Preliminary validation of the A1 and A2 sub-levels of the CEFR-J

Article appearing in Shiken 17.1 (May 2013) pp. 3-10.

Author: Judith Runnels
Hiroshima Bunko Women's University

Abstract:
The newly released Common European Framework of Reference Japan (CEFR-J) was designed to address the issue that a consistent system for measuring learner proficiency and progress in foreign language pedagogy in Japan is lacking. This tailored version of the Common Europe Framework of Reference (CEFR) was developed to better discriminate incremental differences in proficiency for Japanese learners of English, who tend to fall mostly within the A1 and A2 levels. Changes from the original CEFR included the creation of can-do illustrative descriptors that separated 4 of the existing 6 levels into sub-levels. The goal of the current analysis is to test the suitability of the new sub-levels of A1 and A2 for target users of the system in two ways: 1) by determining if newly developed descriptors are empirically rank or- dered by difficulty as specified by the CEFR-J, and 2) by testing the statistical significance of differences in difficulty ratings between the sub-levels. The current analysis found that the rank ordering of levels was the same as predicted by the CEFR-J, and that the higher-order A1 and A2 levels varied in difficulty to a statistically significant degree, but significant differences between adjacent CEFR-J sub-levels were not found. This raises questions about how users of the system can effectively distinguish features representative of each level and whether the additional sub-levels in the CEFR-J can function as intended. Limitations of using a system of illustrative descriptors based primarily on estimates of difficulty and the process of contextualizing a generalized framework are discussed.

Keywords: Common European Framework of Reference, CEFR-J, can-do statements, difficulty, contextualization

Download full article (PDF)