Development and validation of a Japanese C-test

A practical proficiency measure, that is carefully designed and accessible, is essential for research that can advance our understandings of L2 acquisition and related phenomena. Despite their potential to fulfill this apparent need, there have been only a few examples of Japanese C-test development (e.g., Noma, 1985, 1991; Roos, 1994, 1996). These pioneering Japanese C-tests were found to distinguish among participants with distinct language proficiency levels to a certain extent. However, their text selection and deletion methods were not as rigorous as typical of C-tests in general. Accordingly, the current project sought to develop a Japanese C-test by carefully considering critical processes, especially text selection and deletion methods. In this study, text selection was guided by a can-do list created to differentiate learners' abilities to process Japanese written materials. 15 texts were selected initially, then reduced to 11 as a result of multi-phased pilot testing with Japanese native speakers. Subsequently, three deletion methods were piloted with native speakers in order to choose one optimal deletion method for use with L2 learners of Japanese. Finally, 40 L2 learners of Japanese in the US, representing four university curricular levels, completed the 11-text C-test. The optimal five texts to be included in the final Japanese C-test were selected based on analysis of the participants' responses. The final five-text C-test was shown to have good psychometric qualities and to reliably distinguish among L2 learners with different proficiency levels. This chapter discusses challenges in developing a Japanese C-test and provides practical suggestions for developing effective C-tests in Japanese and other languages.