Autosuggestion Quiz

About

Generative AI tools, like GitHub Copilot, are becoming an industry standard by offering real-time code suggestions that streamline the coding process. Although these systems improve productivity, they also introduce pedagogical challenges. Students may become overly reliant on AI-generated code suggestions, accepting them without critical thought, potentially reducing their ability to engage with the underlying logic of the code. We developed an interactive quiz system within a simulated IDE environment designed to help students think critically about autogenerated code suggestions. Instructors use the tool to create contextualized coding quizzes that present multiple code suggestions at each line. Students must choose the correct option to move on to the next step. Survey responses suggest that this approach could promote critical thinking and scaffold metacognitive skills like planning and reflection. Students reported that the system helped them distinguish between good and bad suggestions. Most students preferred this experience to traditional quizzes or GitHub Copilot. These findings show the potential to scaffold more critical use of generative AI coding tools.