Context: I am an assistant professor of mathematics at a small liberal arts college in the US.

I generally find student questions during tests and quizzes to be somewhat annoying. I disliked them when I was a student, and they're even worse as an instructor.

To elaborate on my dislike:

As a student I found such questioning very distracting. I'd be in the midst of trying to solve a difficult problem when suddenly people around me are talking in hushed voices about a problem I had already done. As one can imagine, it was challenging to pull my attention back to what I was doing.

As an instructor, I want to do my best to provide a good test environment, so I would like to minimize questions that I see mostly as a distraction. In addition, most courses contain a student or two who I don't entirely trust, so I like to be able to watch the students during the exam to ensure test security, and answering questions can hinder that. I also don't want students getting up during a test to ask questions, for obvious security-related reasons. (This problem is easily solved by telling them to raise their hand.)

Most questions fall into the following categories:

  1. Legitimate questions aiming to clarify vague instructions or a typo.
  2. Questions fishing for a hint.
  3. Questions seeking support of a solution method.
  4. Questions asking for confirmation about answer format.

Type 1 questions are not a problem. Type 2 and 3 questions are foolish, because I hardly ever tell students anything remotely useful. Type 4 questions are irritating, because my questions specifically say what format to use (i.e. "show two decimal places," "you do not need to simplify," etc.).

What are some strategies for eliminating questions of type 2, 3, and 4 while not discouraging type 1 questions?

Of course, I can just straight-up tell them that they're not to ask questions unless they believe them to be very legitimate, but I'd like to use a less "adversarial" approach, if possible.

  • 25
    I fail to see the difference between type 1 and type 4. Oct 20, 2017 at 17:55
  • 11
    @DSVA Why not saying openly "we cannot answer that -- it's part of your exam to figure out the answer to this question"? Oct 21, 2017 at 17:16
  • 18
    Just ignoring the question is terrible pedagogy. It is far more helpful to be told "I cannot provide that information during the test" and possibly "your answers must be your own, I can't give hints" than wonder if you are going to check and come back to a student with the information.
    – Nij
    Oct 21, 2017 at 21:07
  • 3
    @FedericoPoloni because it makes noise and distracts others, and that's the whole point of this question. You can just shake your head, or in case of point 4 where people ask things where the information is provided you can simply point to that. A lot of communication can happen without talking.
    – user64845
    Oct 22, 2017 at 1:20
  • 2
    I really just hope teachers wouldn't write vague or unclear questions in their exams. After all, isn't making a fair exam part of their job? After that, if there is some lack of clarity left, the students taking the exam should be told to assume a sensible (in the context) situation, and document their assumptions. Actually, they should know to do that even without being told. Afterwards, if a question is deemed to be totally broken, take that into account in the grading, possibly even ignoring that question completely (but you shouldn't need to go that far).
    – ilkkachu
    Oct 23, 2017 at 18:48

14 Answers 14


I explain at the start of the first test that, "Questions clarifying what you're reading are okay (raise your hand, I'll come over). I cannot tell you how to solve a problem. My response may be, 'I can't tell you that during the test.'"

If Question #1 comes up, I do as stated. Questions #2-3 get a fairly loud (but polite) repetition of, "I can't tell you that during the test." (Or, "That's what you're being tested on."). Question #4 may get, "Just read the direction carefully" (which itself is a repetition of something I need to say during almost every single class meeting).

I find this public act tamps down on questions as much as I can hope for. The main thing is to be solidly consistent. Counterexample: I was co-proctoring with another professor in their class. That professor got a question, shook his head in disgust, muttered "You should know that", and walked away. Then turned around, walked back, and actually did talk the student through how to solve the problem. This can only encourage more fishing questions like that.

Due diligence: In liberal-arts math courses I have occasionally had students express disbelief and/or outrage that I wouldn't help them during a test. "But you're the teacher, that's your job" kind of thing.

