28

I'm facing a situation at my university that raises some ethical concerns and I'd like to get your perspective on it. One of our professors has a consistent practice of reusing the same exam questions every year for their course. The questions remain unchanged from year to year. Additionally, this professor does not share either the questions or the solutions with students.

My concerns are twofold. Firstly, this approach could potentially benefit students who have access to previous years' exams through their peers, creating an uneven playing field. Secondly, the lack of transparency and the repetitive nature of the exams could impact the overall learning experience and assessment integrity.

I am unsure about the ethical implications of this practice in an academic setting. Is this approach considered acceptable in academia? Are there ethical standards or guidelines that address such practices?

13
  • 8
    If it is broadly known, which it seems to be, why does anyone have difficulty in finding the information?
    – Jon Custer
    Commented Jan 16 at 19:34
  • 4
    Why would the class I took this term being the same as the class taught this term matter? Commented Jan 16 at 23:08
  • 6
    Frankly, it is obviously unfair (to some extent), but how exactly to justify that is really a matter of your own ethical principles, and not academic norms. It is also common to find professors who requires students to buy rather expensive textbooks written by themselves. Obviously unfair too, but common too.
    – user21820
    Commented Jan 17 at 4:58
  • 6
    It isn't unethical unless the school has a rule against it. It's sloppy. It's probably unwise. But it is not inherently evil.
    – keshlam
    Commented Jan 17 at 22:26
  • 2
    @JonCuster At some universities, only wealthy students will benefit. For example, expensive Fraternities and Sororities may keep copies of prior exams for their paying members, and not make them available to any other students (or make them available only to other students paying to be part of the pay-to-play "Greek" system). Commented Jan 18 at 9:00

10 Answers 10

38

The ethical concern is that a professor needs to provide an environment in which every student can learn. This practice, which I consider foolish, isn't necessarily contradictory to that ethical concern. We are supposed to be educators, not graders.

The fairness concern may be misplaced. Certainly students will be aware of this (most students). Old exam questions might be a good way for students to review what has been given otherwise. Some student organizations (in the US) have kept files of old exams along with answers for (more than?) a hundred years. They may give some students a grading advantage, and they may give some students a learning advantage. The latter is fine. But whether the overall learning environment is proper and fair is beyond such a narrow concern.

It is considered "acceptable" in the sense that it is common enough and that there are few rules that prevent it. I don't adhere to that practice, not condone it, but look deeper. I'd recommend other practices, but, at base, it is the learning that is the important thing.


Personal note: When I was a student (previous century) there were student organizations that kept files of old questions along with (correct/acceptable) answers. These were open to members, which I was not. I never considered this to be a problem as I was willing to work hard to learn things. If someone else got a good grade not working hard, it didn't affect me if I also got a good grade.

It is one reason, however, that I've become an opponent of grading "on the curve" which is, in effect, making the grades of one student dependent in some (any) way on the grades of another. Just as we need to be educators, we need to judge students individually as needed. I told students at the beginning of important courses that everyone could earn an A and that everyone could earn an F. And, no, not everyone wanted to hear that.

8
  • 16
    When&where I was a student (this century, central Europe), these collections were open to any student and professors actually encouraged updating these collections since they reflect the actual course contents. I may add that we did not have courses that really followed a given textbook, we had courses and book recommendations. Commented Jan 16 at 22:38
  • 3
    At the university I went to, all past examination papers were available in the library. (They probably had to be by law -- the university library was a copyright library, so if the exam papers were copyrighted, this follows. )
    – nigel222
    Commented Jan 17 at 9:56
  • 1
    @nigel222 in general for the last 50 years or so (worldwide, earlier for many countries) if something is written down and moderately novel (i.e. more than a list of things) it's probably under copyright (or explicitly gone into the public domain. Most of the versions of legal deposit I'm aware of are based on publication.
    – origimbo
    Commented Jan 17 at 13:43
  • 4
    While such an archive might not have effected you, it has an impact for students where a major reason they are seeking a degree is to obtain employment. Many companies have strict GPA cutoffs for internships, and when comparing new hires the difference in GPA between two candidates from the same school is something that is looked at.
    – Chuu
    Commented Jan 17 at 15:18
  • 3
    "We are supposed to be educators, not graders." Supposed by whom? Most people (e.g. future employers) certainly expect a person with a university degree to have passed some kind of scrutiny on what they have learned, not just have been in an environment in which they can learn for N years.
    – JiK
    Commented Jan 18 at 11:23
24

I had one course where the same (pool of questions) was used in a way that I think fair. I may add that the course was considered rather hard to pass.

