Comparing Boston naming test short forms in a rehabilitation sample

Area of Science:

• Neuropsychology
• Rehabilitation Medicine
• Cognitive Assessment

Background:

• The Boston Naming Test (BNT) is a common tool for assessing confrontational naming.
• Existing BNT short forms have been primarily studied in dementia populations.
• The utility of BNT short forms in physical medicine and rehabilitation (PMR) settings requires further investigation.

Purpose of the Study:

• To compare the performance of several BNT short forms against the standard BNT administration (BNT-S) in PMR outpatients.
• To evaluate the reliability and accuracy of BNT short forms in identifying naming impairments within this patient group.

Main Methods:

• Five 15-item and three 30-item BNT short forms were derived from the BNT-S.
• Spearman correlations were used to compare BNT-S and short form scores.
• Cronbach's alpha assessed internal consistency for all forms.
• Norm-referenced scores (T < 36 and T < 40) determined BNT impairment.
• Area under the curve (AUC) values evaluated classification accuracy.

Main Results:

• BNT-S demonstrated strong correlations with most 30-item (rho = 0.92-0.93) and 15-item (rho = 0.80-0.87) short forms, with CERAD-15 showing a weaker correlation (rho = 0.69).
• Internal consistency was acceptable across all short forms (alpha = 0.72-0.86).
• BNT short forms exhibited excellent to outstanding accuracy in predicting BNT-S impairment.

Conclusions:

• BNT short forms are highly correlated with the standard BNT in PMR patients.
• These short forms demonstrate strong internal consistency and excellent classification accuracy for identifying naming impairments.
• BNT short forms are promising tools for use in rehabilitation settings and warrant further study.