The SCALE AI Repository brings together more than 1,700 studies on AI in education. Together they form a growing map of what has already been studied: the questions researchers have asked, the methods they used, from randomized controlled trials to design studies, user studies and qualitative work, the AI tools they looked at, and the data behind the findings. The Evidence Explorer turns that body of work into something you can search.
See what's already been asked
Search the research questions behind every study. Check whether your question has been studied, see how others framed it, and find the gap.
Research Questions →Find studies that use your method
Browse studies by research design, from randomized controlled trials and quasi-experiments to user studies, benchmark evaluations and qualitative interviews.
Methods →See which AI tools were studied
Find studies by the AI they examine: ChatGPT, Gemini, Claude, Llama, DeepSeek, Khanmigo, custom-built research tools and more.
AI Tools →Find data for your next study
See which datasets researchers have used (654 so far, from math benchmarks such as GSM8K and MMLU to essay-scoring sets such as ASAP and learning logs such as EdNet) and the studies that use them.
Datasets →Questions, methods, AI tools and datasets are drawn from each study's full text (or abstract, when the full text is not available) with AI; see each study for details.
