This obviously will vary across answerers, but I'm curious for your thoughts. Are answerers more likely to open and/or answer the question of someone with a high reputation?

I am very new to the site, and had all my questions answered wonderfully without having any reputation. But, I'm just one example. Not being one who can answer many questions on this site myself, I'm wondering what the answerer perspective is.

  • 5
    $\begingroup$ This is a difficult question. All of the relevant data is available on data.stackexchange.com but it's not clear how to address the correlation vs. causation problem. Perhaps higher-rep users naturally ask more interesting questions, so attract more views in that way. $\endgroup$ Aug 18 '11 at 3:08
  • $\begingroup$ @Qiaochu: Thanks! I did not know about data.stackexchange.com. That's great to know. $\endgroup$
    – OctaviaQ
    Aug 18 '11 at 3:11
  • 5
    $\begingroup$ At least in my case, I evaluate the "interestingness" of a question on its own merits, not by the identity of the asker. $\endgroup$ Aug 18 '11 at 8:47
  • 2
    $\begingroup$ I would not take rep into account when giving an answer. Maybe when I see a bad question, and am ready to vote to close, if I see the questioner is a new user (low rep) my response may be different and more lenient. $\endgroup$
    – GEdgar
    Aug 24 '11 at 15:00

I would be extremely surprised if this were the case (independent of the fact that higher-rep users may ask better questions). There is no incentive I can think of that would cause anyone to preferentially answer questions from other higher-rep users (again, independent etc.). This isn't a definite answer, of course.

In my experience, there are a few other factors which are much more relevant to having your question answered:

  • Clarity of language
  • Good formatting ($\LaTeX$, spacing, paragraph breaks)
  • Adherence to the guidelines in the FAQ
  • Concision

If someone asks a good question, why should I care who it's from? Of course, higher-rep users, having been around more often, are more likely to be aware of and know how to conform to the standards above.

  • 6
    $\begingroup$ I don't think I have ever read the word "Concision" on a web page before. Odd. $\endgroup$ Aug 18 '11 at 12:07
  • 2
    $\begingroup$ The obvious incentive would be status. Mathematicians systematically underestimate psychological influences. That said, I agree with your list of more important factors. $\endgroup$
    – Phira
    Aug 27 '11 at 9:56
  • $\begingroup$ @thei: I still don't understand how this would work. I am well aware of the impact of psychological influences, but status is much more easily accrued by giving good answers than by answering the questions of other high-status users. Are you suggesting that math.SE users want to please other high-status users? I have seen zero evidence for such a motive among the active users I can think of (although high-rep users ask questions so rarely that I think it's impossible to separate out the effect of those questions being better). $\endgroup$ Aug 27 '11 at 16:27

The question does not compute. Before I open a question, I don't know whose question it is. All I know is who was the last one to modify it. And (speaking just for myself, of course) seeing that the last person to engage with a question was one of our high-rep people makes me less likely to open it, on the grounds that I'm unlikely to have anything useful to contribute after said person has already had a crack at it.


The "reputation" of the person asking the question is largely irrelevant. I am at least at the conscious level unaware of it when I decide whether or not to look at a question beyond its title.

However, reputation has an indirect effect on my subsequent behaviour. The level at which a question should be answered is often not completely clear from the question itself. It is not rare for me to click on the OP's name in order to be able to make a more informed guess about the appropriate level. Very low to zero reputation is then unhelpful, for it means that virtually no questions have been asked. Any snippet of biographical information can also be helpful.


Although high-rep users collectively ask questions that are better formed and more appealing more often than the average low-rep's, I believe I open and answer much more from low-rep users because

  1. Low-rep users ask questions much more frequently than high-rep users do, and
  2. They more frequently ask questions that are at or below my level of understanding.

I'll also echo what Gerry said about not even knowing the OPs outside the threads themselves. Hell, the best indicator of an asker's rep level before opening a question is the title itself. We're here for math; social activity is tangential.

(That said, I do think high-rep users often let lower-rep users take the low-hanging fruit so that work ends up distributed according to ability and interest.)


The best way to sort out correlation versus causation is by a random experiment. Let some high rep user who generates a lot of questions pre-commit to, before the next 20 questions they ask, flip a coin and ask either under their own login or as anonymous. I did this with answering questions here and concluded that, if there was an effect, it was smaller than the variation within answers.

Unfortunately for the proposed experiment, very high rep users don't seem to ask a lot of questions! Among the 10K+ crowd, I think no one has asked more than 30. This is an interesting contrast to MO -- I believe MO started in Oct 2009, so 22 months ago while math.SE started in Jul 2010, so 11 months ago. That would suggest that high rep MO users should have roughly twice as many questions as high rep math.SE users. In fact, the ratio is much higher than that. I didn't collect data carefully, but it is quite common for a high rep MO user to have more than 60 questions, whereas a high rep math.SE user never has more than 30.

If someone down around the 3K or 5K range wanted to do the experiment, it would probably still be interesting. Any volunteers?

  • $\begingroup$ But the volunteer shouldn't speak aloud here, should he? $\endgroup$
    – Rasmus
    Aug 23 '11 at 21:01
  • 1
    $\begingroup$ While in principle you are correct, Qiaochu asked 32 questions and myself 26 thus far. $\endgroup$
    – Asaf Karagila Mod
    Aug 23 '11 at 21:49
  • $\begingroup$ Another important factor is that upvotes on questions generate half as much rep on math.SE as on MO, which means that even if the distribution of asking/answering were exactly the same, MO would have more high rep question askers. $\endgroup$ Aug 23 '11 at 22:40
  • $\begingroup$ Here's a SQL query with which you can explore (roughly two months old) data relevant to your post: data.stackexchange.com/mathematics/s/664/power-askers $\endgroup$
    – t.b.
    Aug 23 '11 at 23:09

You must log in to answer this question.

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