New preprint: I got 400 postgrad health management students to use generative AI in an assessment, rather than banning its use. An honest write-up of what worked and what didn't.
https://osf.io/preprints/edarxiv/js6kw_v1
The good: most students found it valuable. The structured scaffolding helped reduce anxiety for for adult learners who'd previously been told never to use AI for any assessments.
The less good: most students engaged instrumentally (Which tool is best? How do I prompt?) rather than critically (Which perspectives does this privilege? What are the equity implications? What is this missing?). A single assessment can help normalise AI engagement but probably can't develop deeper critical AI literacy on its own.
If you're designing courses this year and wrestling with AI integration, I hope the case study and design principles we lay out are useful. It's a preprint - feedback welcome.