How and why do students use GenAI feedback? Currently reading Henderson et al. (2025)

May 16th, 2025
Posted in literature
Tagged GenAItrust

More reading on trust in GenAI, today on how and why do students use GenAI feedback. Henderson et al. (2025), “Comparing Generative AI and teacher feedback: student perceptions of usefulness and trustworthiness“, use responses from approximately 7000 students to understand their thinking regarding to their use of GenAI.

They find that 1/2 of the 7000 respondends used GenAI feedback (which is, as they mention, not necessarily negative — feedback seeking is good for learning!), while the other half did not.

Of those that use GenAI feedback, more than 4/5 found both GenAI and teacher feedback helpful, but 9/10 rated teacher feedback as trustworthy, whereas only 6/10 did that for GenAI feedback. Students generally valued GenAI feedback because it is available at any time, anywhere, can be as much as they like, and is a lot easier to ask for than asking a teacher for feedback, which is perceived as “risky”.

Of those that did not use GenAI feedback, about 3/10 said they would not know how to even do it (e.g. how to upload an essay and prompt for feedback). This is concerning, seeing that of those that use it, more than 4/5 say that it is helpful to them, so all students should at least be able to get feedback, even if they then choose to not do it. About 1/5 students that did not use GenAI feedback also commented on values as reasons for why they did not want to use it (e.g. valuing human connection in the learning process, or concerns about pricacy and integrity).

But all students, whether they use GenAI feedback or not, report concerns about reliability as well as missing context. Teachers are of course in a much better position to know about the assessment, course, students, and “as a result are able to identify ‘what matters’” much better than GenAI, and students are aware of that.

Henderson et al. (2025) conclude that “GenAI and teacher feedback appear to serve different needs, and therefore are complementary but not interchangeable“.

I think it is really interesting to see that there is this big divide between students that use GenAI purposefully to support their learning (maybe not exclusively, but at least partly) and then other students who do not know how to do that. As Møgelvang et al. (2025) write, missing GenAI literacy has effects not just on a personal but also society level. I also think that it is important to stress the point that students understand, and also policy makers need to understand, that while GenAI feedback might be helpful, it cannot replace teacher feedback for many reasons, not just of reliability (which, of course, also needs to be investigated!) but also of the human connection that is important for learning.

Henderson, M., Bearman, M., Chung, J., Fawns, T., Buckingham Shum, S., Matthews, K. E., & de Mello Heredia, J. (2025). Comparing Generative AI and teacher feedback: student perceptions of usefulness and trustworthiness. Assessment & Evaluation in Higher Education, 1–16. https://doi.org/10.1080/02602938.2025.2502582

And in #WaveWatching news: It is finally summer!

There was this whole bunch of small fish right next to Långabryggan, and it was fun to see how they tried to avoid being too close to the shadow of my head as I walked there!