Cheating is a major concern when it comes to remote exams. Many people believe that a lot of $$$ can solve this problem, invested mostly in surveillance technology and requisite massive bandwidth, and that it’s the only way to go. But is this the only option or is it the best option in all circumstances? For some very high stakes certification exams and practice that prepares students for these, that probably is the case, but for many lower-stakes quizzes, and even some exams (lower stakes midterms), there are simpler ways worth exploring, that don’t necessarily make students feel like potential cheaters that someone needs to discipline and punish, locked in a surreal exam pandemicoptikon …
Here are a few strategies that will minimize the chances of students cheating, and a fantastic review technique I owe to one of my most memorable grad school professors at UW-Madison, a few decades ago… While this approach does not make cheating impossible, it deliberately imposes reduces the chances by using a few psychological tricks, and imposing a high “price” on cheating, by indirectly turning any cheating attempt into a learning activity. Here is how you could set up your graded exam (or quiz):
- Divide the quiz into sections by major areas or topics covered. For each section, write a few more questions that will actually be used: for example, if you are asking 10 questions on respiratory system, write 15 good multiple-choice questions. Most modern Learning Management Systems (LMS) will let you pull a random 10 out of 15 questions for a section of your exam. This way, there is a good chance different students will get slightly different sets of questions, making consults like “What did you answer on this question about….” more difficult. [NOTE: post-publish date – a faculty with whom I had a conversation on this very topic, based on actual experience from this summer’s class they taught, wisely commented that you have to be very careful to “equalize” the difficulty of questions within each set, so that students do not feel that some of them got easier random questions, and some got tougher ones…]
- Randomize the order of questions, and the order of answers within each question, if possible – both. This technique, combined with the one above, makes each two exams so noticeably different, that matching a question from one to the other is a very hard and time-consuming work… and a good course material review (although somewhat belated if done during the exam)!
- If your LMS allows it, set the quiz to show one question at a time, with no option to go back and change the answer once the question is submitted. Don’t feel bad about it: research shows that most students who go back and change their mind, actually wind up choosing the incorrect answer – apparently overthinking it, when you are not very confident about something, does not really help… This is also how most professional certification exams are set up, so students who will need to take one of these later (NCLEX, BAR, CPA, software certifications), will benefit in more than one way… [NOTE: post-publish date – another faculty pointed out that in some LMSs single-question-display, recommended here, makes navigation difficult – previewing the exam or quiz, to verify that what students will actually see is acceptable should be a requisite step in the process.]
- Limit the time, but allow a time window. Set a reasonable time for the exam to be completed (for example, 45 minutes – some, more advanced LMSs, let you see the average time taken by students on a particular exam, so after offering it once, you may want to check this, and tweak the time, as needed), but allow students to make this any 45 minutes between 7:00 am, and 7:00 PM, for instance. This will let students with different schedules, home situations, network stability, and work commitments to take the exam without the extra stress of making extraordinary arrangements… And it makes it more equitable for those whose life circumstances impose complex, stressful, less flexible schedules in their lives outside of the class bubble… You may think that this may increase cheating, instead of reducing it, but there is a method in this madness, a way to eat the cake and have it too: make sure to explicitly remind students in a text they see right before starting the exam, that you know that with more time flexibility they might be tempted to take advantage of it and cheat, and that you trust them to do the right thing. Activating their “moral responsibility code,” will reduce the chances of cheating; having them sign some sort of “honor code,” will help even more (see Dan Arielly eloquently explain why). [NOTE: post-publish date – not all specialists on T&L are in favor of limiting exam time at all – see recent 2020 article by UW-Madison’s M. A. Gernsbacher et al. for a counterpoint view.)
- Finally, and this is the best part of it, give your students a chance to learn BEFORE the exam. This very smart method, essentially a very clever psychological “cheat” in its own right, was used in a Non-Verbal Communication class I took decades ago in grad school with Professor Mary Ann Fitzpatrick. Prof Fitzpatrick would give all of us (it was a large 100+ class of undergrads and grad students) in advance of all major exams a complete set of questions (I think about 80-100) that she would later use on her exams; well, to be exact she didn’t give us questions – she gave us all the answers to her multiple choice exams, with questions removed… I had never seen anyone do this before, or after; I have since used this “trick” in some of my Spanish art, film, and lit courses I had taught over the years, always with the due credit given, because it is an absolute genius move. Try it once, and you’ll see what I mean: I had never seen undergrads study so hard for an exam before… Combine it with the Dan Arielly method (bullet point above), and a few very simple LMS techniques listed after the first three bullet points on this list, and with all that, if a student cheats – let them – it is their loss, after all…