Should students have access to past exams or an exam bank?

Question

Something I never understood is how in so many undergrad classes North America, the students are allowed to have past exams that they are allowed to study over.

It never made sense to me because it is apparent that this practice could cause so many problems, especially as exams accumulate over the years, e.g. leading to inflated grades. I do not exclude myself from this practice.

As an example, in one of the classes, I knew only a small number of students had access to an exam around that was given around 5 years ago (literally gotten through one of the student's relatives who took the exams years before), and one of the question from that past exam was very similar to another one that was given in the actual exam. This doesn't seem very fair to me.

Finally, I always felt that my work was delegitimized with access to these resources, and over the year sort of adopted a pessimistic view on learning and knowledge, seeing that so many of my peers became arrogant and prideful despite having had access to these external resources. It was not until I did grad school much later on (where past exams are virtually non-existent) did I "let go" of those feelings and regain a bit of confidence in my own capabilities.

Should students have access to past exams or an exam bank that they can study? Why does it not constitute as cheating? My background is in STEM.

Answer 1

As a professor, one has to assume that all past exam questions are available to students. Some student organizations debrief members after exams and capture the questions for such an exam bank.

Some professors realize this and take action accordingly, not reusing questions, or not using enough of them to matter substantially. Others think, probably mistakenly, that they can set rules forbidding the use of such "resources". The rules will be broken by at least some of the students. I don't know of any effective way to actually prevent access to old questions. So, just assume that they are available.

My own opinion is that the professor should just assume that it will happen and, if they want to make the test taking "fair" will make old exams available for study. Some even go so far as to give out the final exam on the first day of class, to be taken later, though I suspect that not many do that.

But old exams are useful for honest students to study from if they don't expect to get those questions on their own exams.

Of course, coming up with new questions each term/year adds a bit of work, but if the course is so "stagnant" that an exam from a few years ago is still being reused then the prof needs to do a better job.

Answer 2

I think the point is, or should be, that having diagnostic question be "secret" is perverse. That is, why not tell people directly what we want them to know?

And why not let them see representative examples of what we want them to do?

Sure, some students are not so much interested in learning as in the grade... but should we corrupt the scholarly aspects by gaming merely to thwart this? I don't think so.

It is true that some of my own colleagues have "criticized" my composition of PhD Written Prelims by observing that there is no secret about the general scope of the questions on the exam, so, "gosh, people can study the stuff and then everyone passes!?!" Um, well, if you tell people what you want them to know, and if they can demonstrate that they know it... and can do things with it... what is the problem?

... do we need to have a significant number of "failures" to prove that "non-failing" is significant? I'd say "no", but quite a few people would say "yes". This becomes perverse when we have a stringent admissions system, so that really everyone admitted to our grad program is capable of doing a PhD. See what I mean?

As with sports: if one can do the thing well, good; if not, not-so-good. But to create fake/artificial strictures by withholding information is just silly. Not even an interesting game.

Answer 3

If it bothers you that a small group of students had access to previous exams then you should understand that a professor is giving all students a fair opportunity by posting previous exams. If a bad student with access to unauthorized past exams gets the same score as a good student with lecture notes only, then there is a problem. If both have access to past exams then the better student will prevail. The point is that cheaters wont stop looking for shortcuts so the best option is to remove that option by making them available to everyone, but write exams that require thinking not memorizing.

Also I find it unusual that you are in STEM, but feel pessimistic about learning. I am trained in engineering and we were given past exams and dedicted tutorial sessions to do practise problems yet only a handful of students got higher than an A and many got Cs. It's so easy to twist an old problem that can't be solved by memory.

Answer 4

Being in a STEM field, I am sure you're familiar with extremely difficult examinations such as the Putnam exam and International Olympiads in Math and Physics. These definitely are exams, even though they're called "competitions". They all share the feature of having open access to all past year questions - solutions, often with multiple methods, are also posted online with alacrity by various enthusiasts.

Why, then, isn't everyone doing fantastically well in these exams? The answer is that the exams are of a sufficient level of difficulty and also set with sufficient creativity and originality (to distinguish them from the previous years' questions) that it's not at all trivial to just practice old papers and hope to ace the new one within strict time controls.

And that's your answer - if your exam was challenging and original enough, you wouldn't have to worry about students having access to questions from previous years. In fact, it's almost imperative that you give them that access, so that they can be prepared for an exam of that difficulty.

Of course, "standard" exams are not designed to be as challenging as competition or olympiad problems, but the same general principle applies. There are also other considerations - students will often create and distribute their own informal past year question "banks" as hand-me-downs, maybe even sell them to their juniors. These practices will make it almost impossible for you to ensure an even playing field for your students since some will have more access than others to previous questions. So, it's actually more equitable for you to simply make all your previous exam questions open to all and just focus on setting a new exam that's interesting and challenging without obvious repetition of any of your old material.

