ĐÁNH GIÁ KHẢ NĂNG NGHE HIỂU CÁC THÔNG TIN TƯỜNG MINH VÀ KHẢ NĂNG SUY LUẬN TRONG KHI NGHE: CÁC CÂU HỎI TRẮC NGHIỆM CÓ THẬT SỰ HIỆU QUẢ?

Abstract

Multiple-choice questions (MCQs) are a widely accepted format used in second language (L2) listening tests. However, there is a lack of literature on the usefulness of this format in measuring the listening construct. This study investigates the extent to which MCQs engage two important components of the default listening construct: understanding explicit information and drawing inferences from the spoken input (Buck, 2001). Verbal reports were employed to explore the participants’ cognitive processes as they answered multiple-choice test items. Taking a qualitative approach, this study draws on the think-aloud protocols provided by 10 English majors from a university in Vietnam. The results show that while the multiple-choice format successfully measured the ability to understand explicit information, it promoted the use of test-taking strategies that compromised cognitive validity of the test items that targeted the ability to draw inferences. Practical suggestions were given so that MCQs can be better prepared to engage the inference-making process. These included input text preparation, item design, and double-play format. The challenges in measuring L2 learners’ ability to make inferences while listening were also discussed.