  • 13
    If I were a professor and I got a student complaining that it's my job to help them do well on tests, my response would be basically "It's my job to teach you about [course topic]. It's your job to learn it, and prove that you've learned it by doing well on the test." There's probably a reason I'm never going to be a professor.
    – anon
    Oct 22, 2017 at 0:25
  • 5
    @QPaysTaxes: Yes, that's probably been said by professors at community-colleges (et. al.) a few tens of thousands of times this year. Oct 22, 2017 at 1:01

You ask about minimizing questions, but really from how you describe your motivation it sounds like the real question is about minimizing distractions to you and to the students; the questions of type 2-4 don't bother you per se, but only to the extent that they create a distraction.

Well, it seems to me that distractions can easily be brought down to a level that no one can reasonably find objectionable by imposing a suitable protocol for students to ask questions, and without compromising any student's right to ask any question they think is appropriate to ask or making the students feel like you are treating them in an adversarial fashion. A couple of possibilities that come to mind for such a protocol are:

  1. A student who wants to ask a question should raise their hand and wait for you to notice them. You will then give them permission to get up and approach you in a part of the class that's far enough away from the other students so as to minimize the noise and distraction that your conversation will create. They then ask the question, you answer it, and they get back to their seat.

Or, if that's not quiet enough, then a slightly more drastic option would be

  1. The student will write their question on a piece of paper and raise their hand. You collect the paper, go back to the front of the classroom, think about the question, write an answer (or something like "no comment" if the question is not legitimate), and give them back the paper. If they are still unsatisfied with your answer, they can ask a followup question by following protocol 1 above.

Finally, I should add that I see some psychological benefit to allowing students to ask even illegitimate questions (e.g., of type 2-4). We should remember that exams are a very stressful situation and induce a high level of stress and anxiety in many students, which can be quite debilitating and hurt their performance on the exam. I think it's important to be as empathetic and mindful as one can to this fact; in particular, giving students the impression that you are friendly and stand ready to answer any questions they may have can go a small way towards reducing the level of anxiety some students will experience, even to the extent of allowing them to perform better (and want to ask fewer questions!). And this is true even if in practice when the students ask an unfair or illegitimate question you give nothing away. So, it may sound a bit Machiavellian, but in my opinion creating the appearance of being helpful can be just as important and useful in an exam situation as actually being helpful.

  • 12
    You make a very good point in your last paragraph--I believe that a significant number of the students who are asking questions are doing it simply to make themselves feel better, whether or not they realize it. I do try to present a very affable personality to my students, as many of them are already scared enough to be in a math class. Your answer reminds me of the importance of doing this on test days as well. Oct 21, 2017 at 1:04
  • 4
    When I read that I thought not about students wanting to make themselves feel better, but the "converse" (?) fact that if the students need to devote mental effort to deciding whether it's safe to ask a question in the first place, it's likely to hurt their performance on the test. That would be an argument against coming down hard on students who ask illegitimate questions.
    – David Z
    Oct 21, 2017 at 6:07
  • 1
    I wish I could upvote only specifically option 2. Sounds like a perfect solution. No talking/whispering, you get to give the exact reply you want without much follow-up arguments, you can still pay attention to the other students while "receiving" the question (hearing vs reading, which you can do while still checking your problematic students), the student will think about how much time do they really want to spend explaining their question, thus making it concise.... tons of advantages and addresses all OP concerns. I loved it!
    – msb
    Oct 23, 2017 at 23:08

My approach is approximately "no questions, period" (not after the first few minutes). I'm only in the room to proctor, not to talk about the exam. Even if there are typos on an exam, it is a very mixed thing to "make corrections" during the exam itself, because some people will already have spent time on the thing, etc. That is, "fairness" is sometimes approximated best by "uniformity".

One could follow up on "refusal to answer questions" by observing (to the students) that understanding expectations of format, this-and-that, are a large part of what such exams are meant to test. This may not console everyone, but it is a genuine point, and deserves repetition.

So, yes, even when there have been ghastly typos, I've just kept things as they were. Yes, this approach does entail announcing that, no matter how scrupulous I've been, there may indeed be typos, and that students should exercise their own judgement about that possibility. (Again, even if there are typos, it is not possible to correct them in-exam in a fashion that is clearly fair to everyone, unless it happens within the first few minutes.)

It did take me a long time (30+ years?) to get to this point of view...

