What if we could know what our students are feeling? (Rebecca Lewis)
When thinking about technology enhanced learning, I often get asked the same question - how do I know if my student is engaged when they have their head buried in their screen?
It’s a challenge that I also face in my own teaching and I am acutely aware, when conducting a follow-along demo on the projector at the front, that the PC lab classroom layout often doesn’t lend itself to easily manoeuvring around the group and checking up on their progress.
Blended approaches (where we can check in with our learners face-to-face and ask for feedback) offer some solution here, but as a true believer in remote learning I wonder about other ways to track engagement in online learning.
It is in thinking about this dilemma that so many teachers face that I came across DAISEE (the Dataset of Affective States in E-Learning Environments). DAISEE is a vast library of video frames of individuals, focusing on facial expressions, recorded while the data subjects undertook various forms of remote e-learning. The hope is that in the future we will be able to utilise this catalogue of human reactions to create automatic engagement detection tools that utilise facial expression tracking, heart rate monitors (in smart watches and other wearable technology) and other sensors to determine if our learners are actually enjoying the learning content they are undertaking. We will be able to track automatically when our students feel boredom, surprise, delight amongst other emotions.
It has big implications for effective curriculum design, but also big implications for the personal privacy of our learners. I will keenly (but cautiously) await seeing if this new technology takes off!
Response:
This is a fascinating sounding project and I believe that this technology can be a very useful tool for the teacher. I teach maths to students who do not necessarily have a technical background and so the results of such technology would be very beneficial to help me develop sessions that are relevant and pitched at the right level to students. However, as a student in information security, I do have concerns about the privacy aspects of this technology. I shall be looking out for future developments and hope that the technology is implemented in a privacy-aware way. (Ashley Fraser, Information Security)
Current literature indicates that evidence-/scientifically-informed approaches powerfully impact learning. This video, reporting on positive effects of incorporating Twitter in the classroom, might be a useful case study for tracking engagement in e-learning. https://www.youtube.com/watch?v=2w9CnaeaiAE (Annabelle Lee, Music) x
This is a fascinating concept, which, as Rebecca mentions, could have significant implications for curriculum design.
Others have already commented on the ethical issues, but I wonder how DAISEE would deal with unconscious bias. This study by Jack, Caldara and Schyns (2012) concluded that the interpretation of facial expressions is not universal, which might mean that the algorithms used to detect the expressions might be influenced by the cultural background of the AI programmer. (Eva Dann, Library)