I have seen that TeX.SE has community polls on their meta site. Those are polls conducted on their meta to find out various things about user base. Similar polls were conducted on mathematica.SE, on academia.SE and maybe on some other SE sites I am not aware of.1

Would it be interesting to create such a poll also here for math.SE community? The questions might be inspired by the questions from the other polls. I can imagine also some questions about:

  • Demographics (e.g. users inside/outside academia, nations or at least continents, time zones, ...).
  • Areas of mathematics users are interested in.
  • Usage of math.SE and other sites in this network (frequency, which other SE sites, tools used to create posts or graphics used in posts...)
  • Usage of various tools and mathematical software.

In fact, some questions here on meta which might be at least partially answered by such poll. A long time ago we had a question about this site usage (although the site has grown considerably since than and those answers are now clearly outdated). In this question (now deleted) somebody wondered whether mathematicians use GitHub. This recent question was about male/female ratio among users. There was also a relatively recent inquiry about polls in general.

We can learn from polls on other site about the form in which such a poll could be conducted. At the moment I am mainly interesting to know:

  • Whether we want to have such polls. (Would their results be in some way useful or at least interesting?)
  • Which areas could be included in the polls? (I am asking about broad areas, like the ones I have mentioned above, rather than specific questions.)
  • Are there also areas which, for some reason, should be excluded?

EDIT: It seems that so far the idea was received relatively well. At the moment it is in my opinion too soon to start the thread with the community poll. (Currently quid's post has score +5, which is good, but probably not enough to call it community consensus.) I'd prefer to wait a bit to see whether there will be some other feedback or at least how the votes on the already existing answer change.

But perhaps it is not too premature to discuss some things about community polls, in case we decide to create them. But perhaps such discussion could be more suitable in chat. This is why I created this chat room. Examples of the topics which I imagine could be discussed there: various technical details, discussing whether some particular question is suitable, what experience from other sites that conducted such polls could be useful for us here, etc.

1 If I learn about other site which have conducted similar polls, I will add links here.

  • 5
    $\begingroup$ So, we would not have polls like: "What is your favorite calculus textbook?" $\endgroup$
    – GEdgar
    Aug 28, 2016 at 14:24
  • $\begingroup$ @GEdgar I'd say this goes under "tools." But perhaps it could be made explicit. $\endgroup$
    – quid Mod
    Aug 28, 2016 at 14:54
  • $\begingroup$ @quid I think that the type of question mentioned in GEdgar's comment is closer to book-recommendations which are valid questions on the main. (Although not formulated like that.) I also think that his comment was tongue-in-cheek. (Although it is a bit more difficult for me to detect sarcasm, as I am not a native speaker.) $\endgroup$ Aug 28, 2016 at 14:58
  • $\begingroup$ "What is your favorite calculus textbook?" is opinion based and/or too broad for main. If at all it is a "poll." On tex.se they have polls "favorite editor/distro/fonts etc" while actual questions on these are still on-topic on main. Beyond that it is not clear why a question which software one uses should be a valid poll while the question which book one uses is not. $\endgroup$
    – quid Mod
    Aug 28, 2016 at 15:48
  • 8
    $\begingroup$ I upvoted this because I think it's a sensible thing to discuss. However, I oppose polls. $\endgroup$
    – Lord_Farin
    Aug 28, 2016 at 16:56
  • 15
    $\begingroup$ Isn't this post itself a poll? $\endgroup$ Aug 28, 2016 at 18:10
  • 3
    $\begingroup$ "Would it be interesting to create such a poll also here for math.SE community?" Yes; it could also be informative and eye-opening. $\endgroup$
    – amWhy
    Aug 28, 2016 at 19:19
  • 2
    $\begingroup$ @Najib: One key property of the meta site is that there is some meta-characteristics to it. Like questions about the meta itself, or a poll about polls. Kinda like a pie chart of your favorite bars and a bar chart of your favorite pies. $\endgroup$
    – Asaf Karagila Mod
    Aug 29, 2016 at 8:43
  • 1
    $\begingroup$ Seems the response is mostly positive. I say, let's make a thread. $\endgroup$ Aug 30, 2016 at 18:42
  • 2
    $\begingroup$ @6005 Personally I'd wait for a bit more feedback. I am not sure +5 is sufficient score to be considered community consensus. (As you can see from Lord_Farin's comment, upvotes on the question does not necessarily mean agreement to creation of a community poll.) $\endgroup$ Aug 31, 2016 at 0:21
  • $\begingroup$ @MartinSleziak OK. Well there were two answers that would have provided a straw poll as to whether the community likes this idea or not, but both were downvoted to oblivion for a reason I can't understand. I cast the last delete vote just because it was already too late for them to provide an accurate poll. $\endgroup$ Aug 31, 2016 at 0:31

