口頭試験における試験形式および対話者の言語レベルと評価 Effects of Test Types and Interlocutors' Proficiency on Oral Performance Assessment
Access this Article
The study aimed to explore the impacts of three types of test and interlocutors' proficiency level on the test scores of Japanese learners of English participating in a single-speaker task, paired oral discussion, and group oral interaction. The assessment was carried out utilizing the criteria of the CEFR-J (Tono, 2013), the Japanese version of the Common European Framework for Reference (CEFR; Council of Europe, 2001). Five raters assessed 24 university students' oral performance. The ratings were then analyzed utilizing multi-faceted Rasch measurement (MFRM), and the author confirmed that the raters fit the model for the subsequent main analysis. In the main analysis, first, the MFRM indicated that the speakers scored higher in the group oral interaction compared to the single-speaker task and paired oral discussion. Second, the MFRM analysis showed significant differences across test types and proficiency levels. In addition, the raw score analysis revealed a certain disparity in terms of proficiency levels in that the middle-level speakers received higher scores in the paired and group oral interactions compared to the lower- and upper-level speakers. With regard to the paring and grouping conditions, it was difficult to identify any particular tendency, as various phenomena were seen. Third, the rating criteria showed that the scale was over-segmented for assessing the speakers and invited further investigation.
- ARELE: Annual Review of English Language Education in Japan
ARELE: Annual Review of English Language Education in Japan 26(0), 333-348, 2015
The Japan Society of English Language Education