The teacher openly announced that they'd use the very same question pool (which I guess did not have more than 5 variants of each question) they had used for some decades, and that old exams are available.

The course had a large number of exams (> 10 in 1 semester) and they said the idea is "to ask everything" (to rule out good/bad luck with the subtopics), and that even if someone would try to "just" memorize the questions and their answers, there's a sufficient number of questions in the pool that one really cannot help learning the principles behind the answers as a side effect also that way.

I guess their real purpose was less to grade us but rather to make us work through a sufficient number of practice questions (exercises) in order to learn.


Not sharing solutions with students would not really be possible where I am since students have a right to see their graded exams and they can ask questions why exactly they lost points - at which point the solution will be revealed.

4
  • 5
    This is a brilliant strategy. Trick 'cheaters' into learning.
    – seldon
    Commented Jan 17 at 14:20
  • @seldon: we actually had a lot of professors who rather than ban on certain types of "cheating" used exam modes where those cheating strategies don't help: open book exams (with the recommendation to prepare cheat sheets as making those summaries is very good for learning), exams like the described, an excercise course where excercises were done in groups during course time, and if the average grade across all groups for the whole semester was sufficiently good (high bar), no separate exam needed. Lots of oral exams. A/B series of exam questions (no use trying to look at neighbour's solutions) Commented Jan 17 at 21:09
  • 1
    Trying to forbid students to speak about the exam questions they met was something that never crossed my mind as something that could possibly be done... Even with oral exams, if the candidate approved, other people could attend and listen to the exam. Some professors actually recommended to listen as part of the preparation of the important oral exams, or to bring a friend if you thought it would help against exam anxiety. (But as I remember, that rarely actually happened) Commented Jan 17 at 21:15
  • 3
    Continuing on the theme of preempting cheating by making it not effective... I definitely have colleagues that just tell everyone what the exam questions will be ahead of time. One even did a controlled trial over several years where some years they told the students and other years they didn't - it made no difference to average grades. Commented Jan 18 at 12:38
8

I'm going to offer a different perspective here, offering that how negligent the practice is depends highly on the institution, classroom size, grading practices, and course content, along with a host of other factors

The main factor in determining how problematic this is is the probability that there will be students who both want to and have the ability to cheat on the exams. The relevant factors in this are:

  1. Classroom Size - Larger classrooms will naturally have a larger likelihood of someone recording the questions and sharing it with future students.

  2. Cultural Background - In some cultures, cheating is more accepted and/or widespread. This is especially true of cultures that place a heavy emphasis on examination. More of these types of students in your classroom will also increase the likelihood of questions being recorded in passed on.

  3. University Selectivity - In general, anecdotally, it seems that more selective universities have more cheating. I would guess that this is because students are more likely to cheat their way to a good GPA in high school, because they are more likely to have merit-based scholarships which require a certain GPA, and because the average student is more likely to care about their grade than, for example, a student in a regional state college with an over 90% acceptance rate.

  4. Grading Structure - Does the course use a "traditional" grading system? Or an alternative one? If you have a course that, for example, uses a mastery system, where students can continue to retry a topic until they pass, then they are less incentivized to cheat, since if they fail, they can just try again.

  5. Course Difficulty - What is the average student grade in the course? If its an "easy A" course, students are less likely to feel the need to cheat, but if its a "filter" course for a major where over a third of students routinely fail, students are much more likely to feel pressured to cheat.

