영어발음 평가에서 한국인 영어교사의 평가결과 양적 연구

A Quantitative Study of Korean English Teachers’ Evaluations of English Pronunciations Spoken by Korean Speakers

In line with the study by 윤원희(2011a), a separate group of 35 Korean English teachers participated in pronunciation evaluations as raters to check their reliability as a group by calculating Cronbach’s alpha and to compare their corrected item-total correlation to their scores of listening, writing and speaking. As we have seen in the previous study, Korean English teachers are reliable raters but their proficiency of English is related with the degree of similarity to other raters’ evaluation in such a way that the higher the rater’s proficiency is, the closer to the others’ evaluation scores he or she is. The correlations between each rater’s corrected item-total correlation calculated with evaluations for segments, and his/her scores for listening, writing, and speaking are high enough to be statistically significant (p ≤ 0.001). This means that the proficiency level is a factor that could cause instability in the reliability of the evaluation, and that the pronunciation norm used as a criteria for pronunciation evaluation could be different depending on the rater’s proficiency of the language, which results in the different criteria used for the same speech.