3 Answers 3


To me this is a rather light-hearted idea that might be useful in some ways and might be a distraction in other ways. I think we should just give this a try and see how it works out. If it becomes too much of a distraction or a source of friction without much in return we can just shut it done again. The way this is envisioned, if the linked to models should be followed, it's just one meta thread, not a big deal. It appears it does work reasonably well on tex.SE and on academia.SE, so why not here.

To be clear, technically, I assume the idea is to follow the linked to models that is there is one poll-thread, a new poll question is added as an answer-post and the answers to the poll are proposed in comments that can be voted on. (Votes on the answer-post can expressappreciation of the particular poll question; deletion of individual polls is possible too.)

Further and crucially, the idea as I understand it is not to decide something via those polls, but instead to have polls to carry out surveys. For those that do not know them I recommend to have a look at the threads on tex.SE and academia.SE questions include: What is your preferred TeX distro? What is your highest degree? On which continent do your reside? Is English your native language? And so on.

Yes, this is not exact, and yes some of it will likely be a bit on the trivial end. But, so what? It might also bring up some interesting piece of information, and those that do not like it should be able to ignore it. Finally, sometimes on main questions that could somewhat reasonably be handled in this way do come up, yet are not suitable there; it would provide a home for them.

tl;dr let's try it, if it does not work we'll just stop.


Personally, I'm not at all interested in having a Community Polls thread on our meta.

I feel that any "results" taken from them are not really going to be representative of the site. I note that meta participation is quite low even among high rep users so the results will likely on really say something about "active meta users", which is a fairly small group of users. (You also have issues with veracity, either in individuals being dishonest in their voting, or having a group of "outsiders" vote en masse to skew results.)

I tend to think that this will be mostly an exercise in navel gazing.

I, for one, will not participate in such a thread, either in proposing a poll or in voting for the options.

  • 2
    $\begingroup$ +1 particularly for "say something about "active meta users", which is a fairly small group of users." $\endgroup$
    – Surb
    Sep 1, 2016 at 12:58
  • 2
    $\begingroup$ As I said in my answer I think the activity is to be carried out in a light-hearted spirit. Re, being dishonest in the voting: an issue in many a survey in the real world, but then why would I lie when anonymously reporting on my favorite linear algebra book? (Now there may be reasons for doing this but then maybe that's sufficiently rare.) Re 'or having a group of "outsiders" vote en masse to skew results' Yeah, textbook authors around the globe might enlist their first-year students for a guerilla-marketing campaign. ;-) $\endgroup$
    – quid Mod
    Sep 1, 2016 at 14:26
  • 5
    $\begingroup$ @quid: My main point is (should be?) that if the point of these polls is to learn something about the community, then there are serious problems with extrapolating the data onto the community. (As far as outside voters are concerned, I'm mainly thinking of users from other SE sites coming in, similar to the situation with HNQs. We might be surprised to learn that the most common highest level of math education achieved for the community is "What's math?" ​ ;-) ​ ) $\endgroup$
    – user642796
    Sep 1, 2016 at 15:07
  • $\begingroup$ There is precedent on other sites. Academia.SE seems to get a fair bit of HNQ traffic. Their poll questions do not seem to be that much affected. The meta there does not seem super-active either. Still turnout on the polls is reasonable. Users do many odd things, but precious few seem to troll random meta sites by casting drive by votes on "are you a native speaker of English" polls. I mean it's just not that interesting trollery-wise. I suppose there will be that jokster that says they are native Klingon but I think we'll be able to deal with that. $\endgroup$
    – quid Mod
    Sep 1, 2016 at 16:16
  • $\begingroup$ @quid: How do you propose we deal with someone whose parents spoke Klingon at home? Call social services? $\endgroup$
    – Asaf Karagila Mod
    Sep 1, 2016 at 16:18
  • 1
    $\begingroup$ @Asaf I'd ask the Science Fiction & Fantasy mods for advice. $\endgroup$
    – quid Mod
    Sep 1, 2016 at 16:21

