What’s going to be on the exam?

Do you love it when students waste office hours with questions that don’t help them learn? Do you want to cultivate anxious emails from students sent at 3 in the morning? Do you want your students to wager their grades by guessing what you think is the most important material?

Then don’t tell your students what is going to be on the exam.

Students often have a perfectly normal question: What’s going to be on the exam?

In my opinion, the worst answer is “Anything is fair game for the exam, if it was mentioned in class.”

I should first confess that at some early point in my teaching career, I used to do this, or something along those lines. Why did I do this? I honestly don’t know, I think I was just teaching the way I was taught as a student, and the way that a lot of my peers were teaching.

This approach makes writing exams easy — you can just come up with obscure facts from the textbook or things mentioned in class as an aside — and then reward the students who happened to know these trivia. Sure, it’s easy, but it’s also crappy pedagogy.

I think a blanket answer of “anything covered in class” essentially amounts to hazing. Because this requires way, way more work from students than should be expected from a 3-unit class. Yes, even a challenging one like yours.

It is entirely legitimate for a student to be told the basis of their evaluation. Students take a course, and earn a grade. They should be made aware, as specifically as possible, the foundation for this grade before they do what it takes to earn it. The less they know about the basis of their evaluation, the less fair we are to our students.

The more specific you are about what is on the exam, the more transparent you are about how students should study and how they are evaluated. This level of specificity gives you control over the material that your students will study. I have heard colleagues frustrated that students aren’t focusing on studying the right material, or asking the right questions while studying. The more specific you are about what you want the students to learn, the less this becomes a problem.

Vague expectations prevent students from focusing on the facts and concepts which are most important, and may lead to some students wasting their time on minutia. The more vague you are about what will be on the exam, the less control you have over what they study. When students worry about the mysterious contents of the exam, they aren’t focused on learning. Vagueness rewards academic savvy over actual learning. Our value in the classroom isn’t our mastery of content knowledge itself, but our expertise that allows us to parse the useful, meaningful and relevant. Asking the students to master too much information will result in no mastery at all.

In my first tenure-track job, I decided that the norm in my department — “anything can be on the exam” — was a bad thing. Since then, I’ve decided to design exams one of three different ways, all of which are designed to remove guesswork on the part of students:

• I give students a full list of potential exam questions in advance. I then select a subset of these questions for the exam itself, choosing at random, haphazardly, or with a specific rationale. Here is an example of one, from a non-majors Environmental Biology lecture course*.

• I give students a comprehensive exam preparation sheet, one or two weeks before the exam. I give them a solemn promise that everything on the exam will be covered by one or more items on the review sheet. Sometimes these items are very narrow but other times they could be rather open-ended. But they are never intended to be vague. I tell the students that if any question on the exam isn’t based on one of these review items, then I’ll drop it from the exam. I am also tempted to hand these out at the beginning of the semester, but I call too many audibles to make this a wise choice. Here is an example of one, from the first exam in a biostatistics course. You’ll note that there’s a lot of material on there. I can’t ask questions to cover every one of those items. But I can make sure that students study them all, but also make the scope narrow enough that it is do-able.

• I can give a take-home exam. I only do this if I’m blessed with a very small class.

When anybody asks, “What’s going to be on the exam?” I usually can answer that with a single piece of paper.

Sometimes some students will email me, “What’s the answer to number 8 on the exam prep sheet,” or they’ll write me an answer and ask me how it meets my expectations. I make a point to not evaluate their responses or give them any information, unless I do so for the entire class. I might clarify a question or an item, if a student doesn’t understand the words.  Under all circumstances, I assiduously avoid evaluating providing privileged information for the students who feel more comfortable with approaching me for private studying advice, because that would be unfair to the students who don’t email me. I might send a reply to a question to the entire course.

I always schedule time during a class session prior to the exam so that students can ask me questions about any of the review items. Sometimes this lasts just a few minutes, and sometimes the bulk of the class period. (I do not hold separate reviews outside regular class hours, as I’ve mentioned before.) Usually when students email me a question, I ask them to save it for class, so that everyone can benefit from their question. But most of the review session is me saying, “I’m not going to reteach that entire lesson, but this is the nutshell version.”

If we give an advantage to students who correctly guess what will be on the exam, we are helping those who are better mind readers. These are the students with a social and cultural background most similar to the professor. If we’re concerned about culturally inclusive teaching — and being fair to students who have backgrounds different than their professors and the students from overrepresented groups — then a fair playing field means making sure nobody has to guess what topics should be studied more than others. I’ve noticed that middle class white students are more able to anticipate what middle class white professors are going to put on their exams. I’m not aware if there is any scholarly research on this, and I haven’t been able to find papers on this topic while doing a quick search. But I have decided that this is a potential source of bias in my teaching that I need to avoid.

The better I construct the exam prep information, the more time students can use inquiry to dig into course content, which is where the real learning takes place after all.

[note: This is a repost, modestly edited, from one from 4 years ago.]

*I look back at a few elements of this exam, and slightly wince, realizing that if I were to administer this exam now, there are several changes I would have made. I don’t think it’s the worst ever, but it’s interesting to see how I’ve evolved.