Application of WebFit for Diagnosis of Lexical Disorders in Korean Aphasia

Ha, Ji-Wan; Sim, Hyun Sub

ORIGINAL ARTICLE

Commun Sci Disord. 2011;16(3): 231-247.

Ji-Wan Ha^` , and Hyun Sub Sim^`

하지완(Ji-Wan Ha)| 심현섭(Hyun Sub Sim)

ABSTRACT

Background & Objectives
The purpose of this study is to examine the feasibility of applying WebFit, a connectionist SP-model program, to the Korean population. This study also aims to determine the appropriateness of classifying Korean aphasiacs into two groups (S-L and L-P) according to lexical retrieval failure pattern using WebFit.
Methods
The SP-model on which the WebFit program is based is aimed at English speakers. Therefore, when applying the model to a language other than English, the error distribution rates should be similar to those for English. Thus, this study reviewed the distribution rates of errors occurring in K-BNT to evaluate whether they are similar to the error distribution rates in English (Experiment 1). Twenty Korean aphasics were classified into the S-L or L-P group using WebFit. The phonemic paraphasia rates, semantic paraphasia rates, and repetition scores of the two groups were compared based on studies by Goodglass (1993) and Hadar et al. (1998) in order to verify the appropriateness of the group classification (Experiment 2).
Results
The word rate and the nonword rate in K-BNT were 18.2% and 81.7%, respectively. Among the error rates in the words, the semantic error rate was 0.7%, while the formal, mixed, and unrelated error rates were 9.8%, 0.5%, and 9%, respectively. These results are very similar to the error distribution rates suggested by Dell et al. (1997). Through the application of WebFit, the 20 subjects were classified ten into the S-L group and ten into the L-P group. The semantic paraphasia rate of the S-L group was significantly higher than that of the L-P group (p ＜ 0.05), while the phonemic paraphasia rate was significantly higher in the L-P group (p ＜ 0.05). Moreover, the repetition score of the S-L group was significantly higher than that of the L-P group (p ＜ 0.01). Discussion & Conclusions: The connectionist program WebFit can be applied to Korean subjects as the error distribution rates in K-BNT are similar to those in English on which the SP model is based. This study also verified that Korean aphasiacs can be classified into the S-L group and the L-P group using WebFit. Moreover, such classification seems appropriate considering that the two groups revealed significant differences in phonemic paraphasia, semantic paraphasia, and repetition scores. Points of consideration and limitations of the application of WebFit to Korean aphasiacs are suggested in the discussion section.