Recent developments in generative AI – not least the notorious ChatGPT – offer opportunities to stimulate long-overdue reforms to assessment practices. But these reforms need careful thought and principled implementation within an already hard-pressed sector.
Carless, D. (2023). How ChatGPT can help disrupt assessment overload. Times Higher Education Campus
Note that the article I link to is about addressing assessment overload. The first sentence is what caught my attention and led to the chain of thought below.
David Carless makes the point that language models have given us an opportunity to reform assessment practices, which sounds like a good thing. But weren’t MOOCs supposed to do that? And blended learning before MOOCs? And the internet before that?
We’ve had many opportunities for assessment reform following the introduction of new technologies that threaten to disrupt the ‘business-as-usual’ of universities. And universities have almost always ignored those calls. But this time I think it really may be different. The degree and speed of response by higher education institutions has been shocking, at least to me.
These are institutions that take 12 months to put together a policy document. But many of these same institutions had position statements on ChatGPT out within a couple of months.
But I don’t think that universities were worried about assessment validity. Not really. Why would higher education, as a sector, pay attention to these ‘new opportunities to reform assessment’, when they haven’t before? Why are they paying attention now?
Could it be that the panic around this technology is because it undermines the accreditation function of universities, which is essentially the only thing they provide that students can’t get elsewhere?
If students use language models to complete assessment tasks, to the point where we can’t trust the accreditation function of universities, what would students be paying for? Do language models, and the risk to assessment validity, threaten universities’ economic models? Is that why the institutional response to ChatGPT has been so swift and so unequivocal?
This is what I’d call a hot-take, having not spent too much time thinking about it. Let me know why I’m wrong in the comments.