated more highly with the BNT scores than did age. Comparisons of the 85- and 60-item versions revealed significant differences between the percentage correct, and that currently published norms for the 60-item version are more likely to classify a subject's score as impaired. Finally, the two short forms were found to be very similar.