Learning of subordinate category names by aphasic subjects: A comparison of deep and surface-level training methods

This study compared the effects of two cueing methods on aphasic subjects' learning and recall of unknown subordinate category names of dogs. One meth od, personalised cueing, required a deep level of stimulus processing. The second, phonological cueing, provided the subject with surface-level inform ation about the target word's phonemic characteristics. A total of 30 aphas ic subjects were assigned randomly to Personalised (PERS) or Phonological ( PHON) training conditions. Training was identical for the groups with the e xception that PERS group subjects created their own cues to aid recall of u nknown dog names (e.g., Kuvasz), whereas PHON group subjects were provided a first phoneme cue (e.g., /kuh/) and the number of syllables in the dog's name by the examiner. During training, the examiner presented the personali sed or phonological cue, and the subject named a coloured picture of the do g. Naming accuracy was measured across the 12 training trials. A dog name w as considered as learned if the subject responded correctly after a persona lised or phonological cue for the final four training trials. Recall of lea rned items was assessed with post-training probes 1 week, 1 month, and 6 mo nths after completion of training. Subjects were not provided with cues for the probes. Subjects in the PERS group had significantly higher levels of naming accuracy after cues and recalled significantly more of the learned d og names on the probes. Findings confirmed the authors' hypothesis regardin g the durability of personalised cueing in aiding subjects in learning unfa miliar names. Results of the study are discussed in terms of three attribut es of traditional aphasic naming treatments put forth by Howard and colleag ues (Howard et al., 1985): prompting, facilitation, and therapy.