25

I've noticed that many professors seem to impose extremely tight time limits on their exams, even though the subject in question would never have such extreme time constraints in real life. For example, I cannot imagine a situation where I have to write 50 lines of code on a piece of paper in less than 15 minutes.

Is it reasonable for me to complain about such constraints or is it normal to expect fast exam performance from students?

Note that I'm not talking about the actual exam difficulty, just the time constraint built around it.

7
  • 15
    As a professional programmer, I can't imagine a situation where I have to write 50 lines of code on a piece of paper in any time period. A quite bad (but not necessarily atrocious) job interview might call for a lot of code on whiteboard or paper, but it's fairly futile to write that amount of code without the capacity to test it. The time limit is not what makes that task artificial. Commented Apr 15, 2014 at 16:27
  • 1
    @SteveJessop: That may depend on the strictness of how correct the code must be - I have often found myself scribbling down an algorithm on paper while being underway (also on my way to or from work), so I don't forget it by the time I get to continue development of the respective project at a computer, and also to make use of the time where I don't have a computer with a keyboard around. Of course, seeing an analogy to an exam task here requires that the exam task be considered correct when it is conceptually correct, with a rather lax interpretation of the intended syntax. Commented Apr 15, 2014 at 20:51
  • 3
    @O.R.Mapper: fair enough. I don't produce what you'd reasonably call "code" under such circumstances, but maybe that's just because I didn't do a CS degree and hence didn't take these exams :-) Rather, I write enough to remind myself how to write the code later, which normally is a key insight or a picture. Commented Apr 15, 2014 at 23:10
  • 2
    If you need 50 lines of code to answer a 15-minute question, the time constraint could be unfair, the instructor might be looking for a reasonable line of attack instead of a fully polished solution, or you might just be doing it wrong.
    – JeffE
    Commented Apr 16, 2014 at 4:15
  • 3
    As my math teacher once said: "I can run 100 meters, too, but I still don't go to the Olympics".
    – Cephalopod
    Commented Apr 16, 2014 at 12:15

8 Answers 8

26

The objective of an assessment can vary from one lecture to another. Quite often, the point is simply to validate whether the expected learning outcomes of the lecture have been met. For instance, if an expected outcome is to know almost by heart how to write an array sorting program, then it's quite reasonable to ask to write a small amount of lines of code in a fixed amount of time.

If, on the other hand, the expected outcome is at a higher level of understanding (in the sense of Bloom's taxonomy), for instance by asking to design and assess a new data structure to handle a new problem, then it could be more reasonable to expect more time.

The problem you are referring to by "would never have such extreme time constraints in real life" is addressed with the notion of authentic learning (Rule, 2006), which identifies the four following themes for a learning to be authentic:

1) the activity involves real-world problems that mimic the work of professionals in the discipline with presentation of findings to audiences beyond the classroom;

2) open-ended inquiry, thinking skills, and metacognition are addressed;

3) students engage in discourse and social learning in a community of learners; and

4) students are empowered through choice to direct their own learning in relevant project work.

Authentic learning comes with its upsides and downsides (the reference linked acts as a survey paper, if you are interested), so it's not necessarily the best approach. In particular, (Lombardi, 2007) note that the reliance on traditional instruction is not simply a choice made by individual faculty—students often prefer it. For instance, not everybody wants to be tested on writing code in a highly complex environment, using bugged code written by other people, implementing specifications that are sub-optimal, but the client want them in this way, which could be a typical real world situation.

I don't think you should complain about time constraints, but if you believe that authentic learning would be more beneficial to you and your fellow students, you should probably discuss with the professor about the objective of the assessment.


Lombardi, M. M. (2007). Authentic learning for the 21st century: An overview. Educause learning initiative, 1(2007), 1-12.

Rule, Audrey C. (2006). The Components of Authentic Learning. Journal of Authentic Learning, Volume 3, Number 1, Pages 1-10, August 2006

2
  • 6
    A huge +1 for an answer that ends with a bibliography. Commented Apr 15, 2014 at 14:13
  • I very much feel the pain for "need to work with bugged code, confused specifications, suboptimal tools". But some form of the above is an essential skill for any professional... and better learn it as a student (where not much is lost if the result explodes) than on the job.
    – vonbrand
    Commented Feb 23, 2020 at 2:09
17

This is not really an answer but some educated guess. Most instructors do not plan to impose a challenging time constraint for exams. Usually we tend to overestimate how easy is the exam. The younger is the instructor, the more optimistic he/she is about students and more eager to write interesting questions. For subjects that require lots of drilling, tight time constraints are helpful in the sense one can immediately screen out the students who did lots of homework/problem solving from those who have to think from scratch.

