Learning New Words from TED Talks: Strategic use of L1 subtitles and L2 captions
Abstract
In recent years, studies examining the effectiveness of audiovisual input and on-screen text for incidental vocabulary learning have proliferated. However, no studies have explored the potential of repeated viewing with an alternation of L1 subtitles and L2 captions for incidental vocabulary learning although both types of on-screen text have been proved to be beneficial for vocabulary acquisition. Given this gap in the literature, we designed the present study, the rationale for which was guided by the notion of desirable difficulty, the role of retrieval from memory, and conflicting findings regarding the benefits of trial-and-error learning. The research questions were whether using an alternation of L1 subtitles, L2 captions, and no onscreen text (henceforth “none”) leads to greater vocabulary learning compared to using only L2 captions repeatedly and whether the sequence of the different kinds of onscreen text makes a difference to learning gains in the case of repeated viewing. The participants (N = 30) were upper intermediate to advanced ESL learners. They were randomly assigned to one of three conditions, which were watching a TED talk video three times with the sequence of 1) none-subtitles-captions (n = 10), 2) subtitles-captions-none (n = 11), and 3) captions-captions-captions (n = 9). Eleven target words were selected from the video. A meaning recall test format was adopted for a pre-test, an immediate post-test, and a delayed post-test. The tests were administered at 1-week intervals. A listening comprehension test was administered after the first viewing to ensure the participants attended to the content of the TED talk video and vocabulary learning could be ascribed to incidental learning. A meaning recognition test was administered as part of delayed post-testing as well. Finally, a questionnaire elicited the participants’ perceptions of the usefulness of the different viewing sequences. The output of mixed-effects logistic regression analysis revealed that incidental vocabulary acquisition definitely happened through the repeated viewing, but no significant difference was found in the effectiveness of the three viewing conditions. That significance was not reached is unsurprising considering the small study sample. However, the descriptive statistics and the questionnaire responses suggested that using a sequence of subtitles, captions, and none may facilitate word learning at the meaning recall level compared to using captions only. The results thus call for more research on the merits of this sequence of viewing a video with decreasing support from onscreen text.