Developing a sensitive but generalizable measurement of vocabulary gains from self-selected extensive reading

Abstract

Abstract Self-selected reading promotes extensive reading and therefore facilitates learning. However, it is difficult to design vocabulary tests that can assess vocabulary growth in a reliable and ecologically valid way when individual students read different texts. To cope with the challenge, this study developed two vocabulary tests adapted from a research design by Horst (2005) and investigated the extent to which results from the two tests are similar to, or different from each other regarding learning gains. A generalized vocabulary test (GVT), developed based on multiple graded readers available to 83 Korean EFL university students, was administered as pre- and posttests. Sixty-two students who read extensively also took an individualized vocabulary test (IVT), designed using specific graded readers read by each student, as a second posttest. Results indicated that the GVT and IVT showed very similar patterns in terms of measuring students’ vocabulary gains from extensive reading, suggesting that a GVT can be developed as a reliable and useable assessment of vocabulary improvement without needing to identify specific words read by individual students (cf. Horst, 2005). Therefore, the GVT appears to be a valid and practical measurement of vocabulary growth through extensive reading in L2 contexts.