Teachers judged LLM-simulated student ideas realistic and most useful when the ideas sparked sensemaking about student thinking and suggested scaffolds¶

Claim · All claims
Evidence · 1 study (2 entries) · design r2 · q2
Currency · newest study 2026 · checked against its sources 2026-10-18 (automated, 2 of 2 entries) · not reviewed by a person

Subclaims¶

q2 i? All six interviewed teachers found the simulated responses overall realistic and aligned with their lessons' objectives. → Nguyen 2026 q2 i? Teachers valued the LLM-generated facilitation prompts for scaffolding student ideas and surfacing unexpected insights, and three noted they could save time preparing new lessons. → Nguyen 2026 (2)

Evidence¶

Nguyen 2026¶

Nguyen, H., & Cao, J. (2026). Exploring the Capacity of Large Language Models to Simulate Students' Scientific Thinking: Insights for Responsive Teaching. Journal of Science Education and Technology. https://doi.org/10.1007/s10956-026-10333-5

q2 · i? · design · r2

Thematic analysis of 45-minute semi-structured interviews with six teachers who evaluated GPT-4o-generated ideas against their own lesson plans. Example: Rylie reasoned about an incorrect roller-coaster energy idea, "This is incorrect and is important for me to think about."

"Teachers stated that the responses were overall realistic and aligned with their lessons' objectives. They mentioned that the responses were most useful when they sparked teachers' sensemaking about what students knew and did not know."

Nguyen 2026 (2)¶

Nguyen, H., & Cao, J. (2026). Exploring the Capacity of Large Language Models to Simulate Students' Scientific Thinking: Insights for Responsive Teaching. Journal of Science Education and Technology. https://doi.org/10.1007/s10956-026-10333-5

q2 · i? · design · r2

Teachers appreciated suggestions like asking for evidence, reframing questions, and providing counterexamples; John noted unexpected penguin examples in a bird-movement modeling lesson. Three teachers said the suggestions could save time and help anticipate student responses.

"Additionally, all teachers found value in the LLM-generated instructional prompts to scaffold student ideas and surface unexpected insights."

Discussion¶

Related Claims¶

• LLM-simulated student ideas moderately overlap with teacher- and student-generated lesson ideas (mean cosine similarity 0.52), with larger LLMs slightly outperforming Mistral-7B — related
• 91% of LLM-simulated ideas fell within grade-level knowledge scope, but within-scope proportions were significantly lower for elementary and middle school than high school — related
• Only 69% of LLM-simulated ideas met the target grade's language level, with elementary grades faring worst and Llama-4 performing best — related
• Claude showed the highest alignment with human BREQ responses, and interview-containing prompts aligned better than baseline prompts — related
• LLM-based simulated students surpass rule-based systems in natural language generation and situational flexibility, but concerns about bias and evaluation reliability persist — related
• Dissociation between scaffolding quality and retrospective reasoning: experienced teachers scaffold better but reason less advanced about student thinking — related