Alternative Approaches to Exam Feedback
With the end of the semester upon us, exams are once again a topic of conversation in the blawgosphere. I’d like to jump into the exam conversation by looking ahead a few weeks to a time when we’ve graded all of our exams and are thinking about how best to provide feedback to students about the exam. I’ve seen some discussion about exam feedback, and I am sure there has been a lot of additional discussion that I have missed, but the subject seems to have been somewhat under-explored relative to other exam-related subjects. My sense is that most folks provide a model answer, model issue analysis, sample student answer, or some combination of the three. I’ve used each of these approaches myself. Lately, however, I’ve been finding myself thinking about alternative methods.
Specifically, inspired in large part by this fantastic post by Orin Kerr from a couple of years back, I’ve been considering whether to try a tiered approach to my model answers by providing three example answers: one “bad” answer, one “good” answer, and one “excellent” answer. Why three example answers? I’ve found a model answer alone to be of relatively limited value to a student who has not done well on an exam and wants to figure out how to improve. While a model answer may give students a good idea of my ideal exam answer, it doesn’t really let students know what it is that distinguishes the model answer from lesser answers. I’ve been struck by this when I’ve met with students to review their exam. I always have them review the model answer prior to our meeting. And, quite often, it seems like they draw the wrong (or, at least, not the most important) lessons from the model answer. They see, for example, that the model answer references some of the cases we read while their answer didn’t mention a single case and they think “if only I’d referenced the cases, I probably would have done great.” Or, they see that the model answer discussed facts that they failed to and think “if only I had simply referenced those facts I would have had an “A” answer.” In short, many students seem to mistakenly believe that each exam has a simple formula where if they spot the issue, state the correct rule, merely mention facts A, B, and C, and make some sort of reference to case Y, they will get all the points and a high grade. As we all know, of course, this is not the case. Indeed, one of the most common exam errors I see is that the student approaches the problem as a sort of a checklist–they mention many of the relevant facts and state the law accurately, but fail to really apply the law to the facts. Meanwhile, students toward the bottom of the class often look at the model answer, see that their answer falls far far short of the model, and wonder “where do I even begin if I want to improve.”
All of this has led me to think about how I might be able to give students a clearer picture of what distinguishes the model answer from a good answer and a good answer from a bad answer. Orin Kerr’s post, which gave brief examples and explanations of bad, good, and terrific answers using a short and fictional legal set-up, seems to provide a pretty effective structure for achieving this goal. Giving students examples of a “bad” and “good” answer, in addition to the traditional model answer, would provide them with a clearer picture for how they can improve. A poorly performing student, for example, would be able to see a detailed explanation of why a “bad” exam answer is “bad” and get a concrete idea for how she might be able to raise her performance to the next level.
Of course, there are some downsides and limitations to this idea. Perhaps the biggest downside is that it would take a lot more time to write up three example answers along with explanations than to write up a single model answer (and even more time that simply distributing the top 2 or 3 student answers as models). As for limitations, there are a wide variety of reasons that an exam answer might be sub-par, so it would be impossible to include them all in any example of a “bad” and “good” answer. Still, on balance, I think that the additional effort needed to write three example answers would likely be worthwhile, especially for dedicated students who truly want to make the time and effort to improve their performance but are honestly unsure of how to do it.
I’d love to hear any and all comments about this approach (especially if others have adopted it themselves) or any other thoughts or suggestions folks might have about alternatives to the standard methods of exam feedback.