Use of metacognitive strategies by English language learners during an online speaking test

Metacognition, or knowledge about and regulation of one's own thinking, has long been recognized as an important skill in learning and using a second/additional/foreign language ("second language" henceforth to align with most literature). However, there is a lack of empirical research specifically on the use of metacognition in supporting speaking a second language in computer-assisted language testing (CALT) situations. This study investigated the metacognitive behavior of test takers—specifically, the use of metacognitive strategies—as they completed two speaking tasks from the Canadian English Language Proficiency Index Program (CELPIP). One entails describing a picture and the other defending an opinion on a controversial topic. By studying the speakers' metacognitive strategic behaviors when completing the speaking tasks, as well as how these behaviors related to their speaking performance as evidenced by their CELPIP test scores, the study contributes to an understanding of how learners can be better supported in developing and using metacognitive skills on the CELPIP speaking test, and speaking English as a second language. Participants were thirty first-year undergraduate engineering students (aged 18-19) taking an English course at a university in China. In total, the study identified ten metacognitive strategies they used in completing the two CELPIP speaking tasks, three of the most frequently reported being "Evaluating language production" (reflecting on how they did post task), "Evaluating performance" (reflecting on how they were doing during task), and "Planning" (pre task). There was a statistically significant relationship between the total number of reported strategies and speaking scores in Task 1 (picture description) but not Task 2 (defending an opinion). Additionally, there were a few statistically significant relationships between individual strategies and speaking scores across the two tasks. I concluded that the impact of metacognitive strategies on CELPIP speaking performance is task dependent rather than uniform, with planning supporting speaking outcomes but overuse of substitution hindering them. These findings highlight the importance of strategic flexibility, task familiarity, and pre-task planning in CALT, and point to future research on strategy training across diverse speaking tasks.