Exam Pre-Mortem

Continuing my “disaster” reading the past week, I finished Meltdown over the weekend. Like common popular books published in the past decade, it’s a light and quick read, styled with the usual vignettes adored by journalist-types, with a smattering of self-help. I prefer the heavier, more theory-laden, analysis – hence my choice of reading Normal Accidents before Meltdown. The latter draws quite heavily from the former, making my read quicker, however it has thought-provoking moments. One of these is the Pre-Mortem.

We’re more familiar with the post-mortem – dissecting what went wrong after the bad event. The pre-mortem attempts the analysis before, rather than after, things go poorly. The example in Meltdown was asking students at a business school what can be improved, but the experimental design cleverly had two different prompts. One was the typical “what do you think can be improved?” (more elegantly phrased than my paraphrase) and generated the range of standard student views one might see on such evaluations. The other prompt asked the students to imagine they were alumni two years after graduation, learning that the business school was doing poorly and suffering in reputation. The question was to imagine: “What went wrong?” Essentially, a pre-mortem.

As you might have guessed, responses to the pre-mortem question showed larger variation in scope and increased depth of analysis. Some ideas were more far-fetched, but interesting to consider nevertheless, especially if you were doing some risk-analysis planning. A number of years back, when I was faced with a very difficult and significant life-career decision, I did a pre-mortem – except I didn’t recognize it as such at the time. It was not the only part of my analysis (some forecasting was required), but it was an important piece. For a couple of years after, I regularly debated with myself as to the wisdom of my choice, but I can now look back and see that it was the right thing to do at that juncture. As a planner-administrator, I still regularly do forecasts and analysis, however Meltdown reminded me that I should employ the pre-mortem more often.

As a first step, I wonder if I can use the pre-mortem to help my students. Here’s the setup I’m thinking for a pre-exam writing assignment. “Imagine you just got back your exam and you earned a D. You were both surprised and disappointed. Looking back through your exam and your preparation, what went wrong? Discuss whatever comes to mind.”

Would this help the students better prepare for exams? I hope so. I already have “how to study for this class” tips on my syllabi (including ones from previous students), and I’ve built in some formative assessment approaches to prepare students. I’ve now taught a full year of G-Chem where the students used take-home exams with post-exam self-annotation. In P-Chem, I’ve encouraged group study and exam question forecasting, and tried to help the students annotate their problem sets with some “what went wrong” self-reflection. However, these are all post-mortems.

I’d like to think that the post-mortem on the first exam (if a student does poorly) helps in preparing for the second exam. (My classes typically have 3 or 4 exams before the Final Exam, so the first exam shows up quite early in the semester.) But upon reflection, I’m not so sure. There are students who do well on the first exam but then bomb the second (for a variety of reasons, overconfidence being one). There are students who come by my office asking how they can do better on the next exam; I typically turn the question around and ask the student to reflect. And there are students who don’t darken the door of my office, some proportion of whom probably do zero post-mortem on a previous exam. I also hear students in the hallways talking about getting back exams (from other classes), and the reflection is often shallow to non-existent.

So I hope the pre-mortem will help. Maybe I should do it more than once. Maybe it will also help avoid a meltdown. One can only hope.