This "answer" isn't necessarily to advocate for one way or the other. I understand that the proposal is to create one meta thread where each "answer" is a questions and the comments to that "answer" are the options in the poll.

On the one hand I would personally be interested in seeing answers to such questions as

  • the demographics (age, professional level, ...) of the users of the site,

  • what software do you use (and maybe even why you use this software)?,

  • what textbooks do you like and use (there could be an endless number of these)?,

  • what philosophical framework do you belong to (if any)?

  • what conferences do you like/have you gone to?

    As a teacher I think it might even be helpful. So personally I think some of the outcomes of these suggested polls could be interesting.

A list of things that I personally really don't care do know about

  • Did you name your pet after a famous mathematician?
  • Do you think about math when you go to the bathroom?

Also, I can see that this might be a way to allow certain interesting questions that currently aren't appropriate for main or meta. Is the purpose of having the polls to indirectly allow questions that we would otherwise not allow?

On the other hand, I don't know if it is entirely appropriate for the site. I really like Math Stack Exchange because of the narrow focus on providing (quality) answers to (good) questions. For me, the site is about helping people advance in their knowledge and understanding of mathematics. Having been on other forums, I think we are unique in that we haven't tried to make this into yet another social media site where anything goes. Certainly having polls doesn't create this, but I wonder if it is a step in that direction.

I also fear that having polls will just create fights in other meta threads about what type of polls are appropriate.

In the end I am leaning towards opposing the suggestion, but I also don't feel strongly about it.

  • 2
    $\begingroup$ As far as I understand the idea proposed, it is about creating a single meta thread to that end, not about allowing a certain type of question. I am in favor to try the former, yet would be completely opposed to the latter. $\endgroup$
    – quid Mod
    Aug 29, 2016 at 12:28
  • $\begingroup$ @quid: I understand that it is one meta thread where each "answer" is a questions and the comments to that "answer" are the options in the poll. $\endgroup$
    – Thomas
    Aug 29, 2016 at 12:30
  • 1
    $\begingroup$ I thought that you personally understood it that way, but wanted to emphasize it. Other than that I agree with almost everything you say. Except that I arrive at a different conclusion. In particular, on "certainly having polls doesn't create this, but I wonder if it is a step in that direction." I mean, TeX - LaTeX has it since years and I don't think it affected the site negatively. One could even argue having the thread will allow to deal with some borderline questions on main more smoothly along the lines of "I can see that this might be a way to allow certain interesting questions[...]". $\endgroup$
    – quid Mod
    Aug 29, 2016 at 12:41
  • 1
    $\begingroup$ @quid: I can see that we are having similar thoughts about the suggestion. And again, I am not strongly opposed, I am just questioning the proposal and wanted to point out a few things for the discussion. I agree that if it works well on TeX.SE, then it could work here as well. We are a bit bigger though, but that might not matter. My main concern about the proposal is that it moves us in a direction that I am not sure I like. I could see a possible slippery slope. But, I would not be opposed to trying it out. $\endgroup$
    – Thomas
    Aug 29, 2016 at 12:46
  • 1
    $\begingroup$ I agree with the list of polls you'd be interested in and the list you would not be interested in seeing. I disagree with the reasons given against having such a poll thread, and the leaning towards opposing. $\endgroup$ Aug 30, 2016 at 18:41

You must log in to answer this question.

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