Abstract
Although a large amount of research has been conducted in strategy use in second language acquisition and language testing fields, not much research has been conducted on detailed strategy use during an entire process of taking an academic speaking test. Given the importance of the TOEFL Internet-based Test (iBT), which includes both speaking-only and integrated speaking tasks, the need to study effective strategy use on the speaking section of the new TOEFL iBT is even more pronounced. This study aims to investigate how Taiwanese test-takers who received different scores in the simulated TOEFL iBT speaking test made use of their input-reception, preparation (pre-task planning), and speaking (within-task planning) times in different types of tasks. It also examines the construct validity of TOEFL iBT speaking test by means of strategies reported by the test-takers. Two main research questions are asked as followed: 1. For each task, do test takers who received high / mid / low scores use the approach, cognitive, communication, and metacognitive strategies differently during their input-reception time, preparation time and speaking time? 2. For each task, which individual strategies are significantly associated with the levels of performance during input-reception time, preparation time, and speaking time? Participants in this study were 39 Taiwanese adult English-as-a-Foreign-Language (EFL) learners who were preparing for the TOEFL iBT exam during 2008-2009. They took one simulated TOEFL iBT speaking test in a one-on-one audio-recorded setting and then had a retrospective interview recalling the strategies that they used to complete the six TOEFL iBT speaking tasks. The test-takers’ performance of each task was rated by two EFL teachers based on the TOEFL iBT speaking scales with the scores ranging from 1 to 4. Based on the ratings, the participants were divided into three proficiency groups in each task. Participants’ strategy uses were coded based on the taxonomies of O’Malley and Chamot (1990) and Swain et al. (2009). The results indicate that the test-takers used different types of strategies to cope with different types of speaking tasks. In the speaking-only tasks (referred to as the “independent” tasks by TOEFL iBT), approach strategies were used most frequently. On the other hand, in the listening-to-speak and the reading-to-listening-to-speak tasks (referred to as the “integrated” tasks by TOEFL iBT), the uses of cognitive strategies were reported more often. Some strategies are positively associated with the performance while some are negatively associated. Besides, some strategies seem to interrelate with each other, finally resulting in the difference in performance. We also found that the strategy use per se would not lead to distinctive performance; rather, it would be the interplay between the proficiency level and the strategy use that leads to the different results. Finally, the examination of strategy use within each task type helps build the argument for the construct validity of the TOEFL iBT speaking section. Overall, the results indicate that the TOEFL iBT speaking test successfully elicited the abilities it wanted to test. However, the authenticity of the speaking activities elicited by the TOEFL iBT is not high, as the test-takers approached the whole speaking section as a test rather than a real-life situations happening on campus. This lack of authenticity echoes the findings by Cohen and Upton (2006, 2007) in the TOEFL iBT reading section. Although this study is limited by its small sample size, it adds to the literature on strategy use in speaking tests, especially strategy use in different speaking task types and during different period of the test-taking process (input reception, preparation, and actual speaking). The study demonstrates the nature of each speaking task and their construct validity. In practical terms, because we find that strategy use may facilitate or hinder test performance, it is suggested that potential TOEFL iBT test takers should be aware of their strategy use while performing the TOEFL iBT speaking tasks in order to show their language ability to the full potential.