Leveraging Context for Computer-Supported Student-generated Questions and EFL Learning in Grammar Instruction: Its Effects on Task Performance

Inspired by the concept of 'learning in context, ' this study aimed toward examining the effect of computer-supported student-generated questions (SQG) with a given reading passage (acting as the context) on the quality of generated questions and task completion for grammar learning. Thirty-eight college sophomores with Intermediate English level taking a two-credit English course at a university in southern Taiwan participated in two back-to-back SQG activities on grammar covered in instruction during regular class. A newly developed online system equipped with computer-supported procedural prompts and design was adopted to support SQG as individual knowledge construction activities. Four grading criteria - Importance, Fluency, Originality, and Elaboration, were devised for the assessment of the quality of student-generated questions. The results of the repeated measures analysis of variance showed significant differences in task performance total scores between the two SQG activities, attesting SQG without a given passage exceling SQG with a given passage. Additionally, SQG without a given reading passage led to significantly better performance in Fluency, Originality and Elaboration indices than SQG with a given passage. Finally, the result of the chi-square test showed a near-significant difference (p = .054) between the two SQG activities in task completion, with comparatively more participants completing the task under the SQG without given context condition than the SQG with context condition.