Modified semantic feature analysis for anomia: a single case study.

Semantic feature analysis (SFA) is a treatment approach designed for patients with lexical retrieval difficulty caused by semantic deficits. During training, a structured framework (e.g. requiring the patient to name the category, function, and colour of the target items) is usually provided to facilitate the patient in thinking of the semantic features of the target items. Nevertheless, the use of a structured framework potentially limits the variety of semantic features activated for lexical retrieval. This study investigated the effectiveness of modified SFA training to address this potential limitation. An odd-man-out task was carried out in the modified SFA training with an anomic patient with impaired access to the phonological output lexicon. The task aimed to encourage a detailed comparison of semantic features among items in the same category. It is hypothesized that activations of more distinctive semantic features belonging to the target items will result in greater extent of generalizations in the lexical processing. The patient's abilities in naming semantic features at the beginning and at the end of each treatment session were compared. Besides, accuracy and error patterns in oral and written naming in the initial and post-treatment assessment were also compared. The results showed a significant increase in the number of semantic features retrieved within treatment sessions. Moreover, a significant improvement in oral picture naming was also observed subsequent to the modified SFA training. Finally, the results of the error analyses further supported that the modified SFA was effective in promoting overall lexical processing. The findings indicated that both quantity and distinctiveness of concepts activated in the semantic system are essential to effective lexical retrieval. Theoretical and clinical implications were also discussed.