Search papers, labs, and topics across Lattice.
The paper introduces SciEval, a new benchmark dataset for Automatic Instructional Materials Evaluation (AIME) in K-12 science education, comprising 273 lesson-level materials evaluated across 13 criteria using the EQuIP rubric. They tested several mainstream LLMs (GPT, Gemini, Llama, Qwen) on SciEval, finding subpar performance, and then demonstrated that fine-tuning Qwen3 on SciEval yields up to 11% performance gains, emphasizing the importance of domain-specific adaptation. This work frames AIME as a generative AI task, predicting scores and evidence based on educator-designed rubrics.
Off-the-shelf LLMs are surprisingly bad at evaluating K-12 science instructional materials, but targeted fine-tuning can significantly improve their performance.
The need to evaluate instructional materials for K-12 science education has become increasingly important, as more educators use generative AI to create instructional materials. However, the review of instructional materials is time-consuming, expertise-intensive, and difficult to scale, motivating interest in automated evaluation approaches. While large language models (LLMs) have shown strong performance on general evaluation tasks, their performance and reliability on instructional materials remain unclear. To address this gap, we formulate Automatic Instructional Materials Evaluation (AIME) as a generative AI task that predicts scores and evidence using the rubric designed by the educator. We create a benchmark dataset and develop baseline models for AIME. First, we curate the first AIME dataset, SciEval, consisting of instructional materials annotated with pedagogy-aligned evaluation scores and evidence-based rationales. Expert annotations achieve high inter-rater reliability, resulting in a dataset of 273 lesson-level instructional materials evaluated across 13 criteria (N=3549) using the EQuIP rubric. Second, we test mainstream LLMs (GPT, Gemini, Llama, and Qwen) on SciEval and find that none achieve strong performance. Then we fine-tune Qwen3 on SciEval. Results on a held-out test set show that domain-aligned fine-tuning can achieve up to 11 percent performance gains, highlighting the importance of domain-specific fine-tuning for AIME and facilitating the use of LLMs in other educational tasks.
