Generative AI and the Changing Nature of Assessment for DataLab Education Community

 

I enjoyed co presenting with Daire Maria Ni Uanachain . Daire's presentation and recording of session is now available on the Data Lab Education  Community Page 

If you are not a member of this community you should join. ( probably most relevant to folks in Scotland)   My own presentation is at foot of this blog post and I think complemented Daire's approach. 

I don't think we are at an inflection point yet. Changes to learning and assessment should have arrived when the internet and search engines became ubiquitous. 

Generative AI poses less of a threat to vocational training than to traditional academic sectors because vocational education relies on authentic assessment where learners must physically demonstrate practical capabilities rather than simply writing essays for these areas.  Digital evidence capture, such as video recording practical assessments (e.g. solar panel installation or working at height), work really well for verification and quality sampling.  Alternative assessments work in every other subject area too.   

To be clear, educational AI isn't about experimental frontier systems, it's about the accessible, everyday tools already at our students' fingertips. Keeping our focus on practical applications helps demystify the technology and prevents unnecessary anxiety from stalling essential digital skills development  (though the danger of armageddon will be this year's excuse for avoidance of any digital skills development) 

The truth is  getting teaching staff to change assessments rather than rely on 'AI Detection Engines' which don't work,  is massively challenging. Bad practice is deeply entrenched. We still have academic staff who ask generations of students the same questions and that should have changed long before the advent of AI.

Overcoming resistance to change is all about having the right governance, time and support in place. 

Unlocking the potential of these tools locally hinges on closer collaboration between IT departments,  academic and support staff. Building trust and dialogue between teams ensures we fully leverage the native AI features already sitting in our VLEs (like Canvas) and the enterprise platforms Microsoft or Google accounts. 

This is essential to prevent growth of shadow IT etc.   

There is useful support from Qualifications Scotland and a host of other sources. I've listed some current ones in slide deck. 

Hot of the press Glasgow Universities Toolkit for Reflecting on the Use of Digital and Artificial Intelligence (AI) Technologies in Learning and Teaching is included on this list.