Constructing a Measurement Model of L2 Complexity in Automated Essay Scoring for Japanese EFL Learners’ Writing: Toward a Qualitative and Analytic Evaluation

ID: 54537 Type: Brief Paper
  1. Takeshi Kato, University of Tsukuba, Japan
  2. Yuichi Ono, University of Tsukba, Japan

Tuesday, March 19 4:15-4:35 PM

No presider for this session.

The current study aims to construct a measurement model of L2 complexity with conventional linguistic features, and examines the applicability of the complexity constructs to Japanese learners’ essay scoring by evaluating the prediction accuracy. Complexity index values are automatically analyzed with natural language processing techniques from 503 essays written by Japanese learners of English, and then, structural equation modeling is performed. As results show, the proposed model of measuring L2 complexity sufficiently fits the data and accounts for greater variance of holistic essay scores than traditional methods. Based on the findings, the researchers provide implications for L2 writing assessment in classroom teaching contexts and propose important suggestions for future research on the automated essay scoring based on educational constructs.

Topics

Conference attendees are able to comment on papers, view the full text and slides, and attend live presentations. If you are an attendee, please login to get full access.