Can There Be a Cheat-Proof Exam?

Cheating is not fun for anyone, except perhaps for the student who does not get caught. At my university I have only one class I teach for non-majors, that is students from around campus who are not majoring in anthropology. It is a class in the anthropology of Tibet, and is a large, lecture class consisting of 150-250 students in any given semester. Each week I lecture twice for fifty minutes, and the students have weekly “recitation” sessions of roughly twenty-five students each where they collectively discuss that week’s readings and lectures with a graduate student Teaching Assistant (TA). This is a classic course model for a large, public university in the USA. It is a course I resisted teaching for years—so many students, I thought. How could I ever teach about Tibet to such a huge audience?—but one I have now taught five times since 2008, and that I have come to love. There is a thrilling combination of reaching an audience for whom this is likely to be the first and only class in anthropology or about Tibet for the great majority of the students. I like to think of the students taking lessons from anthropology back with them to their home majors, whether it is biology or business or neuroscience or journalism. I [optimistically] like to think of them rethinking aspects of their studies, or the world around them, with the introduction to anthropology via Tibet they have received. Of the many things I like about the course, there is one thing I do not: it is the only course in which I catch students cheating.

My initial explanation for this was due to the fact that the course was mostly non-majors. Anthropology students were more committed, less likely to cheat, I thought. Non-majors took the course as a novelty, it seemed, thinking it would be interesting but easy. Some were not amused when they found out it was not easy but actually required attending lecture and recitation, and reading, and thinking. Other students loved the class, and over the years, a number have changed their majors to anthropology after taking this class (and other ones my colleagues teach similar to this—“gateway” classes, we sometimes call them). These students, and the overwhelming majority of the students did not cheat, but instead enjoyed a semester devoted to a topic often radically different than that what they usually studied. For some students it was the only time they had written papers in their college career. Others had no idea how to study for the exam. “Its all stories,” they would say. “And do we need to know the theories?” Exactly, and yes. Welcome to anthropology.

Yesterday, a Twitter thread on cheating helped me realize that my non-major explanation for why students cheated in this class was not the only reason. As I pondered cheating on exams, I realized I did not have issues with cheating in the other large lecture class I regularly teach (and none at all in the small, seminar-style classes for junior and senior anthropology majors where all assignments are essays or research papers). This other lecture class is Introduction to Cultural Anthropology, designed specifically for anthropology majors. As titled, the course introduces them to anthropological concepts, theories, and the history of cultural anthropology. On any given semester, this usually means 100-150 students who are eager and excited to be there and to learn. But the difference between the two classes is not just that one class is for non-majors and one is for majors. The other key difference is that I use radically different exam formats in each class. It turns out one style lends itself to cheating, while the other does not.

In my Tibet lecture class, students study for the exam based on their notes and readings. They know the categories of questions that will be on the exam (e.g., multiple choice, Tibetan vocabulary, essays, etc.), and the course material to study (e.g., everything from the mid-term to the last day of class). But they do not know the content of the actual exam until they sit down to take it. This is the sort of exam I regularly took as an undergraduate, and proctored as a graduate student myself. Until, that is, I learned of another exam style, and convinced the professor for whom I was teaching to test it out. My interest in this exam style was pedagogical in that I thought it was a better way to test students’ knowledge. It was a fairer test, and one that asked for learning rather than memorization. I first used this test style as a TA at the University of Michigan in 1995 in a huge (550 student) introductory lecture course. It worked beautifully then, and I have been using it ever since. Here it is:

One week before the exam date, I hand out the exam to the students in the class. It consists of twenty-five multiple-choice questions. I tell the students that they need to determine the correct answer to each of the twenty-five questions, and be able to write an essay explaining why the answer they chose is correct (and the other answers wrong). In essence, they have to “teach” the material back to me and their TAs. They are welcome to study in groups or alone, whatever suits their individual preference, although I recommend working in groups to talk through their ideas and arguments.