1
  • 1
    When starting out teaching, I was given the rule of thumb: Take the time you need to solve the exam, multiply by four. Not that I follow it each time rigurously, but it stuck.
    – vonbrand
    Commented Feb 23, 2020 at 2:11
7

In my experience, challenging time constraints are a hindrance because they induce panic, which inhibits the ability to reason through a problem and make a thoughtful answer to it.

I'm currently tutoring two students on grammar for a business English class. Their teacher sets 100 questions for each 50 minute exam. That comes down to 30 seconds for each question. This tactic makes a mockery of all my attempts to teach them the logic and reason behind grammar. The students barely have enough time to figure out what's being asked of them, let alone apply any rules or reasoning principles.

The students feel this is unfair; they learned the rules, did the homework (which average 300 questions per set), and now they're failing the tests because they can't go through the whole process of reading a problem, conjuring up the correct rule, and finding the correct answer in 30 seconds. This causes bitterness and low morale. It makes them feel that all the work they put in to the homework was a waste of time.

It's not feasible in the standard method of testing to give students unlimited time, and most schools don't have the resources to set aside computer labs for exams so CS students can write code on them. But there's no reason to make the time constraints more of a problem than they need to be. Make the questions easier; put fewer questions; and split long questions into multiple parts, which are graded independently. I've found all of these reduce the pressure, and the second two options let you still put challenging problems on the exam.

In short, yes, I do think it's wrong to impose a challenging time constraint on an exam. Not only because it's unrealistic, but also because it puts artificial limits on the students' performance, which the students will recognize as artificial and react badly to. Unless you're trying to teach people to program on the bridge of the Starship Enterprise during a Red Alert while an anomaly is sucking the ship in, lighten up the time limits.

8
  • 2
    +1, but I'm not so convinced. In the real world, conversation requires a quick and natural stream of thought and ease of usage of language. While a 100 Qs in 50 mins is definitely harsh, I think it's not completely unjustified because your students need to develop the ability to quickly use their grammatical skills in conversation. Commented Apr 16, 2014 at 8:17
  • @shortstheory: seems like both things are reasonable to test, and the problem here is that the tutor thinks one thing should be tested and the examiner thinks another thing. In conversation you just have to get it right, you don't have to show your working. I sort of wonder whether a typical native speaker of English would have a sporting chance on the test, though, given that we can spot most errors and correct them extremely quickly, but might be totally ignorant of the theory required to justify our answers. Commented Aug 22, 2014 at 10:14
  • ... that is to say, is the correct answer of form, "If I were you" (in which case the exam is testing fluency, which for conversation must be fast), or is the correct answer of form, "one must use the subjunctive mood" (which is testing academic knowledge of grammar, and where the argument that you have to do it instantly in real life business English doesn't really hold)? Commented Aug 22, 2014 at 10:16
  • 1
    @SteveJessop Actually, the students were native speakers of English, albeit low-income ones unfamiliar with formal written English. The class was supposed to teach them about business writing, but it did so by using (from my knowledge of linguistics, confusing and borderline wrong) grammatical explanations like you'd usually see in an ESL class. You didn't need to show any work, but the questions were purposely designed to confuse people with their background, so you did need to apply the theory to find the right answer.
    – tsleyson
    Commented Aug 23, 2014 at 23:30
  • 1
    I think what shortstheory says in the first comment is valid if you're testing non-native speakers. There is some justification. But in this context, the knowledge was meant to be applied to business emails and letters, so, as Steve Jessop says, the argument that you have to do this instantly doesn't really hold. You have time to think "do I need 's or s' here?", which you didn't have on the exam. I found this unfair and counterproductive. I told the students to learn the theory, even if they didn't have time to use it, because it would help them in the long run.
    – tsleyson
    Commented Aug 23, 2014 at 23:34
4

It is reasonable to bring up the issue of exam difficulty, although I would try to bring it up in a way that is not accusing the professor of being unfair, perhaps bringing it up as part of a question of how to study next time, or asking if this is going to be typical of his exams.

