Should previous final exams be released to students?

Question

In general, you can find the past exams for the past 15 years or so at my school. However I had one prof in Calculus III who chose to withhold his exams. This was the only instance in which a prof has chosen to do so in my two years at university studying engineering. His justification was that he didn't want students studying exams rather than the actual course material which I find to be interesting. As a student, I notice that solving past exams from the same instructor as practice can be helpful for a couple of reasons:

• it helps you get accustomed to the instructor's style of exam so you're less nervous when you're writing the actual exam. I find that even seeing the familiar cover page helps in calming me down just before an exam starts.
• it helps you figure out what type of questions the instructor likes to put the focus on (i.e instructor A prefers to put conceptual questions, instructor B hasn't put a question on the laplace transform in 5 years so maybe I'll study that later if I have time, instructor C always has a 2nd-order circuit as question 1, etc...)

Of course, the 2nd point is exacly why my calculus instructor chose to withhold his exams. But at the same time, some students may benefit if they use them responsibly as challenging practice problems without jeopardizing their learning experience (i.e someone with the mentality of "I know instructor B hasn't put a question on the laplace transform in 5 years, but I need to study it anyway because it's important"). Most of my other instructors release their past final exams, but do not post full solutions (or full solutions to just last year's exam). A few have chosen to release just the final answers to the questions without the full solution.

Is it better for the students if the instructors released the past final exams? What about full solutions (or just final answers) to the finals exams? Are there studies that show if one strategy is better than another?

I understand that the answers and arguments may vary depending on the field so I should add that I am particularly interested in an answer for the fields of mathematics, science, and engineering.

Answer 1

I don't have an opinion on releasing solutions, but the exams should definitely be released. Some students will have access to them through previous students, one way or another. Making them officially available makes sure everyone has the same access.

Answer 2

I would strongly say yes.

Personally, I am very unconvinced by this argument:

His justification was that he didn't want students studying exams rather than the actual course material which I find to be interesting.

Why is there a mismatch between the course material and the exams? This angers students and doesn't accomplish anything. Well-designed exams don't admit shortcuts, but rather force the students to learn the material which the instructor wants to teach.

It is human nature to respond to carrots and sticks --- ask anyone employed as a personal trainer. These carrots and sticks can be lined up with whatever intrinsic goals the instructor wants to promote.

Answer 3

The question seems to assume that all professors write completely new exams every semester and never reuse an exam question. That's not the case for me, and I don't know of any colleague who does so. In fact, many people at my school use the same stack of photocopied exams over and over, because they don't want to be bothered with the extra work of making fresh copies. I usually write some new questions every time, but I also recycle a lot of questions -- the recycled questions usually constitute the majority of the questions.

Aside from laziness, there are positive educational advantages to reusing questions. Sometimes it is difficult to predict how students will do on a particular question, and most teachers find that they systematically underestimate difficulty. If I've used a question before, I have some idea of how students will do on it.

If any questions are being reused, then it doesn't make sense to release 100% of old exams. Some students will acquire third-party solutions to the problems. (There are sophisticated for-profit web sites that do this sort of thing worldwide.) The students who do this will have an unfair advantage over those who don't.

In a STEM subject that focuses on problem-solving, such as physics or math, it's not necessary to make old exams available so that students can see what kind of questions are likely to be asked. They can tell this from the problem sets that were assigned.

Answer 4

Since you cannot prevent students from writing down from memory the questions they just got on an exam -

Yes, the only coherent choice is to fully release these exams

Otherwise, you're just putting an artificial barrier before this release: The barrier of someone coming to the exam and wasting some of his/her time copying its contents. Or of having a student with eidetic memory simply take the exam as usual, then reproduce it later. Any half-decent student union would get these exams (re)produced and publish them, possibly with solutions, as a service to its members. That's what happens in my Alma Mater, anyway - they had a nice store of printed exam booklets.

Answer 5

Availability of worked answers/past papers etc should ideally be consistent across all sections of all courses within a given degree program, be it a pure STEM subject or a dual award/tripos structure. Anything else risks the level playing field that supposedly exists. In practice the only way to ensure fairness is to let all the information be freely available.

E.g. I went to a collegiate university, and found that the availability of past papers, worked solutions, already completed lecture notes (my courses typically had handouts that had gaps to be filled in at the lecture) varied between the colleges. The variation was most prominent in engineering, where some colleges had access to past students notes and solutions to problems as well as past papers, all of which had been scanned and uploaded to the intranet, some had access only to past papers, and some had no such access.

Obviously this creates quite a disparity in the amount of material one has to look over to understand a concept or problem and furthermore, with access to such material crucially one learns how to best layout an answer for maximum marks, something that wasn't otherwise shown. If papers/worked solutions are formally withheld, you can almost guarantee that some will be available informally to some small part of the student group through connections to higher years (in this case collegiate). This results in an advantage unrelated to your actual ability in comprehension and implementation which is the supposed testing target of university exams.