  • Comments are not for extended discussion; this conversation has been moved to chat.
    – StrongBad
    Oct 23, 2017 at 14:35
  • 1
    Do you feel the "yes the typos exist, deal with it" approach properly prepares students for life after university? Because in industry it is stressed that the earlier a problem is identified and correct, the less cost and risk there is to the firm. How does that factor in to your approach?
    – corsiKa
    Oct 23, 2017 at 20:06
  • 1
    @corsiKa, I see your point, but/and typical "exam" environments are soooo artificial, so unreal, with so many unrealistic/anti-realistic constraints, that normal best practices are not easy to apply sensibly, I think. It is nearly impossible to make an exam that truly tests what one would like to test, I claim... so we inevitably make caricatures. And so on. Oct 24, 2017 at 0:19

Some of my old professors followed this procedure:

  • Sit down and get test sheet.
  • Get 5 minutes to read the test carefully (pens down), we were able to ask clarifying questions during this time. This period did not count towards the time limit.
  • Start of the test, no more questions beyond this point.
  • 1
    Yes, this is a good procedure! Oct 20, 2017 at 22:35
  • 37
    I strongly disagree with the procedure. As a student, many of the times I needed some relevant clarification was mid-test, when I was actively trying to solve the question. If the professor refuses to accept questions after the 5-minute period even if there are typos or ambiguous instructions, then the student suffers from the professor's mistake.
    – user63725
    Oct 20, 2017 at 22:56
  • 14
    As someone who has encountered problems and issues in tests that only become apparent after some effort to solve them, I think this is a rather poor approach. -- Not all issues (by which I mean actual mistakes in question phrasing) are immediately apparent during a brief 5 minute examination. Poor wording or unspecified edge cases can raise their heads in the midst of solving it, rather than being immediately apparent. (For example, due to a typo on one test I took, a calculated ratio which needed to be approximately an integer for the result to make sense wasn't even close.)
    – R.M.
    Oct 20, 2017 at 23:07
  • 1
    Due to the structure of my exams and the nature of the questions, I don't see this approach working especially well. That said, I think it could definitely work is some settings! Oct 21, 2017 at 0:59
  • 3
    I've followed the first two steps for years, but I strongly disagree with the third.
    – JeffE
    Oct 21, 2017 at 20:50

Some practical points:

  • Spend the time and re-read the exam text "adversarially", trying to think up any possible way to misinterpret them, and correct for it. On the other hand, better to leave some option for mis-interpretation than spending half a page on spelling out the interpretation.
  • Have another teacher/TA in the course sit down and solve the exam in its entirety, both to catch potential errors/unclear points and to time it and assess difficulty.
  • Do answer trivial questions (e.g. confirmations) briefly, but kindly and in a supportive rather than derogatory tone. Example: Q: "Am I really supposed to do XYZ?" A: "Yes, those are the instructions." Don't try lecturing people about asking such questions; don't repeat the demand not ask such questions. Instead, try to smile (if you can manage doing that in a non-creepy way), expressing the pride of having brought your student from relative ignorance to the brink of being proven to have mastered the subject matter. I always felt proud of my students taking their exam!
  • Remember that students' questions while taking the exam may uncover an unforeseen problem with it. This has happened to me quite a few times - both as a student as a TA.
  • When faced with a "How do I solve this?" question, and declining to answer - have a look at the student's face. If they appear super-stressed, shaking, desperate - consider suggesting they try moving on to the next question/problem and go back to this one later; or that they go have a drink of water and catch their breath. Consider asking "Are you all right?" if you're actually concerned about them. Some people occasionally have 'blackouts', or otherwise become very anxious, and this sometimes helps. It actually 'discourages' useless questions in my experience, because even noticing that a stressed-out student is being attended to with a kind demeanor reduces stress somewhat. That might sound irrational, childish, overly soft, immature to you - but remember you're testing the student's command of math, not the quality of their character or their maturity. Plus - perhaps, for you, an excellent academic who made it to professorship the exams were not that hard and stressful as for other students - don't judge them until you've walked a mile in their shoes.
  • If you get a lot of questions overall, or about some point in particular - when the exam ends, either stay there for 10 more minutes and indicate that you will listen to issues regarding that question, or hold extra 'office hours' - for hearing and addressing questions about the exam in a little more detail - once you are able to talk about the actual answer, and at more length. This will help you understand what it is people were finding difficult to understand or to solve. Also, this will likely have a positive psychological effect on the students - whether they did well or poorly.
  • 4
    These are some really good points. I especially like your point about keeping a supportive tone. The exam is already hard enough for them; it's not necessary to be overly strict with your responses. Oct 23, 2017 at 16:39
  • I love your point about tailoring my responses to their emotional state--I've been doing this unconsciously, but I think I can be even more effective now that I recognize the value of doing it deliberately. Oct 23, 2017 at 18:31
  • Having a TA take the test is a great idea. When I was a grad student, "locking in" concepts that I learned as an undergrad was a big part of my learning process.
    – sig_seg_v
    Oct 23, 2017 at 22:51
  • @sig_seg_v: I'm not sure how your two sentences relate to each other, but thanks for the compliment. Actually, it was an old tradition already in the mass first-semester CS introductory course in my alma mater.
    – einpoklum
    Oct 23, 2017 at 23:00