  6. Grade Importance - How vital is getting a good grade in the course to the student? If it is a gen-ed class, then the answer is probably pretty low, beyond keeping the grade high enough for merit-based scholarships. If it is a course where a good grade is required to declare their major, then the incentive to cheat is going to be high.

  7. Future Relevance - If a class imparts a skillset which is going to be particularly relevant for later coursework, and for after college (e.g. on a job search), and the instructor actively ensures students know this, then students are more likely to view learning the skillset as important (and thus not cheat).

  8. Attitudes Towards Professor - When students particularly respect and like a professor, they are less likely to cheat. Similarly, if they detest and don't respect a professor, then they are more likely to view cheating as acceptable. As an anecdote, I remember in my undergrad a student using Chegg to cheat on some homework for a disliked professor, while this student also mentioned how they would never cheat for a specific other professor's class, who was known for being kind, passionate about teaching, and a good educator in general (it is worth noting despite these qualities the professor did not give easy As).

  9. Cheating Punishment - When professors use systems to identify cheating, and institutions take it seriously and exact strong punishments on students, students are less likely to cheat than when professors don't pay attention and where students only get a slap on the wrist when caught.

  10. Institutional Culture - Related to all of the above, if students generally recognize that cheating is rampant, then they are more likely to join in.

The second factor is, upon questions being shared, how easy is it for students to cheat? The main factors for this are:

  1. Number of Questions - If you have a database of, let's say, 500 questions, and each student will be asked only 20 questions a year, than memorizing each answer is much more difficult and less rewarding.

  2. Answer Involvement - As Oбжорoв's answer indicated, if your questions are, let's say, an essay question, it is much more difficult to cheat on it than, let's say, multiple choice.

  3. Exam Format - An unproctored exam is much easy to cheat on than a proctored one, and a digital exam is easier to cheat on than one with a paper and pencil.

Finally, I suppose one could also consider how "equal opportunity" the cheating will be. If it is likely answers will be posted online and distributed to the entire class, it's probably less discriminatory than a group sharing it among themselves, though I would think this is only a little solace for an already bad scenario. Beyond that, it's difficult to tell what would lead to more "public" cheating, beyond large class sizes and more homogenous students.

So, putting this altogether with two scenarios:

Most Problematic - A professor is teaching a very large course (i.e.: 100's to 1000's of students) where at least a B is required to declare some (popular) major. The institution is an extremely selective university, and many of the students in the class come from cultures where cheating is more acceptable. In fact, most of them are close friends with people that took the class last year with similar descriptions. There are also students, particularly those from poorer backgrounds, that don't have any network within the university. The university famously does nothing to cheaters, and even students caught several times continue to be enrolled in the institution. Despite the course grade importance, the content is of little relevance for students after passing the class. The professor is a jerk and is hated for it. Its also a very difficult class, with more students failing than passing. Exams only have a single opportunity, and a poor performance cannot be made up for. Despite the difficulty, exams are multiple choice and taken online at home alone, and the exact same questions are used every year, for every student, with absolutely no variation.

Least Problematic - A professor is teaching a course at a public regional university with very high acceptance rates. The classroom is very small with less than 10 students, who are all from similar local backgrounds. The instructor is beloved and respected by the students, as both a great educator and a kind soul. The instructor uses a grading system which lets students in some way make up failed exams, and almost everyone gets an A in this course. The professor and institution take cheating seriously, with first offenses resulting in failing the course and second offenses resulting in expulsion. The course grade isn't relevant for their futures, but the content of the course is of tremendous significance, and the students know this. The professor has a large database they pull from, but may be repeated in multiple years by chance. These questions are challenging, involving writing an essay, writing a mathematical proof, or writing a computer program. Exams are always proctored in person and done via paper and pencil.

Where the cutoff for being "unethical" between the least problematic and most problematic scenario is depends on your own sense of metaethics.

I will note that, ironically, the most "valuable" institutions to have on one's résumé are much more likely to have courses which sound like that problematic scenario, while the least "valuable" ones are going to sound like the environment where cheating is rare. If I can get on a pedestal for a moment, the fact that generally R1s systematically disincentivize their tenure track professors from caring about teaching in general, have giant course sizes that depersonalizes everything, and that many R1s refuse to hire teaching-track professors and give them salary, job security, and college governance parity with their research-oriented counterparts, is a large reason for many of the issues in higher ed, including this cheating one.