On exam day, the students receive the exam but now the twenty-five multiple-choice questions are in scrambled order from the study exam they received the week before. Together with the course TAs, I have also chosen four of the twenty-five multiple-choice questions as essay questions. Of the four possible essays, the students must choose three. They are required to write essays for each of the three questions explaining and defending the answer they chose as correct. The multiple-choice section is worth 25 points, and each essay is worth 25 points for a total of 100.

This exam format requires a rethinking of the multiple-choice question. You need to imagine anew the multiple-choice question as an essay question in disguise. What were the ideas you most wanted students to learn in the class? What questions will give students the best chance to show the real substance of what they have learned? Here are some of the multiple-choice questions I’ve used in the past:

1. Laura Bohannan’s fieldwork story about explaining Hamlet to the Tiv (“Shakespeare in the Bush”) illustrated what anthropological concern?:

          (a) the relevance of grid-group classification

           (b) the difficulties of cultural translation

          (c) the difference between social organization and social structure

          (d) the universal importance of liminality in rites of passage

          (e) the move away from the ethnographic present

If writing an essay on this question, students would have to identify (b) as the correct answer, explain the concept of cultural translation, and why this was Bohannan’s chief concern in “Shakespeare in the Bush.” They would next need to say why the other answers were incorrect. For example, they might explain (d) in relation to the rise of symbolic anthropology and the writings of Victor Turner they had read for class, and thus as key to 1960s anthropology but not the point of the Bohannan article.

2. Using Arjun Appadurai’s outline of global cultural flows, which “scape” provides the main backdrop for the Miss Tibet pageant?

          (a) ethnoscape

          (b) technoscape

          (c) finanscape

         (d) ideoscape

         (e) mediascape

Here students would need to explain the correct “-scape” in the context of globalization and the Miss Tibet beauty pageant, while also explaining each of the other scapes, and why they were not the main issue involved in the pageant.

3. Which of the following statements about gender is true?

          (a) Gender is represented and constructed in the same way across cultures

          (b) A linear relationship exists between sex and gender

          (c) Gender is biologically based

          (d) Gender systems vary across cultures

          (e) Gender can only be analyzed using feminist anthropology

For this question, students would need to identify and explain the correct answer (d), while also successfully explaining why the remaining answers are not true. We would expect them to use ethnographic or theoretical materials from course lectures and readings to illustrate their points.

4. An anthropologist comes to Boulder to study “Hipster Farming: Class and Privilege in Boulder’s Locavore Movement.” Based on the title of her project, what theory is likely to be the main one she is using?

          (a) Feminism

          (b) Poststructuralism

          (c) Cultural ecology

          (d) Marxist anthropology

          (e) Practice Theory

Working with theory is a key learning goal in my Introduction to Cultural Anthropology lecture course. This question requires the students to teach back to us each of these five theories, as well as choosing which one best fits a project on class and privilege. I teach them that contemporary cultural anthropology is theoretically plural. That is, we do not (usually)  use single theoretical models in our research and writing (as we did in earlier eras in the discipline), hence the need to underline the word “main” in the question. In an essay on a question like this, students are often able to give fairly sophisticated takes on how many of these theories could be relevant to the research along with identifying the main theory.

5. Which statement best matches what a Structuralist would say?

          (a) “Each part of culture matches a universal human need”

           (b) “Performance is part of conforming to normative ideas and behaviors”

          (c) “Myths are evidence that there are universal binary divisions in human thought”

          (d) “Gender is a social construction”

          (e) “A search for meaning rather than law is best done through interpretation”

Another theory question, this time working in the opposite direction: identify the theory from the claim made. They know one is structuralism, but all other theories they must come up with, match, and explain on their own.

Not all of the multiple-choice questions translate to essay questions. Here are three that require knowledge of concepts and texts, but that I would not use as one of the essay options:

6. Among the Awlad ‘Ali, a patron/client relationship gives what to each participant?

           (a) Honor/dignity

           (b) Gender/sexuality

           (c) Wealth/independence

          (d) Land/cattle

          (e) A wife/a daughter

7. Among the Azande, which of the following events would witchcraft explain?

           (a) A balcony collapsing while people were standing on it

          (b) A double rainbow

          (c) Earning a bad grade on a test for which you did not study

         (d) The exchange of cattle between age-mates

         (e) A coconut falling from a tree while you are a safe distance away

 8. Which system of marriage would best describe the system in which a woman can marry more than one husband who must be part of another group?