I am also a mathematics professor, though at a large state university. To be honest, I'm not sure why the questions are "annoying" to you. Maybe you could clarify this, since it could help you get a good answer. (In particular you say you disliked these questions as a student. Again, honestly, I'm not sure why you cared about that.) I also want to add though that I don't get that many questions on exams: probably about one question per 20 students per exam. It is plausible to me that it might be part of the culture of SLACs that students ask more questions on exams, but I don't really know.

With regard to the undesirable kinds of questions: I actually find 2) rather weird. I cannot think of a single instance in 11 years of teaching at my university where a student has asked me for a hint on an exam. Only very rarely have they asked a question where I think they are hoping to get information out of me that will make the problem easier to solve. This actually strikes me as being mildly inappropriate behavior. A lot of SLACs have strong honor code cultures, so you might think about whether/how these kinds of questions are not abiding by the spirit of the honor code. If you feel that way, you could put something in the syllabus about it. Then, if a student asks such a question during an exam, you can say "I really can't answer that kind of question. I'll explain more later." And then you can follow up with the student afterwards that they shouldn't be fishing for answers to exam questions.

With regard to 3): it's similar to 2) but more understandable/legitimate: after all, some solution methods are not going to get full credit. These types of questions make sense when they already have an answer. If they are asking something like "Would it be a good idea to integrate by parts?" then that really falls under 2). You can try to be as clear as possible in your course before the exam and in the instructions of the exam which solution methods you regard as legitimate, but often there is some real doubt.

With regard to 4): if they are asking things that are already stated on the exam, just answer quickly by calling attention to the fact that it's already stated on the exam.

Coming back to the beginning of my answer, I think you should reflect on whether, how and why the SLAC culture encourages students to ask too many questions on exams and also whether you think that students at your SLAC in general get rewarded for asking lots of questions. (E.g. do you think that other instructors give out information to individual students in response to questions? It seems a bit unlikely to me, but I suppose it's possible.) If it's clear to you that students are not getting rewarded for asking these questions then maybe you could frame it as advice. A student who is coming up every five minutes to ask a question is probably not getting the undisturbed thinking time that they need to do well on a challenging math exam. On the other hand, if the practice is really not hurting them at all and just mildly annoying you...maybe live with it?

  • 6
    FYI, in my lower-level courses at a community college, this is an everyday occurrence. Students will not even be able to tell what's appropriate or not. I usually have one student per course who will ask 10 questions per exam. Putting a statement on the syllabus is of no help, because they don't read that either. It's a widespread problem, and for many of us the cycle is frustrating and tiring some days. (c.f., question on certain needy students expecting to use all your office hours every day.) Oct 20, 2017 at 19:01
  • 1
    Also, I don't mean to rub it in, but this semester "all of my office hours" consists of two hours, both on Wednesday. So it would be okay if a student wanted to show up to both of them....though so far as I can recall that has not happened yet. Oct 21, 2017 at 1:41
  • 2
    Would someone mind qualifying the term "SLAC"? Throughout my (UK) education hardly anyone ever asked an question during an exam; when they have, it was often followed by an announcement to the whole room that "Paper Z, Question X has a typo...", or "Diagram Y is wrong, heed the text...", or the invigilators suddenly decide to write the Date on the board. At my university, the instructors were rarely in the exam room, but were always contactable by the invigilators to clarify any errors in the papers, and the layout of the rooms made communication with individuals rather difficult. Oct 21, 2017 at 9:16
  • 1
    @VisualMelon, "small liberal arts college". A peculiarity of the US system which aims to produce all-rounders by forcing people to take a wide mixture of courses even in tertiary education. At the risk of caricaturing, where the UK system abandoned the trivium and the quadrivium and introduced specialisation from age 16, the US just updated the contents. Oct 21, 2017 at 16:49
  • 1
    some solution methods are not going to get full credit — Say what now? Why would some solution methods be worth more credit than others, if they correctly lead to the same correct answer?
    – JeffE
    Oct 21, 2017 at 20:53

