Semantic feature analysis: incorporating typicality treatment and mediating strategy training to promote generalization.

PURPOSE: This investigation was designed to examine the generalization effects of semantic treatment for word retrieval deficits in people with aphasia. Semantic feature analysis (SFA; Boyle & Coelho, 1995), typicality treatment (Kiran & Thompson, 2003), and mediating strategy training were combined to maximize potential generalization effects. METHOD: Treatment, which included SFA and a semantic feature judgment task, was conducted with 9 participants with chronic aphasia in the context of multiple baseline designs across behaviors. Typical and atypical exemplars were trained across animate and inanimate categories. Treatment was sequentially modified to overtly train the use of SFA as a mediating strategy. RESULTS: Eight of the 9 participants demonstrated improvements in naming of trained stimuli. Positive generalization effects were limited overall; possible response generalization was evident for 5 participants. Instruction in the use of a mediating strategy resulted in improved naming of treated words for all participants; however, generalization to untreated words did not occur. CONCLUSION: Treatment using SFA resulted in improved naming of treated typical and atypical exemplars in both animate and inanimate categories for 8 of 9 participants. Training in a mediating strategy also resulted in improved retrieval of experimental words. Regardless of intervention approach, generalization to untreated items was limited.