Masao AIKAWA
   Department   Kyoto University of Foreign Studies  Department of British and American Studies, Faculty of Foreign Studies
   Position   Professor
Date 2016/09/01
Presentation Theme The Development of the Grammar/Text/Error Profiles for the CEFR-J: ​Application of the Profile Information
Conference JACET 55th International Convention
Promoters 大学英語教育学会
Conference Type International
Presentation Type Panelist at Symposium/Workshop (Applied)
Contribution Type Collaborative
Venue 北星学園大学
Publisher and common publisher Tono Yukio,Ishii Yasutake, Uchida Satoru, Notohara Yoshiyuki, Aikawa,Masao
Details There is a growing interest in applying the CEFR in various areas of English language teaching in Japan. The CEFR-J is one of such examples. Our project team aims to create resources for using the CEFR-J more effectively. In this symposium we will report on the latest development of three outcomes of CEFR-J reference level descriptions (RLDs): the CEFR-J Grammar Profile, the CEFR-J Text Profile, and the CEFR-J Error Profile. First, Yukio Tono will briefly report on the CEFR-J project as a whole and how RLDs have been conducted using various corpus resources. Then Yasutake Ishii will report on the Grammar Profile by explaining how grammatical constructions were selected and extracted from ELT coursebook corpora and how grading and sequencing of grammar items were determined based on the CEFR levels, using machine learning. Then, Satoru Uchida will go on to report on the Text Profile, where various metrics for text fluency, vocabulary profile, and text complexity were extracted from the texts classified according to the CEFR-J levels and the frequencies of those different metrics were processed in terms of their relative values as criterial features. Yoshiyuki Notohara will then discuss the process of analyzing Japanese EFL learners' errors as criterial features for CEFR-J levels and the development of the CEFR-J Error Profile. Finally, Masao Aikawa will present how teachers, policy makers, materials designers, and test developers can exploit these profiles for improving their existing curriculums, teaching materials, tasks or tests.