It may be that the professor sees value in the tight time constraint that you aren’t aware of. For example, they might be asking something very challenging, knowing that very few students will succeed completely. It is very common to have test questions ranging from fairly easy to very difficult, with the expectation that most students will miss the most difficult questions. Partially correct answers show partial understanding. In this case, a coding question with a time limit will show up problems that might be harder to see in an assignment where a student has unlimited time. One sort of issue that will show up is that there's big difference between the uncompleted code of somebody who starts writing random code in the hopes it will sort itself out as it comes, vs somebody who takes 3 minutes to plan an approach, and the rest of the time implementing that approach. But if all you are looking at is the final result is a problem set, the difference is less obvious, even if the first person takes twice as long to reach the same result. There could be other possible benefits to this type of question as well.

It’s also possible that the professor isn’t aware of just how difficult the test is, and your question will bring that to his attention. Although if nobody finishes that question, that alone will give most instructors a clue that their estimated difficulty was off. But generally, exams have very little to do with career prep, and more to do with assessing student ability.

4

Opinions on this vary, but personally I prefer to avoid cases where the major difficulty of the exam is the time limit. If the time constraints are too challenging then the process becomes a test of who can write the fastest, rather than who knows the material well. Personally, I prefer to give a time constraint that allows students to think about each question and go at a reasonable pace, with questions that range from easy marks to hard marks, to test their actual understanding of the subject matter. I have also adopted the following rules for time limits in exams:

  • Exam length for undergraduate-level courses: As a rule-of-thumb, the teacher should be able to correctly complete the exam in one-third of the time allocated to students, from a position of initial ignorance of the answers. So if you are setting a three-hour exam, you should be able to do the whole thing, including all working out and writing up of answers, in one hour.

  • Exam length postgraduate-level courses: As a rule-of-thumb, the teacher should be able to correctly complete the exam in one-half of the time allocated to students, from a position of initial ignorance of the answers. So if you are setting a three-hour exam, you should be able to do the whole thing, including all working out and writing up of answers, in one-and-a-half hours.

These time limits have served me well in my courses, and it ensures that I am testing students on their knowledge of the material, rather than their ability to write really fast.

3

Some professors seem to be using this paradigm of testing; instead of giving ample time so that it's reasonable to finish the test, they'll intentionally make it so even the best-prepared testee will only have time to finish 75% of the test. Of course it will be curved based on this. It separates those who know the material really well from the people who only know a little, and apparently creates a pretty good bell curve, which seems to be what everyone is after these days.

3
  • 8
    It will also separate the people who have test anxiety and the people who don't, since anyone with the slightest amount of anxiety will fail. Commented Apr 15, 2014 at 21:47
  • 2
    anyone with the slightest amount of anxiety will fail — [citation needed]
    – JeffE
    Commented Apr 16, 2014 at 4:16
  • 5
    "It separates those who know the material really well from the people who only know a little" -- [citation needed]
    – Lii
    Commented Apr 16, 2014 at 13:35
3

It's not uncommon to have challenging time constraints. Some American tests like the ACT and SAT have a tight limit built in to reward students for being fast.

However, many of these tests are scaled, in that no matter how well or how badly the students do, the grades are normally distributed..

1

Tests with tight time constraints are probably created by instructors who are not familiar with personality profiles like those revealed in Meyers-Briggs personality tests. People with some personality types tend to test slowly but deliberately, while others will test quickly but haphazardly. Of course, I'm also sure that there are those that test quickly and deliberately, as well as slowly and haphazardly. But, in the end, if instructors thought about it, they would find that they most likely prefer students go slowly but deliberately, thereby raising their percentage of correct answers. I imagine that most employers prefer slow production in order to achieve a high rate of perfection, thereby limiting wasted time and material.

3
  • I am an industry retiree. My personal observation is, most employers prefer to quick production in order to achieve a high production rate, thereby limiting short term cost.
    – Nobody
    Commented Apr 16, 2014 at 6:16
  • 1
    It's all relative, though. My employer wants me to get a bit of a move on, sure, but very rarely do I work at the intensity that I achieved for 3-hour blocks during my finals exam. Quite aside from anything else, I was intellectually exhausted and even physically fatigued after two such exams in a day, where in a normal work day I'd still have an hour and a half to do! Commented Aug 22, 2014 at 10:03
  • 1
    Furthermore, I reckon my error rate was higher in exams than in my professional work, and that's fine because the mark scheme in my exams was to do as much correct work as possible in the time allowed, whereas the "mark scheme" in my professional work puts high negative weight on errors. I mean, we don't have an actual mark scheme, but we'd rather put out one page of correct work than three pages of work, one of which is entirely incorrect and we don't know which. So yes, my employer is more error-averse than my examiners were. And a thesis assessor is more error-averse than a finals examiner. Commented Aug 22, 2014 at 10:07

You must log in to answer this question.