Empirical comparison of alternate forms of the Boston Naming Test

Various short forms of the Boston Naming Test (BNT) are compared including an empirically derived 30-item form, odd and even items split-half forms, four 15-item forms, and a rationally derived 15-item form used in conjunction with the Consortium to Establish a Registry in Alzheimer's Disease (CERAD). The present analysis was conducted using a sample of 320 individuals with diagnoses including dementia (n = 194), thought disorder (n = 46), depression (n = 16), general neuropsychiatric disorders (n = 12), and cerebral tumors (n = 52). Results indicated that all forms possess adequate, although variable internal consistency, and correlations between forms were reasonable. Average item difficulty indices also differed with the CERAD version being least desirable. Finally, classification rates were different by forms, indicating limitations on the extent to which the forms may be used interchangeably.