Answer 5

Access to past (or practice) exams is good test design

Any test or experiment will have "noise". That is, beside the thing you want to test, there will always be other factors affecting your results. This is, of course, also true of exams.

As others have explained, there are nuances to any question that the examiner might not even be aware of. Local academic culture, implicit biases and assumptions, the level and scope of a specific course. A student hoping to do well on a test has to take all of these into account. However, these details are usually not what you're testing for. You want to know whether a student understood and memorized the subject matter, not if they remember which notation a specific professor prefers.

The more noise you introduce into an exam, the more effort is required from your students (if they want to do well) to mitigate its efffects. "Secret" tests are especially noisy because the student, by design, cannot know which aspects of the course material will become relevant. This includes things only tangentially related to the subject matter.

Did the professor choose axiom X on purpose or can I also use the more generally applicable Y? They mentioned Y in a lecture, but didn't go into detail. Do they always use this notation or only in special cases?

In my experience, this often leads to last minute "cramming": immersing yourself in the course material and (indiscriminately) memorizing as much as possible shortly before the exam. This allows the student to guess secondary details ("in this kind of question, we used axiom X, but we didn't cover Y so that's probably out of scope") more easily since they're fresh in their mind. However it is usually not a good learning technique. It is something students adopt to cope with the uncertainties that come with the test method*.

You can reduce noise by exposing students to the types of question, the expected scope of the exam and the answer style you expect throughout the course. Letting them practice on old (or mock) exams will allow them to test their knowledge, assimilate the cultural details over time and concentrate on understanding the subject they're actually supposed to study.

What if, after releasing past exams, almost no one fails anymore?

Congratulations, you have made your students more successful by allowing them to prepare better. This is a good thing. If you feel that the exam is too easy now, add more challenging content. You'll produce more competent graduates.

Caveat: This line of reasoning assumes that you prepare a "fresh" exam each time and that it is reasonably well designed to test for understanding of the subject matter in the first place. Reusing previous questions with minimal changes will defeat the purpose.

*Yes, laziness is also a factor.

Answer 6

Setting rules on the ban of those ressources will only help those who won't respect them even further. After all, how would those rules actually be enforced?

Instead, professors should make exams that actually measure the students' understanding of the subject, rather than their ability to memorize previously posed questions. If a student can pass an exam by just copy-pasting answers from previous one, I'd say it's the professor's fault rather than the student's.

So, yes. As many other answers have pointed out, "exam banks" will appear whether you like it or not, so maybe it is a good idea to make past exams public so that everyone plays by the same rules. Reusing questions is, in my honest opinion, a very bad practice

Answer 7

There are implicit differences across fields, examiners and countries. It is not fair to test students on knowledge without them knowing precisely what you're asking. No matter how specific you think an exam's questions may be, there will be certain expectations which have not been laid out.

I came from a physics/mathematics background but studied some softer science subjects and really struggled with some exams because I answered in the wrong style. What was technically correct was not necessarily what was being asked.

If an exam asks:

• "Give an example of symbiosis in nature."

Do you give a specific example, referring to two named species? Does this question refer to a mutualistic, commensalistic or parasitic relationship? This was covered in the class, should I use the example given in the lectures? Do I need to include a specific definition of symbiosis or can I just write two species?

If an exam asks:

• "Give the Taylor series expansion of sin(x)."

You might happen to know the answer off-hand; is it ok to just write that? Is some working expected? Do I need to write it as a sum or can I just write the first few terms? How many terms of the sequence are needed?

You may think you write perfectly concrete exams where every question has an exact answer but I can guarantee you do not because in my experience, no one does.

Answer 8

It really depends on where you are. I have never understood why many in North American can't see past exam papers.

I did maths at the University of Cambridge (England). Most of the work we did during the term to learn things was to answer questions from past exams and have these answers reviewed by our tutors.

The last 20 odd years past exams are actually available to all from the maths department web page https://www.maths.cam.ac.uk/undergrad/pastpapers/past-ia-ib-and-ii-examination-papers

OK the examiner has to do more work to invent new questions, but the students can never learn answers by rote as they do not know what the questions will be.

Answer 9

At my university in Sweden it is common practice to have previous exams available. I would almost like to state that they are required to due to the university's rules, but I am unsure about this; I strongly seem to remember reading something like it once.

I can let you know that the passing rate is not increased at all due to this; there are still the "Moby Dicks" of courses around there, gathering some 70% fail rate each year. As someone else said, it is good to know what is expected of you.

