코퍼스를 이용한 성인 EFL 학생들의 영어 말하기 문법 오류 분석: ‘be ～ing’ 구문을 중심으로

Abstract

This study purports to investigate a grammatical error pattern that a group of college-level EFL students exhibited in an English oral proficiency test. To achieve this goal, 1,208 speech samples of 619 college-level EFL students were transcribed from text recorded with a computer-based speaking test system called K-COPI (Computerized Oral Proficiency Interview) developed at Korea Military Academy. Then, the researchers concentrated on the grammatical structure ‘be ～ing’ to investigate through focus group discussions among the speaking raters and English conversation instructors within the Korea Military Academy faculty. The transcribed text files were analyzed with the corpus analysis tool WordSmith using the Concordance and Keywords functions. This compare and contrast analysis displayed the frequencies of various erroneous patterns in which the ‘be’ verb was missing, in contrast to the correct patterns in which the ‘be’ verb was included. In terms of descriptive statistics, the structures containing a pronoun + ‘be ～ing’ showed generally lower error rates than the ones containing a modal verb + ‘be ～ing’ did. Then, a series of X2 tests were conducted using the students’ level of reading comprehension scores as the independent variable, and the oral patterns of erroneously omitting the ‘be’ verb as the dependent variable. Based on these X2 results, EFL students with a higher average score of reading comprehension are likely to omit the ‘be’ verb -- when saying the ‘I am/was ～ing’ and ‘you are/were ～ing’ structures -- at a significantly lower rate than students with a lower average reading comprehension score.