         (a) Endogamous polygamy

         (b) Endogamous monogamy

         (c) Exogamous polygamy

        (d) Exogamous polyandry

        (e) Exogamous polygyny

Students tell me they are often able to divide the questions into three groups—definitely a candidate for an essay question, definitely not a candidate, and maybe a candidate. This also helps them focus and refine their studying and knowledge.

That is the exam. I’ve used it successfully in my Introduction to Cultural Anthropology lecture course for two decades. Students find it fair, and like having the exam ahead of time. They tell me it gives them a sense of control, and an ability to focus their learning on what matters, and to make sure they know the material. Knowing to teach rather than to regurgitate is a different type of knowing. Pedagogically, I find this exam structure invaluable, but I now realize it also discourages cheating. Students have the exam ahead of time, and by the time the exam date arrives should have determined the correct answer to each of the multiple-choice questions. This, combined with the fact that is a course of motivated students who are mostly anthropology majors, leads to zero instances of (caught) cheating using this exam format in my class.

But, is it cheat proof?

It is as close as I have gotten. In the context of a room full of 100+ students taking an exam with paper and pen at the introductory level, I think this is pretty good. Pedagogically in terms of testing the students on the ideas I most want them to learn and take from the class, it is excellent. And in terms of dissuading cheating, it certainly seems to work.

I want to note that I did not come up with this exam structure. I don’t know who did. I found the original idea for this format in a box of teaching ideas kept in the TA lounge of the University of Michigan anthropology department. Many of us loved to talk about ideas for our classes and assignments and exams, as well as share hard copies of handouts we used. One day I discovered this exam format on a handout in the box. It was so long ago—twenty-two years ago—that I no longer have the original write-up of the exam format, nor do I recall where it was from other than perhaps a university in Ohio. If anyone reading this knows who came up with this exam format, please do let me know. I would love to thank them.

If anyone else out there has a great exam format for a large, lecture class, let’s hear it. Sharing ideas for teaching is always welcome here.


Carole McGranahan

I am an anthropologist and historian of Tibet, and a professor at the University of Colorado. I conduct research, write, lecture, and teach. At any given time, I am probably working on one of the following projects: Tibet, British empire, and the Pangdatsang family; the CIA as an ethnographic subject; contemporary US empire; the ongoing self-immolations in Tibet; the Chushi Gangdrug resistance army; refugee citizenship in the Tibetan diaspora (Canada, India, Nepal, USA); and, anthropology as theoretical storytelling.

6 thoughts on “Can There Be a Cheat-Proof Exam?

  1. Carole: I love the format of the exam you describe. You discuss it in terms of cheating, but I wonder if/how you contend with another issue it presents. In a large intro course, the regular dilemma is that grading is seductively easy with ‘objective’ exams — multiple choice, true-false, fill-in-the-blank — while essays, even short essays, are time consuming to grade, especially (I confess) at the end of the semester when I’m dashing to get everything done in time to leave town. I do not let my TAs grade essays, feeling that I have an obligation to students to evaluate their more creative work myself — and to apply a single, consistent set of standards to all students — but grading 300 essays, even short essays, from 100 students might be beyond my patience! Have you used TAs to grade essays on an exam like this? Has that worked well?

  2. Different pedagogical styles, I guess. I learned to grade essays and essay exams as a grad student TA myself, and consider my grad student TAs fully competent for the task. In my estimation, this is also part of my job as professor supervising grad student TAs—teaching them how to teach, to craft assignments and exams, and to evaluate/grade. Some of this work is more hands-on than others, some is facilitating and supporting, and some is creating opportunities for them, but all in the name of teaching our undergrads well (and enjoying it in the process). In any given semester, my grad students might be anywhere from the MA level to ABD PhD candidates who are finishing their dissertations. If I did have a TA who I deemed not competent for the job, I would have them replaced with another grad student TA; in my seventeen years of teaching, this has happened only once.

