Just What Is Your Best Exam Format?

Now that I’m done with hours upon hours of post-semester meetings and painting my house’s front door, I can comment on Simon’s recent post about open-book exams.

Abandon all hope, ye who enter here.

One’s choice of exam format reflects two questions that are often in conflict. Will the exam give students a reasonable chance of demonstrating whether they have acquired the knowledge that they were supposed to have acquired? Can the instructor accurately, impartially, and practically assess the exam results? For example . . .

  • Oral exam: great for testing exegetical ability on the fly, but extremely tiresome and unfeasible if the instructor is teaching more than a couple dozen students in a course.
  • Multiple choice questions: very easy for the instructor to grade, minimizes student complaints, but encourages binge and purge memorization.
  • The timed essay exam, whether open- or closed-book: also tiresome to grade, often susceptible to instructor bias, and, perhaps most importantly, reinforces the unproductive notion that writing (and thus thinking) does not need to be a careful, deliberative process.

How does all this affect me? Over the years I have moved away from formal exams and toward a final culminating assignment — such as a take-home essay question that I reveal in the last week of the semester — intended to test how well students are able to apply concepts to an unfamiliar situation. But lately I’ve become disenchanted with this format, too.

Simon’s post prompted me to think back to my own days as a student. Exams in physics, mathematics, and engineering consisted of, essentially, solving a variety of puzzles — full marks required both supplying the correct solution and documenting how one arrived at it. The primary ability being tested was concept application. One prepared for these exams by working on practice puzzles involving the same concepts. Courses in political science, history, and whatnot had timed essay exams. To prepare for these, I would guess at likely questions and create outlines of essays that answered these questions. I would repeatedly hand-write an outline to memorize it, then turn it into prose during the exam. Even if my guesses weren’t entirely accurate, they were often close enough for the outlines to be very useful.

I’m now wondering if there is a way to turn the outline creation process into the equivalent of an exam. Something that gets graded, but not as part of a scaffolded research paper assignment.