Alternate forms of the Boston Naming Test in Alzheimer's disease

A 15-item version of the Boston Naming Test (BNT) is used as a measure of language ability in the Con sortium to Establish a Registry for Alzheimer' s disease (CERAD) battery. The purpose of this study was to examine the rate of agreement between the 15-item BNT and the full 60-item BNT in the classification of language impairment in 58 patients meeting NINCDS/ADRDA criteria for probable Alzheimer' s disease (AD). All patients were administered both versions of the BNT and were classified as normal, borderline, or impaired. The null hypothesis of independence of classification was examined using a chi-square analysis and Pearson's correlation. Although the two versions of the BNT were strongly associated (r = .83), the rate of agreement between the 15-item and 60-item BNT was only 70% in classifying patients as impaired on language abilities. These findings suggest poor criterion validity of the 15-item version of the BNT used on the CERAD battery compared to the full 60-item BNT in correctly classifying language abilities in patients with AD.