If students get questions clarified when they ask, suddenly the incentive is for every student to ask about every test problem (even if they don't need it). This is obviously suboptimal.

Answering many questions is also unfair, as two students might both be confused by a test problem but only one asked -- hence, they are not being tested on the material, but on their willingness to ask a question during an exam.

My solution is to answer almost no questions. My typical answer is "knowing the answer to your question is part of what's being tested here". What's being tested includes literacy (reading the directions to the exam and the directions for the problem and interpreting same). I make exceptions for two cases:

  1. Typos in the problem.
  2. Weird overanalysis of the problem.

In the former case (I made a mistake), I announce to the whole class the correction. The latter case is when a student is contemplating using very powerful tools from another course, that he or she doesn't really understand, rather than the simple tools available in this course. Then I tell the student that it's not necessary to do it the complicated way. This happens rarely, so I am not concerned about fairness.

  • Case 2 is essentially something that falls into "type 4" in my original question, and I think you've done a really good job of capturing the essence of this behavior. Oct 23, 2017 at 18:29

Consider that in many places, for many exams, the proctors have nothing at all to do with the class or the subject matter. They are just there to supervise the room. Clearly, in that environment, there are no questions. If there are flaws in the exam, the faculty will have to make whatever compensations seem least bad after the fact. It might, then, be easiest to set expectations that all exams work this way, even if the proctor happens to be someone who could possibly answer a question. When I was a student of math and engineering back in the dark ages, that's how it worked. You got a test, you did the best you could.

Every student should know better than to use up all the time on one question. Even if the exam is perfect to the letter, there's always the chance that the student just isn't going to find the path to an answer in the available time. Telling students to (a) put a question aside if they are not finding a solution in reasonable amount of time and (b) write an explanation of why you think the question is insoluble is better than answering questions. If the question is broken, then (b) helps. If the question isn't broken, perhaps the student will earn partial credit for a thought process that is partially correct even if it hit a pothole.

  • Actually there is a half-way house: the proctors are from a different department and know nothing, but the question setters are available outside the room for the first 20 minutes, and the proctors will relay questions of the type "Is this an error in the exam?" and announce any replies which contain corrections. Oct 22, 2017 at 7:45

First, make sure that your questions really are as error-free and unambiguous as possible. Achieving that quality takes effort, though. You will need to have one or two trusted TAs actually take the test to ensure that there are no obvious issues.

At the exam itself, announce a policy that if you have difficulty understanding a question, try your best to solve it on your own, and write down your assumptions along with the answer. It would probably be helpful to state that you have planted no trick questions, and that smart-alecky assumptions won't fly. Then, stick to your word and accept any reasonable alternative interpretations when grading the exams.

If you do get requests for clarification anyway, answer most of them by simply reiterating the "state your assumptions" policy. If it turns out that there really was a bug in the exam, own up to it publicly by writing an announcement on the board.

  • Anecdotally, the assumptions list is generally the root of most ambiguous questions I've experienced. When I was a student and encountered ambiguous questions, I'd present to the proctor the "assumed interpretation" along with either the literal or alternative interpretations that would lead to different answers or a different question entirely. Alternatively, if I didn't trust the proctor to admit the ambiguousness, on the answer sheet I'd list out multiple answers to the question based on each interpretation.
    – user45909
    Oct 23, 2017 at 22:04

I find magu_'s approach the best if you have to proctor your own class. Our institution solved the issue by having exams proctored by non-academic personnel.