1
  • 2
    In a difficult Biology course that was also required for pre-meds (in the days when med schools were very hard to get into), during an exam, a student in the row next to me introduced himself to me, followed by, "What's the answer to number 3?" I later found out that his nickname was Cheatin' Ed. Can one imagine that this had no effect on the patient care he gave? Commented Jan 18 at 13:38
4

I agree that all students should have equal access to past exams, and if I were in charge of your program, I would hope students would bring their concerns to me.

I would never repeat the exact same exam every year, but how bad it is from a pedagogical perspective answer depends somewhat on the subject matter, the type of questions, and whether the exam is written or oral.

For example, there is old joke that some economics professors "don't change the questions, they just change the answers". I have heard recent versions of this joke about professors in fast moving fields (e.g. computer science, biotech, …) doing this since the answers this year may actually be somewhat different from last year.

More seriously, occasionally a final exam is essentially a project presentation. Here is an example of a political science course whose final exam was always a single question: "Propose a solution to a major social problem. The answer must incorporate political ideology, economic impact, and media representation". In that case, however, it is best to use an oral exam format where the examiner(s) can pose follow-up questions based on what is presented, so even though the main question is always the same, the follow-up questions are not identical and can quickly establish how well the student understands what they have just presented.

4

Existing answers are good, but I think they omit one relevant point... Having some of the questions on an exam identical to questions that have been repeatedly used in previous years can be extremely valuable as a way of calibrating the difficulty of the other, newly-written questions, to achieve a consistent, criterion-referenced grading standard from year to year. In that sense, reusing questions is not just ethically acceptable, but ethically beneficial.

3

My feeling is it's totally wrong and unethical.

Some students are better at memorization than others, and could just regurgitate verbatim an answer composed at leisure before the exam. This, even if the pool of possible questions is quite large. Some would get their prepared answers checked over by another person and incorporate their corrections and suggestions. Some would go even further, and pay someone else to write the answers for them. That would be cheating, but only if the person writing the answer for them plagiarised a known text could that be proved.

The only way the examination system can be fair is if none of the students have ever seen the question before the exam. (Or, they could be marked on coursework, but there is then a risk of plagiarism and other cheating as above).

13
  • 4
    I'd be less concerned about "fairness of grading" and more concerned about "learning opportunity". Practicing past exams is usually a great way to learn. But if there is only one past exam and the teacher has to make it unavailable to the students because they plan on reusing it, then the teacher is actively removing a studying opportunity from the students.
    – Stef
    Commented Jan 17 at 10:37
  • 3
    @Abion47: Cheating is not a victimless offense. It diminishes the value of the degree / harms the reputation of the school when half its graduates show up to interviews knowing nothing at all.
    – Ben Voigt
    Commented Jan 17 at 16:25
  • 1
    @IanSudbery You or I might think that an answer needs figuring out because it's (e.g.) an explanatory text or a verbal-reasoning argument or a first-principles algebraic derivation of considerable length, but there are some students (not many, but some) who have an extraordinary ability to learn by rote very long explanatory texts and verbal-reasoning arguments and first-principles algebraic derivations (which ends up getting revealed when they regurgitate them in answer to a completely different question). Commented Jan 18 at 16:45
  • 1
    @IanSudbery: The answer specifically says "Some students would ... regurgitate verbatim an answer composed at leisure before the exam. Some would get their prepared answers checked over by another person and incorporate their corrections and suggestions. Some would go even further, and pay someone else to write the answers for them." Or wherever they got a copy of the old exam (frat house files?) might have included correct answers with it. And today you can add "Generative AI" into the mix. In none of these cases is there any evidence the student came up with the answer.
    – Ben Voigt
    Commented Jan 18 at 17:07
  • 1
    @BenVoigt I don't see someone reguritating an answer they have composed at leisure as being particularly problematic. They have still demonstrated the skills they supposedly needed for the exam. Obviously paying someone else to compose the answer in its entirety is problematic, but I suspect this is rare enough (at least in my context) that it wouldn't bring down the value of the degree any more than is already the case with people paying for their dissertations (40% of their degree here) to be ghost written. Commented Jan 18 at 17:24
3

It is somewhat unethical, particularly if the professors know that there are students in the class who can't join the 'group' to get previous exam answers, for example, due to racial and language barriers.

