Impact of scaffold structures on student-generated questions in English grammar learning

Language learners' output plays a crucial role in Second Language Acquisition, fostering hypothesis testing and metalinguistic awareness. However, novice learners, particularly in English as a Foreign Language (EFL) context, often face challenges in producing assessable language production. While student-generated questions (SGQ) offer a promising pedagogy, research on the comparative effects of different scaffolding types for SGQ in English grammar learning remains notably underexplored. This study investigated whether A2-level EFL learners benefit differentially from two distinct scaffold types, the less structured ‘main idea’ and the more structured ‘what-if-not’, on their grammar learning through SGQ. Adopting a quantitative, within-subjects quasi-experimental design, a convenience sample of 55 first-year non-English majors at a university in southern Taiwan participated. Data were collected via an online learning system (QuARKS) that delivered the scaffolded SGQ tasks, and student output was assessed using a multi-dimensional rubric measuring fluency, flexibility, and creativity (novelty and usefulness). Results from paired-sample t -tests indicated that the ‘main idea’ scaffold led to significantly higher fluency and novelty. Conversely, the ‘what-if-not’ scaffold resulted in significantly greater usefulness. No significant difference was observed for flexibility. The findings highlight the practical value of SGQ as a dual instructional and formative assessment tool in grammar learning, supported by technology. The multi-dimensional assessment approach provides nuanced insights into learner productive performance. This study suggests larger-scale, longer-term, and more diverse studies to evaluate sustained effects and generalizability across language proficiency levels.