Effects of random selection tests on second language vocabulary learning: a comparison with cumulative tests

Abstract Cumulative tests (CTs) require learners to recall previously learned words. The scope of the tested words gradually increases with increased study sessions. A potential limitation of CTs is that less learning may occur for words introduced later than earlier. This study proposes random selection tests (RSTs), which may compensate for this problem. We randomly selected the tested words from all target words, such that learners can review all words in a balanced manner. To investigate whether RSTs compensate for CT limitations, the study randomly assigned 69 Japanese learners of English to the RST or CT group. They were given 45 Japanese and English word pairs and weekly small tests across three weeks. The CT group recalled 15 words per small test, in which the scope of the tested words increased (15 → 30 → 45 words). Alternatively, the RST group recalled 15 words, which were randomly selected from 45 words, per small test. The posttest revealed that the RST group compensated for the limitation and retained the second and last 15 words better than did the CT group. The findings suggest that RSTs enable L2 learners to study all words in a balanced manner.