#AI #Education experiments can be great, if careful.
A 6-week #writing course gave one group an in-person instructor and the other group a #languageModel.
But the AI group accessed the #LLM *during* the pre-test?
Shouldn’t baseline conditions be equal?
4 media
