3
$\begingroup$

I use several Stack Exchange sites, with Stack Overflow being the most usual for me. I strongly agree with the points threshold (needed for every action in the Stack Exchange sites). I've never had trouble earning enough points to allow me to do various things (down vote, comment, edit). I earn most of my points because of the questions I ask. I don't think that I ask too many questions, so I think I've got good balance so far.

I have noticed that here on Cross Validated my questions (and many others that I see) get few views and also few upvotes. As far as I understand the norms of the site, if the question shows it was well asked and it deserves attention, you should upvote. I try to follow that when I read other's questions. I understand that Stack Overflow might have 500 times more users than Cross Validated and that might account for the difference. Also, given the characteristics that questions about statistics typically have (they need lots of comments to understand what the OP was referring to and if they tried this or that), I believe that many questions end up with a lot of discussion in the comment section and zero answers.

Finally, I am astonished by the diversity of the questions being asked here and the amount of time and knowledge that is needed in order to approximate to a solution. I believe that we need tons of new members in order to cover so much variability. The statistical knowledge of people asking here is also very variable (I consider myself as one who is just starting to understand and apply it to my research but I've seen some questions that fly really high). So I wonder how to raise the happiness of the users. Questions come in at a much higher rate than the answering rate. I have asked some questions that, in spite of having upvotes, do not get much attention. My only question that did grab some attention was 99% about programming in R (in fact was migrated from Stack Overflow).

I have found questions I want to comment on to get further into the analysis and help but I can't get the points needed (again, first you need points for question or answers and it seems a bit like a vicious cycle). All these situations keep me wondering if I'm the only one that is having this concern.

$\endgroup$
  • $\begingroup$ I upvoted three of your questions. Hopefully you will soon get across 50 rep barrier and will be able to comment. Personally, I find this 50 rep limit quite unnecessary. $\endgroup$ – amoeba Jul 15 '15 at 20:11
  • $\begingroup$ I have noticed (thank you). What do you think about the discusion I tried to start? $\endgroup$ – Matias Andina Jul 15 '15 at 20:13
  • 8
    $\begingroup$ You raise many issues, but note that your (lengthy) post does not have a single question mark. What exactly are you asking? $\endgroup$ – amoeba Jul 15 '15 at 20:30
  • $\begingroup$ My concearn is mostly around the low rates of answers in Cross Validated compared to other Stack Exchange sites. If found some other meta posts regarding to this issue. I'm just wondering if there's any way to tackle that by promoting different scoring and/or new member policies. $\endgroup$ – Matias Andina Jul 15 '15 at 20:50
  • $\begingroup$ Several of these issues have been much discussed on Meta. The most distinctive feature that results in low average votes here is, arguably, a large number of rather poor questions from people obliged to do data analysis but who lack the time, inclination or ability to study statistics in any depth. One does see some questions on Stack Overflow that amount to "I have no code, so please write some for me" but questions showing little or no research are rather common here. I don't understand your question about reputation, however, as anybody can answer! $\endgroup$ – Nick Cox Jul 16 '15 at 0:16
  • 1
    $\begingroup$ The threshold for commenting is supposedly to cut down on poor quality comments, including offensive or spam remarks. One argument in favour is that it apparently does work! $\endgroup$ – Nick Cox Jul 16 '15 at 0:22
  • 3
    $\begingroup$ We certainly can't control the scoring; it's the same across the SE network. I do notice that some sites are more aggressive about closing or even deleting what I'd regard as rescuable questions, which presents an unfriendly face to newcomers who often need help here at a technical level far beyond their ability to clearly ask for. There's the fact that each application area uses different terminology and focuses on different techniques, so unless you know what every area calls something (will the average econometrician know that Jarque-Bera is usually called something else among statistcians? $\endgroup$ – Glen_b Jul 16 '15 at 0:52
  • 2
    $\begingroup$ ... Or the average med stats person know that Bland Altman is just a Tukey mean-difference plot? Such splintering of jargon leads to mutual incomprehension). It's also relatively easy to ask stats questions that nobody has good answers for yet. We could cut down our unanswered questions by closing some old ones that have no prospect of good answers. One thing we could really use on the site is more regular active answerers with the reputation to engage in helping improve questions (across a variety of application areas), but I don't have a good idea how to achieve that. $\endgroup$ – Glen_b Jul 16 '15 at 0:52
  • 3
    $\begingroup$ For your issue of few votes -- some of your questions are not clearly posed, or unclear. In the case of your recent question, it's quite a vague question, and there's also an issue with the premise of the question -- to even get it to a state where it's really asking something well answerable would require a lot of effort. I had intended to come back to that question when I had time to discuss the issues in comments, but don't always manage to find time before still more things come up. $\endgroup$ – Glen_b Jul 16 '15 at 1:42
  • $\begingroup$ I agree 100% with you, and I understand that the questions in this site are difficult to answer and also difficult to ask because they need a lot more specificity than the other questions on SE network. They usually get into a lot of comments and everyone lose track. $\endgroup$ – Matias Andina Jul 16 '15 at 3:52
  • 6
    $\begingroup$ Another problem is that it seems harder to ask a good and answerable statistics question than a good and answerable programming question $\endgroup$ – shadowtalker Jul 16 '15 at 13:10

You must log in to answer this question.

Browse other questions tagged .