By Stacy Palen
We are just past finals here at Weber State, and we have been having a lot of discussions about how the transition to online learning went. Among those discussions is a big piece about student cheating. This was prompted by a faculty member who has taught the online astronomy course for a long time (27 times!) and has usually proctored closed-note exams. During the second half of this semester, those exams changed to open-note exams taken at home (presumably!) without a proctor. The average of student course scores rose 8%, and for the first time ever, no one earned a “D” or failed the class.
Clearly, this was not a controlled experiment. There are several possibilities for why student scores rose, which are not mutually exclusive:
- Students who are uncomfortable with going someplace new to take proctored exams were more comfortable at home.
- Students who are normally overwhelmed by a closed-book exam did better with an open-book exam.
- Students cheated with one another by sharing answers.
- Students cheated by looking things up online.
- Students used a “service,” such as “Take My Online Exam” or “Online Class Hero,” or something similar.
- Something else we haven’t thought of yet.
Figuring out what’s going on here, and why, will take more data, and probably some more experiments. We feel compelled to figure it out, because we want to maintain the integrity of the profession, and we want to help guide students to be better people.
Fighting against cheating can be draining. I recall a professor, for whom I was a TA in grad school, who carried out quite sophisticated statistical cross-correlation analyses of the in-class multiple-choice tests in order to catch people cheating on exams. He seemed to enjoy the challenge. I did not, and I found that spending so much mental effort on distrust really damaged my ability to find joy in my job. (Not to make it all about me…but I think students benefit when I’m full of joy, rather than furious.)
I take a different track. Even in my face-to-face classes, I give take-home, open-book, open-note, and written exams that students have several days to work on. I came to this solution by focusing really hard on what I actually want students to know or be able to do.
I don’t actually care if students can recall things; I care if they can figure out things. I also care to give them feedback about their reasoning. Consequently, I don’t give multiple-choice exams. All by itself, that makes cheating a lot harder. (Yes, it’s a gigantic pain to grade 120 final exams by hand. But it’s also a gigantic pain to run sophisticated statistical cross-correlation analyses, and change them every time the testing software changes.)
I don’t think students will ever not have Google (or something similar) at their fingertips, so it’s fine with me if they look things up. I write an exam that presumes that they actually do have Google, or the textbook, at their fingertips to look things up. Making this assumption lets me ask questions that are a lot harder to figure out, and therefore a lot harder to Google directly.
I do care that they “attend” class (for a certain pandemic value of “attend”), so on the exam, I ask several questions that are trivial if they’ve actually been in this class for this semester, but are impossible if they haven’t. (And I vary my lectures and materials accordingly.) This has the added benefit of automatically penalizing students using a test bank or an online service; they can’t get those points, and I don’t have to do anything special about it.
But the most important thing I do is try to make the exams personally compelling. I write a story for them, and then I drop them in it. For several years now, I’ve used a “zombie apocalypse” narrative. (In the post-COVID world, that might not be the best choice!) I used to drop them on a desert island.
In the future, I might try having them imagine they are teaching their kids about the sky, or they’ve been abducted by aliens, or that a time machine has transported them to the deep future. The common feature underlying all these scenarios is that they are on their own. I think the story matters; it makes the material feel relevant (even if they know zombies aren’t real), and it gives them an incentive to try to solve it themselves.
This semester, in the abruptly online experiment that we were all thrown into, I found that I had little to change about this practice. I had to think a little harder about the fraction of “attendance” questions that I wanted to ask, and what made for a fair question of this type. (I specifically referenced Astronomy in Action videos instead of our in-class lecture.) In my class, the average on the pre-COVID midterm was within a few points of the average on the post-COVID final, and the overall course average was just over 75%, which is where it usually is.
There are so many tools now, and so many different ways, to carve up a class into compartments that teach or test each content area, skill, or attitude; take a look at them, and figure out which ones will do the best job for the things you care about. Be sure to start by asking yourself this important question: What is it that you want them to know or be able to do? What are the deeper values you bring with you to the classroom?
Then, figure out how to design your assessments to reach those goals and teach those values. This may be overwhelming this semester, but as you look ahead to future semesters, you may find that you are changing a lot of things anyway. Changing the way you write your exams may save you the time and effort you currently spend on arranging proctors or catching cheaters, which will ultimately make it harder and less rewarding for students to cheat.