In life generally I am embarrassingly bad at keeping track of people with whom I’ve had only limited interaction. There are many people who I remember visually and may even know something about, but who I cannot name. There are probably even more people who look vaguely familiar, but for whom I couldn’t say if I know them from one place or another.
This all gets exaggerated in the context of working as a teaching assistant. University of Toronto “tutorials” routinely have more than 30 students enrolled, though attendance may be only 50-70% in a given week. For a single class, I will usually have 3-4 tutorials and I sometimes teach as many as three classes at once. Often, courses only last for the fall or the winter term, so I get a new set of students after the winter break.
From the perspective of a student, even if they take a full course load of five courses at a time and every class includes a tutorial, that’s only a maximum of 10 teaching assistants per year, or perhaps 40 during a bachelor’s degree. Furthermore, in all but the most active tutorials I will be speaking for a good portion of the time, so every student has reasons for their memory of me to be reinforced. Many of them will also see me in office hours to discuss essay drafts or grades.
From my perspective, during the seven years of my PhD I will probably have an average of two courses at a time with a 50/50 mix of year-long and one-term courses. At three tutorials per course, that’s 270 students per year, many of who I will only see in a minority of tutorials. As a result, there are probably already well over 1,000 students who remember me as a TA, and of those there are probably only 5% or so who participated enough for me to have any kind of clear memory of them and even then it’s probably just a “former student” flag without details about which course, much less what they said, their name, or any personal details about them.
It would be less socially awkward if I could remember at least those basic fields of data about everyone in the large set, but it’s possibly beyond what a reasonable human memory and cognitive capacity can achieve. It’s certainly well beyond what I can achieve, as someone who has routinely worked in groups or offices of 30 people or so in which I knew only a minority of names, even after being involved for months or years. In a way it may even be a good thing. I would prefer if tests and assignments at U of T were marked with student numbers only, not names. Inevitably, seeing a name you remember will subconsciously influence grading (though the impact is hard to predict: do I interpret the work of more involved students more leniently, or do I expect more from them?). I avoid looking at names on documents I am grading and it may be a further protection that even for the majority of students who I am currently teaching their name won’t be more than vaguely familiar to me and not tied to any specific memories of them as a person.
One awkward dimension of all this is the frequent expectation in the syllabus of courses that TAs will grade students for both attendance and participation. If participation is to be assessed by spoken contributions, I don’t think I can track it accurately at the same time as I am trying to facilitate a worthwhile discussion. I buy and distribute name cards to raise the odds that I will be able to identify students, but people don’t always use them consistently and it would be too embarrassing to tell a student months into the year that I have no idea who they are and can they please give me their name for my participation records.
I also question the fairness of grading people based on spoken contributions, since the people who feel empowered to speak may just be the most extroverted, confident, and privileged. Based on experience I can say conclusively that the most talkative aren’t necessarily the best informed or those who contribute the most value to the discussion. Students are also smart enough to game any system, so if they know that I am trying to check people off when they make a substantive contribution to each week’s discussion they know how to do just enough to get a checkmark.
One approach I learned from another TA is to give everyone a 10 minute writing task during each 50 minute tutorial, then rapidly scan that people have actually done it as they are leaving, checking it off against the attendance list. It means sacrificing some discussion time, but it also means that people cannot be entirely tuned out during the tutorial. It’s impossible for me to tell whether a silent person is listening attentively or thinking about something entirely unrelated. Teaching assessments suggest that students don’t find these writing tasks pointless or distasteful.
It would be interesting to try teaching undergraduates at a school that emphasizes teaching more and is willing to constrain tutorials to a size where it’s actually possible to know most students, and where most people will actually speak in any given tutorial. In my ideal world, I would also implement the system we used in graduate tutorials in Oxford, where someone at random is called upon to briefly summarize each reading in 2-3 minutes. It definitely drove me to do the readings then, out of fear of embarrassment. I did try the system at U of T, stressing how the summary can be very brief and how these records will be perfect study notes, but students hated it, complained about it in their teaching assessments, and said that it drove them to not attend class.
Related:
Encouraging equitable tutorial participation
Tutorial attendance
Attendance histogram