Last weekend, I stumbled upon Zoo Tycoon in the Xbox Store. Within a short minutes of playing, I was reminded of the ‘Sims’-like indicators that highlight how the guests and the animals are feeling. I found myself fascinated by the idea of the indicators in relation to the visualisations from a teaching standpoint:
what is the value of a teacher having similar indicators for students to highlight if students are happy, sad, angry, or simply feeling 'meh' (i.e. feeling the pandemic wall)?
From this context, I tried to track my own emotions for the week. For inspiration, I referred to Week 11 of Dear Data, in particular Stefanie’s drawing of colored lines.
In comparison to Stefanie’s drawing, I decided to create the visualisation with colors flowing from one to the other as emotions are:
- not perfect or precise in timing
- are sometimes fleeting, and other times long-lasting
- can be in direct result of something, or seem spontaneous to others because they were brought upon by a thought
Datafication of emotions is a difficult task for this very reason – emotions are personal; however, facial recognition in education is on the rise (Williamson et al, 2020). Looking back at my week, there are gaps in the data and estimations of when I shifted from one emotion to the other.
As a technologist, I often wonder not if, but how long, it will take for facial recognition AI to have an accuracy rating of 99% across a spectrum of emotions. Even more interesting is if you can ‘trick’ the AI, or know if the student on the camera is indeed a student and not a deep fake?
Rather than limit or reduce the view of students from a teacher’s perspective (Williamson et al, 2020), I am hoping a dashboard highlighting emotion would prompt action or provide a different perspective. For example, it could help identify if just one student is anxious, or if the class as a whole is anxious.
As highlighted by Bulger (pg. 4, 2016), in classrooms teachers are able leverage learner-centered instruction and personlise their teaching based on “interpersonal cues…. subject matter expertise… knowledge of how people learn, and knowledge of each student, to determine individual needs, adjusting their lessons in response to questions and behaviors”. A major concern here, as with the sleep data I considered last week, is data privacy and ethical implications. A teacher may ask if certain emotions should allow for more lenient grading, or how the teacher themselves can remain objective by constantly being exposed to the emotions of their students.
From a personal standpoint, I have seen many instances over the last year where the data in the dashboards that I present to clients are sometimes seen as ‘useless’ because of COVID-19. With emotions, I wonder whether this data is truly useful to the teacher. However, one thing that I have learned through experience is that in a remote world where we need to over-communicate every action and emotion even to those close to us because what we are going through and feeling is unprecedented.
As a final reflection for this emotional data, it may be more important that the teacher have high emotional intelligence and/or understanding of the emotions tracked rather than a deep understanding of the facial recognition AI and training data sets behind it.
Bulger, M. 2016. Personalized Learning: The Conversations We’re Not Having. Data & Society working paper. Available: https://datasociety.net/pubs/ecl/PersonalizedLearning_primer_2016.pdf
Lupi, Giorgia and Stefanie Posavec. Dear Data Project (2015). Accessed via http://www.dear-data.com/all
Williamson, B. Bayne, S. Shay, S. 2020. The datafication of teaching in Higher Education: critical issues and perspectives. Teaching in Higher Education. 25(4), pp. 351-365.