Abstracts

Auditory naming tasks may be more appropriate than visual ones as a measure of potential word-finding difficulties in patients who are candidates for resection from the dominant anterior temporal lobe (1). We wished to optimize clinical usefulness of the existing auditory naming task (ANT) by creating a French version that would yield results similar to those obtained in English, and by increasing the number of items so that auditory naming results could be compared more directly with those from visual naming. To facilitate a more direct comparison of auditory naming with the visual naming task used most commonly (Boston Naming Test; BNT), we increased the number of items on the existing auditory naming task from 50 to 69 by adding 19 new items, then testing 119 healthy volunteers (53 francophone, 66 anglophone) on this longer version, and finally keeping the 60 items that gave the clearest results (i.e., elicited greatest agreement, fewest synonyms). To maximize equivalence between English and French versions, we compared results from anglophone and francophone subjects for each item, rejecting those on which language-group differences were greater than 25%. These normative data were then used to rearrange items in order of increasing difficulty separately in each language. We are now comparing performance on the revised ANT to BNT in patients (N = 20) with epileptic focus in a temporal lobe. Comparison of mean correct responses on the 60-item ANT with those from the BNT showed a small, in absolute terms, but significant (p [lt] .01) difference between tasks for healthy subjects, with BNT yielding the higher scores. There was a main effect of language, showing higher English than French scores on both tasks (p [lt] .01). Patients with a left temporal-lobe focus earned lower scores on both naming tasks compared to those with a right temporal-lobe focus (p = .03); again ANT scores were lower overall than BNT scores. As reported also for the 50-item ANT (1), auditory naming remains more difficult than visual naming for healthy subjects. Thus, despite equating the ANT and BNT for number of items, scores cannot be compared directly, as the difference in difficulty should be taken into account. Our French version of the ANT yields scores that are highly similar to the English, although the French are slightly lower. The French ANT is already proving useful in language evaluation of francophone patients in Quebec.
1. Hamberger M, Seidel W. JINS (2003);9:479-89. (Supported by Canadian Institutes of Health Research.)