    However, I also think that this particular exam lends itself to essays that are more structured and focused, so per your concerns about the time involved, might be “easier” to grade.

    One other thing: I can’t imagine giving exams in anthropology that do not involve writing. Not everyone feels this way. My original impetus as a grad student for instituting this exam involved a strong dislike of the multiple choice format which the course for which I was TAing used. The entire exam was fifty multiple choice questions. I didn’t feel it fit the sort of anthropological learning I wanted the students to be doing, but you’re right, had to convince my fellow grad student TAs and professor that it was worth it to give ourselves the extra work of grading exams rather than just feeding fill-in-the-bubble Scantron sheets through some university computer. Luckily for me, the other grad student TAs and professors were willing to test it out. As professor, my grad student TAs usually teach for several different professors over the course of their TA-ing career so get to see different styles of lecturing as well as exams. All of this is great for developing their own sense of pedagogy, some of which is personal style.

    Finally, I didn’t write this in the original essay, but both of these lecture classes involve three-four essay assignments in addition to a midterm and final exam. The TAs grade the essay assignments as well. And do a great job at it! So, definitely something to consider.

  3. Carole: Thanks for your rely. As I noted, I have always been reluctant to have my teaching assistants grade student essays, and this reluctance springs from several concerns, including the fact that — in our introductory courses — TAs are often archaeology or biological anthropology students, and even some of our cultural graduate students have come from undergraduate programs in arts and humanities, and are themselves still learning some of the basics of anthropology. Perhaps I am overly cautious about grading fairly and consistently from student to student, which is difficult with multiple TAs. When I have done this in the past, I seem to end every semester with the TAs gone for the winter break or for the summer, and a line of unhappy students at my office wondering why they got a B when their essays were just as good as those of a friend who got an A, and I have to work out what TA Bob was thinking and what TA Mary was thinking. However, I applaud your willingness to use these opportunities to mentor TAs, and your stamina!

    That said, I do use essay type exams whenever possible, which I find give students more opportunity to explore ideas and to be creative — one of the best “essays” I ever received was a long poem, a brilliant piece of work by an English/Creative Writing student who went on to become a well-known author. With up to about 40 students in a course, I typically pass out a list of 5 or 6 essay topics a week before the exam, and tell students that on the day of the exam I will pick 3 or 4 of those topics and will ask students to write on 2. They are free to write all 5 over the week and pick the appropriate 2 on exam day, but my standards will be high. This has worked very effectively — it doesn’t focus on what students can memorize for an exam, but their skills at researching a topic and synthesizing information into a coherent essay. Virtually all students write several substantial essays over the week, which means that I lure them into working on several key topics in some depth in essays that they will not even submit. Students seem to like this approach, which maximizes their control over the material, and even though this is one of those cases in which I end up reading as many as 80 essays for each exam, I find this manageable.

  4. I love the foundational idea here but am a little concerned with how the execution (MC and 3xEssays) would play out. What I read from Barbara’s comment is a concern about how long grading three essays can take. It usually takes longer than I have ever been assigned to do it, and so I’m weary to assign more than 1 essay/per exam out of concern for my TA’s time.
    To carefully and thoughtfully grade three essays takes about 45 minutes. My concern would be TAs feeling so rushed that they glaze over. I admit it would be hard for me to grade dozens of 3-essay exams. I’m also wondering how many are assigned to each TA?

    And by “cheat,” do you mean just looking at other students’ exams during the exam? Because I’m not certain this mode would prevent other forms of cheating, like reading off hidden notes (esp. if essays prepared in advance) and multiple students writing near-identical essays with one brainchild.

    I’ll repeat though, I love the idea of choosing, defending, explaining the MC answers. Brilliant.

  5. I am late to the conversation but wanted to say thank you for this. I think I will stick to my all multiple-choice for intro, but really looking forward to trying this twist on an essay exam out with my new mid-level course next fall. My class sizes are thirty to fifty studentsish, no TA. I do “fieldwork”-based essay assignments but haven’t had my students (suburban community college, very diverse on all axes) do any in class exam writing yet…

Comments are closed.