Return and Reflect …
No no, whilst the title might be suggestive of the sector’s go to defence against the dark arts of AI, today is about giving back the feedback and collecting the student voice. If you wanted hints into obtaining ethics of pedagogic research, wrong day, try day 4!
Firstly, the return of the AI feedback. Whilst it would be lovely to suggest that we had the perfect Virtual Learning Environment (VLE) solution for this… we did find the level of customisation that might be required to pull that off natively. I will outline two approaches, one (what we did for this) custom python code, two using Power Automate (low code).
Rich, being the data science master that he is, wrote Python code that took the .docx file that was produced by AI, matched it to the student for whom it belonged, called outlook, attached it, placed boiler plate text + mail merge type customisation and sent it direct to the student via a shared mailbox! Genius right!? However, the same outcome can be achieved in a low code way using Power Automate within Microsoft platform, in short it is about setting up a spreadsheet with the fields and addressed much like a normal mail merge but it gives you more options to attach and customise. Do get in contact if you want to explore any of these tools, although there is lots of good help on LinkedIn learning etc.!?
What did we learn from students? Well, there is still work to do on coding the data properly etc. however some helpful anecdotes:
- Somewhere ethically opposed to the use of the tools, which for me suggests the importance of opt-in.
- 71% of those that received it thought it was quite or very useful
- 89% of students made changes to their work as a result of the feedback provided
- There were mixed feelings about the lack of perfection of the output… (aspects of which were intentional…)
- That said, they wanted more opportunities to obtain feedback like this
- I think there was a slight misunderstanding regarding the infallibility of humanity attached to this also.
Going forward, our little team is likely to embed more of this into our first year unit and take a more intentional and aligned approach to its use. This year showed that for this cohort of Engineering students, they were broadly comfortable with its use but had not always appreciated the limitations of what AI can do. A more systematic approach and development over the year is likely to address this, but in the high-speed development of AI, we were including the opportunity in a Just In Time (JIT) delivery model.
And that’s ‘a wrap’, I believe they say!
With many thanks to Dr Aisling Tiereny, the BILT Team, many from IT services and a whole lot more, that made this activity and mini blog series possible!
P.S. Do be in contact if you want to hear more, or we would be happy to give a short seminar if interested!
Leave a Reply