Some professors even hand out best student answer to questions from previous exams as well to have a reference to go after.

I am also in a STEM university, MSc and BSc level so far. (Let's hope for PhD as well!!)

Answer 10

Yes, of course.

They already are available. There were students there, present and able to remember, or keep copies. Which means, if you keep them "secret" - there will be a bias among your subjects, err, students. Those who have the old tests and those who don't. Likely there is a correlation - this bias is higher among those who either A) have money or B) have academic connections. So unless you find it perfectly acceptable that your meritocratic system (which in my view, is academia as it should be) degenerates into a nepotic or oligarchic system - you need to make that information available to all.

Besides, reusing old exams, by parts or whole, is an abhorrence of lazyness and shoddy workmanship. Be better. How can you be outstanding, if you accept such ineptitude.

Answer 11

Not necessarily past exams, but practice tests with worked solutions would help. Actively discourage students from just memorizing the answers at the last minute.

Look at MCAT prep. It's basically all practice testing. If you took the SAT or ACT or AP exams, did you take practice tests?

Here are several reasons to provode practice exams.

1. Practice testing helps students learn. Students are quite literally practicing the things you would like them to know.

2. Practice tests help students gauge their level of understanding. How many times have you heard a student say, "I thought I was going to do better on that test?" That would not be as likely to happen if they took a practice test or two well in advance.

3. It is more effective than other studying techniques. A lot of students, if they don't have questions to practice with, will reread the book, flip through their notes, and say they get it. They could be fooling themselves because they don't know if they can answer questions about the subject.

Answer 12

I think it's absolutely necessary to allow students to view past exams, at least in my subject (law). Students with good connections will get the past exams either way. It wouldn't be fair not to give the exams to everyone else, allowing those few with good connections to prevail without learning while everyone else struggles.

Past exams are also a good way of getting a feeling for what answer a professor wants to read when asking a certain question. Many professors have their very own style in HOW they ask questions if they want to hear a certain thing. It's only beneficial.

Answer 13

Two points here:

How do students prepare for an exam?

• Read textbooks.
• Read course materials.
• Attend class and take notes.
• Review their own notes.
• Review notes taken by other students, of the same or previous years.
• Do homework assigned in class and have it graded.
• Do extra exercises from textbooks.
• Do extra exercises under exam-like conditions. (Regarding quantity and time, allowed materials, etc.)

Thinking back to my university days, the last was more energy than the average student would spend on the average exam. But it would be a sign of dedication, not cheating, if one did it. Now what is a legitimate source of these exam-style exercises?

What do exams test?

If the ability to memorize the previous half dozen exams means one can pass the next one, either the lecturer is recycling questions, or the course is really basic. In that case, wouldn't a textbook contain all the things one has to memorize to pass?

Answer 14

One thing that always works, but is a hard sell with students, is that they should work out as many problems as possible without looking at solutions. I'm for having lots of problems available, I'm very much against having lots of answers available. My exams are often modified (simplified) homework questions, and so the book generally has tons of sample exam questions to begin with.

I teach Mathematics, it may be somewhat different in other fields. In math it is also common, at least in the US, to not teach the same class over and over, so being able to come up with new exams is not so bad.

Answer 15

I am not sure there is a universal answer to this: the correct answer may depend on the discipline, the level of the course, the number of students in the class, etc.

Amongst arguments in favour of this:

1. It levels the playing field for students: by making sure that all students have access to exams, there can be no subset of students with a putative advantage of having some exams while others do not,

2. In principle, if a student can correctly answer most of the questions on the previous 3-4 final exams, this student should do well on most questions of the current exam (assuming the core material is the same). Thus, past exams provide guidelines on the topics covered in the exam, the emphasis of some of the material, and also allow students to become familiar with the style of the instructor.

3. In principle it forces the instructor to produce some additional questions every year and thus keeps the finals more “current” than old ones.

The major argument against this is that some students will study the past exams rather than the whole contents of the course, so that any parts of the course not emphasized in previous exams will get comparatively less attention, and the student will have a narrower grasp of the material. This can set up some students for disappointment and false claims that questions are overly different from previous exams.

Answer 16

Yes, in the current state of US university education, every student should have explicit, formally-accepted access to all available former exams for every class.

Why? Because of three problems.

Problem #1 - former exams are the only decent resource for studying for university-level exams. Quoting from Trylks' answer: http://academia.stackexchange.com/a/23128/56720 on what to do about students acquiring old exams:

The questions in an exam and the exercises done before the exam should not be more different than an exam and a previous exam, actually old exams make for perfect exercises and practice

The point is that studying the courses and making a few exams/exercises to practice should be easier and lead to greater success than checking the whole compilation of past exams, which should anyway lead to a good knowledge of the contents of the course (in a more tedious way than reading the theory and checking this with a few exams).

