by Gordon Cooper
Like my colleagues Kate and Gareth, I have been advocating the benefits of the approach of audio/video feedback for several years. After a number of iterations the drive is linked to allowing students a key opportunity to develop, by providing extensive feedback on a draft submission of a project or dissertation report using Kaltura. From a personal perspective, it is satisfying to be able to provide a very detailed critique to a student, emphasising key points with voice tone, stressing positives and discussing areas for improvement, while being able to directly point to the text and images on a screen. Data from a case study has shown that in general students have found the Kaltura output more useful than traditional written approaches feedback. My perception is that follow up meetings have been more positive and focused.
Can GenAI provide an effective way for students to close the audio/video feedback loop?
GenAI - The elephant in the room. Generated by Apple Image Playground
At this stage in our discussion it feels a fitting to turn to the elephant in the room when we talk about teaching and assessment in HE- Generative AI.
Why is this important?
Organisations such as the World Economic Forum provide regular forecasts considering what skills employers will be looking for in the coming years. In the figure below, on the left we show the predictions made in 2020 for today, while on the right is the vision for 2030. While not even on the radar in 2020, Generative AI is now the fastest growing skill for 2030. We have a responsibility to students to embrace and embed GenAI in our curriculum, training them for future roles. By this I do not mean using ChatGPT to write their essays, but taking advantage of other tools such as NotebookLM which are excellent for drawing together information sources the user provides and summarising trends in the data.
Information from the World Economic Forum, from 2020 predicting the top skills required today
Information from the World Economic Forum from January 2025, evaluating where we will be in five years time.
What can we try? - Using NotebookLM for feedback appraisal
A current pilot study is evaluating whether there are benefits to using NotebookLM in using feedback in a positive and productive manner to close the loop. Uploading draft documents and the transcripts from audio/video recordings allows a summary to be produced indicating key common areas that require attention. Prior to submission a final version can be reviewed to understand whether feedback points have been addressed. Provisional tests of this method look to produce an output that will help students improve their submissions. An important caveat of this approach is that it is looking at overarching trends in the work, and won’t consider some of the more specific points raised in feedback. In this way it supplements listening to the audio/video feedback rather than replacing it.
Could this be expanded to map a student’s feedback journey?
Thinking forward even further, it will be exciting to investigate whether NotebookLM could be used in this fashion to chart the feedback journey of a student as they progress through their degree demonstrating how they have developed with time.
Gordon Cooper is a Senior University Teacher and Examination Officer at the School of Biosciences.