A Comparative Study of Cognitive Load and Test-Taking Strategy Use in Video-Based While-Listening Vs. Post-Listening Performance Tests
Abstract
While-listening performance (WLP) and post-listening performance (PLP) tests have been well established as two critical test formats for assessing listening comprehension competence under most second language (L2) assessment conditions. However, there remain a multitude of unknowns including the varying degrees of cognitive load that WLP and PLP tests impose on L2 learners’ working memory (WM), the different types of test-taking strategies elicited by WLP and PLP tests that potentially affect test validity, the reasons why L2 listeners consider one test condition to be more cognitively demanding over the other, and the extent to which the various degrees of cognitive load and strategy use in the WLP vs. PLP test conditions may interact with L2 learners’ listening performance. To address this gap, the present study aimed to explore L2 learners’ cognitive load, strategy use, and reasons for considering a specific test condition to be difficult in video-based WLP vs. PLP tests. Using a mixed-methods research design, this study triangulated three types of data: test performance data, questionnaire data, and interview data from 30 L2 speakers of English aged 19 to 30. The qualitative results indicated that test takers employed a wide range of strategies during the WLP and PLP tests, with the majority of them finding the PLP condition to be more challenging than the WLP condition. While cognitive load measure was not found to be a significant predictor of L2 listening proficiency, the quantitative results demonstrated the role of strategy-related factors in introducing some construct-irrelevant variance. This study implies the significance of integrating some effective strategies into L2 listening instruction to improve learners’ test performance and using more robust instruments such as eye trackers or functional near-infrared spectroscopy (fNIRS) tools in research on cognitive processes.