This semester, I’ve experimented with anonymous grading for the first time. Now that I think about it, it is a mystery why it took me so long to realize the obviousness of it, but better late than not at all, I suppose. As many other countries, the Netherlands does not have a tradition of anonymous grading at all, but I recently found out that in the UK it is fairly common practice, showing that it can be done. This was one of the topics of Jennifer Saul’s recent Aspasia Lecture in Groningen, and I am happy to report that she made such a good case for it that my colleagues in the evaluation board of the Faculty are already looking into adopting anonymous grading systematically.
Why should it be done? Well, for those of you familiar with the literature on implicit biases, the answer will not be hard to find: we inevitably rely on stereotypes and preconceptions to perceive and judge people, which serve as convenient heuristic shortcuts. This can have a negative effect on how we judge members of stigmatized groups (based on gender, ethnicity, class, geographical origin etc.), and it can also unfairly boost our judgment of privileged groups. With grading in particular, it has been noticed that anonymity significantly increases the average grades of members of these stigmatized groups, simply because their work is looked upon more objectively without the association to a particular person. (See this informative report by the British National Union of Students.)
One might think that, while anonymity may make sense for e.g. journal refereeing, it makes less sense when it comes to students, in particular in small groups. After all, after a few classes you should be able to overcome your implicit biases and form a neutral, accurate image of each student, right? Well, no; in fact, this very familiarity may blur your objective judgment of the quality of a student’s work in a different way. A well-known psychological phenomenon is the halo effect, i.e. our tendency to judge character or quality of a person’s performance on the basis of overall impression. So the talkative student who makes a good impression at the very first class will have an unfair advantage over the somewhat quieter student sitting at the back, and this is likely going to influence how the lecturer evaluates their respective performances. The talkative student will have an unfair advantage over the shy one (that is, if you are the kind of lecturer who likes talkative students!).
Once you’ve decided to engage in anonymous grading, you have the issue of how to implement it, from a practical point of view. The Blackboard system we use here in Groningen is entirely anonymity-unfriendly, and only allows for assignments to be uploaded under the students’ name. What I did instead was to ask the students to have only their student numbers on their essays (as well as naming the file with their numbers), and send them to me per email. I created a folder in my computer especially for the essays, and downloaded each essay as they came in, paying special attention not to create associations between names and numbers; it’s not difficult if you do the downloading in a fairly mechanical way. (Alternatively, you can ask someone else to do this for you and upload the assignments to a drop-box folder or something.) Here is a post over at the Feminist Philosophers with additional tips and resources.
One interesting thing I noticed while reading the essays is that my mind was always desperately trying to figure out who wrote each paper – understandably, as from its point of view, knowing the author would make the whole task considerably less time- and energy-consuming. I would then just resort to my preexisting ideas about each student and would thus be able to skip a couple of passages here and there, or so it thought. (I was mostly worried about the halo effect, but the effect of implicit biases concerning stigmatized groups was not to be neglected.) I did recognize some the essays as having been written by a particular student (especially for those who had chosen unusual topics and had discussed them with me beforehand), but for the most part I really did not know who had written what during the whole process.
All in all, I’m happy with the results of this ‘experiment’, and intend to adopt anonymous grading permanently, whenever feasible (i.e. not for seminar presentations, obviously). I am not sure to what extent my judgment did in fact become more objective in this manner, but I am pretty sure it cannot have become less objective, and that’s a good enough reason to continue. Philosophers or not, we are all just 'ordinary people', and need all the external help we can get to counter some of our cognitive tendencies when they are not conducive to optimal cognitive performance in specific situations -- such as grading.