CAF기반 토익스피킹 대안모델 구현을 위한 기초연구

Exploratory research on developing an alternative TOEIC speaking model based on ‘Complexity’, ‘Accuracy’ and ‘Fluency’

The purpose of this study is to explore the concurrent validity between TOEIC SPEAKING and English speaking ability based on language ‘Complexity’, ‘Accuracy’ and ‘Fluency’, and to analyze the difference of CAF between upper and lower English speaking groups. In total 29 university students participated in the study and carried out the speaking test. The results of Pearson analysis revealed a high correlation (.796) between the TOEIC speaking scores and CAF. Specifically, ‘Complexity’ and ‘Fluency’ showed a high correlation and ‘Accuracy’ showed a moderate correlation with TOEIC speaking scores. There were also statistical differences between the upper and lower groups in correlation with CAF. ‘Complexity’ was the highest correlation in upper group whereas ‘Accuracy’ was the highest correlation in the lower group. To assess causal relationship of CAF on TOEIC speaking score, the study performed regression analysis between TOIEC speaking score and CAF. The results showed the English speaking ability based on CAF had 63.1% effect size on TOEIC speaking score. This suggests TOEIC speaking test can be replaced by CAF. Individual CAF regression analysis revealed Fluency’ had the strongest causal relationship (62.5%) followed by ‘Complexity’ (54.6%), and the ‘Accuracy’ had the least (10.4%). There was a statistical difference between upper and lower speaking ability groups: upper group had 34.3% whereas lower group had 16%, which means higher English speaking group can get the higher TOEIC speaking score.