Huixin Gao: Audience Design as a Generative Mechanism in Proof Learning: Comparing Self-Explanation and Explanation to a Fictitious Other

Explanation tasks support this goal by encouraging students to articulate not just that a statement is true but why it is true, making explicit the chain of reasoning underpinning each inferential step. Existing research has largely focused on examining explanation strategies in isolation, leaving it unclear which approach best supports the development of high-quality explanations.

This study investigated how two explanation strategies—self-explanation (SE) and explanation to a fictitious low-prior-knowledge audience (EFO-weak)—affect proof-related explanatory performance, and whether the assumed prior knowledge of the audience (weak vs. high) modulates this performance. Undergraduate students submitted oral video explanations and anonymously assessed peers’ submissions using structured rubrics. Quantitative analyses revealed that EFO-weak encourages better explanations than SE, suggesting that framing explanations for an external audience promotes process-oriented reasoning and the careful articulation of warrants to a greater extent.