Personal story: When I was in high-school the homework would be much harder than the exams, anyone making the homework (optional) would get good qualifications and the qualifications would reflect actual good knowledge about the subject. In the university there were exercises, but they were explanatory and very basic, the questions in the exam were much harder. This made the exercises useless, students needed exams from previous years just to practice in answering the questions, and I hated that.

I agree wholeheartedly with Trylk's assessment of modern university exams - they are written so differently from the rest of the coursework that they are often impossible to effectively prepare for without former exams. Professors don't provide students with any better alternative for study, so students have to rely on the tedious practice of reviewing every former exam they can scrounge up.

Problem #2 - some students will always get ahold of former exams.

They simply will. Some frats are able to consistently find recruits on the strength of their extensive record of former exams in verious courses. Some students have brothers, sisters, friends, etc. that took the course last year. Some are willing to outright buy former exams. There's simply no stopping it, and even a dedicated effort to prevent it will still only catch a few.

Problem #3 - because of problem #1 and #2, students who can find a former exam have an unfair advantage that's unlikely to be addressed

If professors upped their game in terms of coursework and exam writing, then you might not need this policy. But, as things stand, I think that the only way to create a fair playing field is for former exams to be made available to everybody.

Answer 17

The exam should be written with awareness of what resources the students have access to. If that condition is met, and all students have access to essentially the same resources, there's no problem here. For example, if a professor plans to re-use a question from a previous exam, then they should either closely guard their old exams or make the exams easily available to everyone. I've had professors do things like tell the class that a homework problem was going to appear on the exam verbatim, but not tell them which one; since it wasn't practical for students to just memorize every homework problem with its solution, this forced the class to work hard to understand the method for solving each of the homework problems. It would be reasonable to try something similar with past exams, e.g.:

Here are all the exams I've written for this class in the last ten years. One of the problems on the exam will be exactly the same as a problem from one of these examples.

It's also worth noting that exams are often different in format than other means of evaluation in the same class – homework questions are usually very different from test questions, and (at least when I teach) weekly quiz questions are very different from exam questions. But it doesn't seem fair to evaluate a student based on how well they handle a completely new style of question, if the goal of the class was to teach key concepts or techniques – to take an analogy, imagine if a student had a hard time reading the exam because it was printed on blue paper instead of white. Would the lower score they received as a result accurately reflect their understanding of the material, or just their discomfort with the way the questions were presented?

Because of these factors, I generally supply my students with past exams, with solutions, and with a "mock exam" that I write fresh alongside the actual exam. This seems to improve student success significantly – and by that I don't mean it just inflates grades, which is definitely a legitimate concern. What I mean is that my students typically demonstrate stronger familiarity with the material after preparing for my exams using resources like these.

That said, there's another style of teaching that would make perfect sense, in principle. If you're teaching something that would be reasonably easy to memorize, or if having to deal with an unfamiliar format is a key skill you're trying to teach, then it might make sense not to release old exams. In that case, though, the professor would need to make sure that none of the students got hold of old exams, because otherwise the exam would be unfair to those who didn't; that means they'd have to disallow students from taking their exams home with them after they've been graded, and they couldn't release solutions after the fact. The example you gave sounds like a situation like this going wrong - access to old exams was intended to be controlled, but that control failed, resulting in an unfair situation. In the modern era of the Internet, controlling information like that is difficult, and probably not worth it; prohibiting students from using old exams is more likely to lead to more cheating and more unfairness than anything else.

In short, a professor needs to pick an option and embrace it. Either make old exams available to everyone and write the exam with the expectation that students are looking at previous ones, or make old exams available to no one.

Answer 18

It's best to ignore this issue, whether some students get an advantage over others at exams this way is irrelevant, because exams are at best only a very crude way to assess students' ability anyway and those students who focus too much on exams tend not to become the best experts in the field later.

So, this is just yet another aspect of the "teaching to the test" problem that should be tackled comprehensively instead of looking at one symptom at a time and trying to invent some ad-hoc band-aid for that. The best solution i.m.o. is to get rid of exams and degrees altogether. Universities should be places where students buy an education of their choice out of pure interest or to get qualified for a job of they later want to do. How does an employer find out whether a job applicant is suitable for the job? By letting the job applicant sit an exam that tests the specific skills required for the job.

The employer can then choose to outsource this to some specialized testing company, or to a university. So, universities may still end up taking exams, but this is then not directly related to the courses given to students. Students can, of course, sit voluntary exams to test whether they can be confident to pass an exam set by a prospective employer.