Respect for students means no surprises
Do you think it's a good idea to leave your students guessing what will be on the test?
I remember a conversation about teaching, with a colleague at a conference. She was lamenting that how students kept asking about what was going to be on the upcoming exam. She wondered how I handle this with my students.
I said that I used to get this a lot, and I addressed it by being very specific about what students could expect to see on the exam.
How do you do that, she asked.
I told her that I give them a review sheet, about one page, that defines the scope of the exam. It’s specific enough that the students will have plenty to do to make sure that they’re prepared for the exam, and it provides reassurance that they won’t be surprised by anything on the day of the exam. While I don’t put everything on this review sheet on the exam, I’d say that most of it ends up on the test.
I was dismayed by her response. She said, “I don’t do review sheets.” She went on to explain that it’s a matter of principle that if she spent many hours in class teaching a broad variety of content, that students should be expected to be able to know it all when the exam rolled around. That it’s a student’s duty to figure out what is important and what’s not, and the more they learn, the better they’ll do on the exam.
I don’t know about you, but I think that’s absurd. I mean, when I have to re-take the annual safety training for campus, I have trouble memorizing whether the eyewash needs to have 12 inches or 18 inches of clearance, and whether the difference between a B fire extinguisher and a C fire extinguisher. But we expect our students to memorize everything that comes out of our mouths over the course of several weeks, and their ability to regurgitate it back to us with fidelity is the basis for their grade? Yowza.
I also think that there’s a huge cultural bias embedded in “knowing what’s important,” because it’s the instructor who gets to be the arbiter of deciding what is important enough to be on the text. And it’s perfectly acceptable and reasonable for other people to think that other content in the course is more important that you do. What this means is that students who are better off at understanding the mindset of the instructor are the ones who will have the best idea about what the instructor will put on the exam. Who is best situated to do that? The students who share similar cultural backgrounds with the professor, who understand where they are coming from, both literally and figuratively.
That was several years ago. I don’t know if my colleague still thinks it’s a good idea to keep people in the dark about what is expected of them.
While we are complex people, I still think we can accurately make a generalization that professors can be sorted into one of two broad groups: one are the kind who behave as if their course is the only thing that should take up their students’ time and concern, and the other kind are the ones who recognize that students are just like us in that we are juggling complex lives involving tradeoffs and challenges involving family, work, money, health, and so on.
If you’re in the first group, well, you’re just an unreasonable human being, and you’re not going to change your mind by reading some schmo’s newsletter. You probably haven’t gotten this far, anyway.
If you’re in the second ground, then you accept that we don’t expect students to dedicate their entire lives to the course that we are teaching throughout the semester, then we should be able to accept the notion that there is a finite amount of time that students have for what we are teaching. It then follows that we should not expect our students to be responsible for fully digesting and memorizing an inordinate amount of information, and it would be unfair to simply reward the students who happen to have more bandwidth. We should set reasonable learning outcomes that are accessible to all of our students. Which means that we need to make sure that the task of preparing for our exams to succeed should involve a reasonable and finite amount of effort.
Which means that we shouldn’t be rewarding the students who happen to be more lucky or better at guessing what we’ll be evaluating them on.
What does a good exam review sheet look like? I mean, it can be a lot of things, and I think this really depends the level of the course, the nature of the content, and so on. I think the key elements involve: a) defining the scope of the exam within the realm of a reasonable amount of content and b) being specific enough that it provides students what they need to know at at what level. For example, “everything in chapter 18” is a bad thing, but “Being able to explain the structure and function of plant tissue types and where they are located within the plants we studied” is better.
When I’ve given exams, I’m a fan of short answer and essay questions. I would give the students one week before the exam a single page with a bunch of questions, maybe 35 or 40 of them. I would tell them that about ten of these would be on the exam. And then I would select (using a stratified randomized design) these questions for the exam. This was I was able to get them to study more than I was going to be testing them on. But if they came to me asking about what was going to be on the exam, we would be talking about the content of the question itself. And if you do this with the whole course (in the LMS, for example), then that discussion is an opportunity for everybody to learn.
What if you’re teaching a ginormous course and are expected/needed to use multiple choice tests? Well, can’t you hand out or assign a bunch of the questions in advance and then draw from them for the exam? Isn’t the point of the course for students to learn, and wouldn’t this exercise support that process?
Now that you’ve gotten to this point, I’d like to ask you an important question: why are you using exams at all? Is it possible that you can use time in your courses more constructively, and find other ways to evaluate student learning? If you’re not down with fully ungrading, then at least see if everybody is better off without big high-stakes exams? Just a thought?
I agree with you 100% about the need to consider equity when we design assessments! I've spent a lot of time thinking about this too, and have also written about my approach here on SubStack. Back before covid, I did what you explain here: lengthy prep sheet preceding in-class assessment. The prep sheet included what I wanted students to learn and I told them the test would be a "sample" of that "population" of content. Students liked it a lot: they knew how to prepare and that's huge! Out of necessity and in the spirit of EDI and UDL I made a dramatic change during the zoom-school phase of the pandemic, and that change worked so well I am still doing it now 4 years later with no intention of going back to my "before-times" approach. If you are interested in my radical redesign and philosophy behind it, check out the newsletter here: https://ericakleinknechtoshea.substack.com/p/trade-tools-mastery-testing