So every semester, as the school starts, I take stock of the students in my classes. I ask myself, who are going to be my best students? As someone who has taught for six years, I would think that I have started to see what makes students successful in my classes. I tend to highlight the students who on the first day of class have read the syllabus, the ones who introduce themselves to me, who introduce themselves to the students around them, who ask good questions and who seem very attentive.

I try to write down who I think will be my best students. Then at the end of the semester, I do a sort on the who has the highest grade in my class.

I am wrong every single semester. EVERY. SINGLE. SEMESTER. What’s worse, I’m usually dead wrong.

I tend to assume that the extroverts will do better. It’s not an impossible thing to imagine, since I teach a team-based class. But it’s not just the great team players. I also tend to assume the students who do best are those who reach out, who check in when they are struggled. There are student often (I teach large classes) where the students who are doing best often haven’t built a relationship with me at all.

I also tend to assume that it’s the students that are outwardly engaged, and by outwardly engaged, I mean students who are friendly. And probably those students are more likely to get the lions share of the recommendations and the mentoring, but in terms of the grades, those people aren’t the people who are killing it.

Sometimes I think it’s my point system, like it rewards some type of perfectionist behaviors (like making sure you do an assignment perfectly) or perhaps overly stresses turning things in over graded work. But I have founded that when I revise my grading schema, I still am surprised by the results.

The point of this story is that you may think that you know your students, and you may consider yourself a “fair” professor, but your attempts to understand what will make someone successful in your class is flawed because we all make assumptions about people. We think we know what types of behavior that makes people successful, in reality we just scratching the surface.

I’m a pretty complicated computer. I know a lot about my students and I know a lot about what makes them successful. I teach the same class over and over again so I see what works for students and what makes them less successful. But even I can’t predict what will make the students successful.

In some ways a computer algorithm is going to be better than me because it probably has less implicit bias. I’m not sure though. A lot of the socio economic indicators that makes a student seem more extroverted or committed to class could also show up in lots of other data. What about the common correlation between “at risk” youth, or students who come from “weak” high schools? They could easier thrive in this new environment. Additionally, as a qualitative human, I process lots of other behaviors and attitudes that a computer would probably not see. For example, I can tell if a student says my name right on the first day of class, or bring a pencil to write on the syllabus, or introduces themselves to classmates. That’s not stuff that I would hope that a university would gather on their students, nor would I think that a computer could guess all the things that might affect how I guess who will be my best students.

What I’m going for here is that people are complicated, and we think that because we give out the grades we understand what going to make students successful. I’m not we do, and I don’t think that we ever will. The important part is to make the interventions and the changes as they happen and when we see them, so that people can become as successful as I would like them all to be, as I eagerly scan my first day of class.