Beside the obvious advantage that they simply would not be able to answer illegitimate questions it is also more cost-efficient. The academics have that time to do their research.


You just have to warn your students beforehand. As in,

Questions about the format of your answer will not be answered. Fishing for hints will not work as I won't give any useful info about the questions to any of you.

Give them a few minutes extra before you start the exam and ask them to read the questions and ask if the questions are clear. Giving them even just five minutes to read and understand the questions and letting them ask anything they want to be clarified will let you have an easier time later on.

Anyway, since you seem to be grading the papers, I'd also like to add that the type two and three are just fine as questions in my opinion, with the appropriate precautions. We had one professor who would give out any hint that you wanted him to give. He would start the exam like so:

You can ask for hints on questions and anything else your want. I will give you whatever you ask for on the question. However, this will mean that I'll deduct points appropriately. If you want the whole answer to a question, of course, I will even give you that, but, that will mean getting no points out of that particular question.

I had to ask for a hint once, it was something minor, so he wrote -2 to the side of my question (it was a 20 pointer out of 100), and just gave me the hint I needed. With the hint, I was able to answer the rest of the question, getting 18 points in total from that one. Though, if I couldn't answer the question, I'd have lost two points from my paper and that would be it.

I was the only one to ask for a hint on any of his exams because people thought he would deduct too many points. He was really fair though, a tiny hint wouldn't mean deducting half the point worth of that question from the paper.


Math student here;

Casting my vote for the direct, adversarial approach (particularly if you're not generally adversarial as it will maximize impact.)

This is always going on even in upper level math exams at my med-lg state uni, and it is really annoying. I am always hoping for a rule-nazi of a professor to lay down the law, but half of them don't even bother closing the door let alone addressing in-class distractions like inappropriate questions during exams.

  • I had similar feelings as a student, but I'm curious if this is a minority point of view. Oct 21, 2017 at 1:06
  • What you consider an inappropriate question could very well be a critically important one to someone else. I find it highly unfair to impose your (pretty drastic) standard on everyone.
    – JS Lavertu
    Oct 21, 2017 at 22:20
  • How is it unfair/drastic? That is, what kind of question would be appropriate beyond a simple clarifying question (e.g. in this question is it implicit that blah is a function of bleh) Oct 22, 2017 at 17:53

During my teaching days, I managed a few exams. I used a strategy that was quite satisfyingly successful:

Right before the exam, I would have told the students that the exam would last at least for two hours. And during the first hour, no one is allowed to talk or ask any questions and I won't be answering any. If you need any clarifications, just move on to the easy or answerable questions during the first hour and give the rest a little thought. I should emphasize again that no one has the permission to talk for one hour.

Then after that, my job was way lot easier because many of the problems were automatically solved already. Although I should confess that my teaching days only lasted for two semesters and this method is not tested thoroughly, yet!

  • This would be a cause of great consternation for many students - judging from my own experience as a teacer.
    – einpoklum
    Oct 23, 2017 at 9:11

I am not entirely sure what you mean by type #4, but I will try to address it. From my experience, there are high schools were it is taught that some things have to been done a certain way (like it is only allowed to use only a certain method of solving equations or written division has to been written down in exactly that way etc.) which pleases the teacher. I worked with 10-year old kids which were familiar with negative integers, but always asked whether it is allowed to use them since their teachers "forbid" them to use those in class. When I announce Gauß algorithm, students ask whether it will still be allowed to use the substitution method to solve equations. My math teacher always marked my test's answers as wrong when I wrote them as fractions or roots insted as a rounded decimal number (even when the problem was purely algebraic without a "real-world-context").

With this I want to say that especially for new students, it is sometimes not easy to get used to the concept "everything which is correct and well-argued is allowed to solve the problem". On the other hand, I often find the university lecturers have a hard time understanding this mindset of the students.

So, if you get a question of type #4, try to think if it can be because of this problem. If so, do not dislike it, but answer it honestly. If possible, say beforehand that "every approach which satisfies the following conditions: [...] is allowed". Of course, if the question literally is "how many decimals should we write down" when it says "write 2 decimals", you can say "read the question again". On the other hand, if you write "you do not need to simplify", this can mean a lot of things depending on the teacher and you may want to clarify this.

You must log in to answer this question.

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