영작문평가 상황에서 FACETS와 Coh-Metrix를 활용한 에세이 분석: 학습자 단면을 중심으로

초록

Lee, Ho. 2019. Analyzing L2 essay with FACETS and Coh-Metrix in L2 writing assessment. Korean Journal of English Language and Linguistics 19-4, 796-816. The current study aims to identify Coh-Metrix indices that significantly predict writing proficiency and well discriminate misfit learners from infit and overfit learners. 85 undergraduate and graduate students took an on-line essay test. Two expert raters scored their essays in terms of three analytic areas such as content, organization, and language use. The researcher conducted FACETS analysis whose purpose is to form sub-groups in regard to proficiency level and fit information. Then, the researcher ran Coh-Metrix program to investigate semantic, syntactic, and linguistic features of the students’ essays. The result showed that sentence length, argument overlap, and syntax similarity in adjacent sentences meaningfully discriminated proficient group from less proficient one. In addition, essays written by misfit learners showed lower value of Flesch Reading Ease than essays by counterparts. This study ended with the implication that further studies in L2 writing assessment need to explore learner dimension.

키워드