This practice was common in Hong Kong 15 years ago. The worst aspects of it were reusing exams and grading on a curve. So, students who had access to previous years' answers could answer perfectly and all receive A's, while those who didn't have access would get C's or D's, even though they performed well in class and exam.

1
  • It is not unethical.
  • Profs often do it because it is a good question, and because they are busy, and because they have different priorities than students do.
  • Profs can do what they want... unfortunately.
  • Professors are experts in their field. Their field is usually not teaching. Few professors bother to take any teaching training. Though of course, some do and there are entire departments at universities devoted to helping professors teach.
  • Looking at past exams is a good study technique. So are reading, attending class, discussing with classmates, meeting with the professor, doing research online, and asking questions.
  • The "problem" is likely as old as education itself, and is likely not going away.

My background: Taught very large courses in-person for over 16 years and then got a Masters of Education.

3
  • "Few professors bother to take any teaching training." 10-15 years ago, many British universities had a policy that formal pedagogic training was compulsory for new faculty members, although I think a lot of them have subsequently retreated from that position. Commented Jan 18 at 12:16
  • 1
    @DanielHatton - Its still the case at my university and all the universities my friends work at. Commented Jan 18 at 16:02
  • 1
    @IanSudbery The university where I worked ditched the requirement c. 2018 as a cost-saving measure. I'm still in touch on Twitter with some of the faculty who used to deliver the training, and they all seem to have moved on to another nearby uni, I presume having taken voluntary redundancy. Commented Jan 18 at 16:34
1

I think "unethical" is an odd word to use here. It's arguably lazy and foolish, but I don't see how it would be "unethical".

One could argue that there's nothing wrong with this practice at all. A good test should presumably test the student's knowledge of the material taught in class. A reasonably intelligent and insightful student should be able to figure out what likely questions on the test are going to be. For example, if a history class spent a lot of time on the causes of the US Civil War, a student would reasonably expect there to be questions on the causes of the US Civil War. I wouldn't expect a test for this class to ask me to find the derivative of the sine function. Indeed, I've taken many classes where, before a major test, the teacher would flat out say, "There will be questions on the test about ..."

I suppose in some classes you could say, Yes, there would have to be questions about X, but you wouldn't know the details of the question. To make a trivial example, I'd expect a class in arithmetic to have addition problems on the test. But whether it asks for 17+29 or 32+97, you wouldn't know. And if you did know in advance, you can spend hours working out the answer if necessary and memorize it. (Of course I wouldn't expect a college class in arithmetic, but you get the idea I hope.)

I could readily see a professor saying, Any reasonable test for this class is going to have to ask these 20 things. I have to ask about all of them or I won't be covering all the material in the class, and I can't fairly ask about other things because those other things were not covered in class. Depending on the class, maybe you could construct questions that asked about the same subject in different ways. But in many classes this would be impractical. I think the wise and diligent professor in such a case would have a pool of questions, select some number from the pool each year, and present them in a different order. Then a student couldn't just study a previous year's test, memorize the answers, and come in and write them down. He'd have to at least understand enough to recognize which question was which. Preferably to have some understanding of what the question is asking, so he knows which answer to give. And at that point, that's basically what you're testing for.

0

Doesn't this depend on the field? Compare these three questions:

  1. What are the ethical consequences, if any, of slaughtering animals without sedation?
  2. What is the influence of Aristotle in Hegel's philosophy?
  3. State and prove the fundamental theorem of linear algebra.

These three questions probe different knowledge of the students, so they may be more or less appropriate.

1
  • why the downvotes? Commented Jan 19 at 13:04

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .