Furthering the argument for visually inclusive L2 academic listening tests: The role of content-rich videos

Abstract

Although the listening skill is generally believed to incorporate visual processing, a second language (L2) academic listening assessment construct is still viewed as visual-free by many high-stakes test developers (Kang et al., 2019). This study attempted to generate evidence regarding the inclusion of visuals into the construct by comparing 143 English learners’ performance on an audio-only vs. a video-based version of an academic listening test. The video-based version of the test used content-rich videos, defined as lecture videos containing both context-related (e.g., gestures) and content-related (e.g., graphs) audio-congruent visual cues. In addition, the video-based version accounted for whether or not individual items in the test were cued by videos (coined as video-cued items). A multi-faceted Rasch interactions between learners’ listening proficiency (lower, higher), item type (local, global), and item video-dependence (video-cued vs. traditional) were run. The interactions revealed that video-cued global items were easier for lower-proficiency learners when videos were present. For the same learners, however, traditional global items were harder with videos. The effects of content-rich videos on local items and on higher-proficiency learners’ comprehension were negligible. These findings are discussed in the light of the argument for including content-rich videos in assessment constructs of L2 academic listening comprehension.