Not sure why, but lately I've noticed that many questions are being downvoted, not because they are "off-topic," but because the CV observers felt the question wasn't clear enough. Is this sort of knee-jerk, wholesale rejection good practice? Is it discouraging to potential participants who may not be well-enough informed to ask a statistically lucid question but are still quite willing to learn from more structured feedback? Is this shaping the demography of this forum away from basic questioners to the highly sophisticated user?

Here are some examples where I felt the trigger is, perhaps, being pulled too fast:

All of them are undeniably vague, unstructured and confusing. However, why should all questioners be expected to cogently articulate a "dissertation-worthy" question? Is it the role of more experienced CV participants to act as "stats police? Or is it better to act more "Socratically," in helping naive participants elicit and better articulate their concerns?

I would argue that the latter role is more rewarding for both the questioner as well as the participant as both can learn from each other's ignorance.

  • 6
    $\begingroup$ [I've deleted the entire exchange here, even the good comments (and those below one of the answers which thread also went in a singularly unproductive direction). I don't think the kind of behavior seen here has any place on stackexchange; indeed now it's closed as a duplicate I'm inclined to remove this post altogether. If anyone wishes to behave in this sort of fashion I suggest you spend your time in a forum where that's accepted. This isn't it. If you can't be considerably more constructive than was displayed here by several users, be elsewhere.] $\endgroup$
    – Glen_b
    Commented Apr 7, 2017 at 2:37
  • 3
    $\begingroup$ If anyone wishes to behave in this sort of fashion I suggest you spend your time in a forum where that's accepted. This isn't it. If you can't be considerably more constructive than was displayed here by several users, be elsewhere. Well said. Lots of unconstructive comments in that thread from several users. It's too bad it escalated as far as it did. Deleting the comments was the right decision, and I doubt much useful content would be lost by deleting the question altogether. $\endgroup$
    – gammer
    Commented Apr 7, 2017 at 5:20
  • 2
    $\begingroup$ @Glen_b Gosh, yesterday was one rare evening when I decided to go to bed early without reading CV - and now it turns out I missed one of the so rare comment fights! Sad. $\endgroup$
    – amoeba
    Commented Apr 7, 2017 at 7:12
  • 3
    $\begingroup$ I also noticed a considerable increase in downvoted (usually 1 per post) on first time posts. Some downvotes are normal, and you don't have to post a comment to explain why you downvoted. However, I think on average newcomers should be treated more gently. If the community gets a reputation of being harsh to new users, this won't be good for the site growth. $\endgroup$
    – DeltaIV
    Commented Apr 7, 2017 at 11:54

2 Answers 2


Interesting and important question. I don't want to add specifically to what I and more crucially many others have said on voting here. There are many such threads. We're not long out of Ganging up with downvotes. How can we discourage mob behaviour? I hope it is fair to say that almost none of the participants in that thread are likely to feel that it went well or pushed anything forward. A minority of those posting feel very unhappy at some aspects of voting on CV. Even the rest should not be happy that those members are unhappy, regardless of not sharing in the gloom. But: to restate what is not a personal opinion, but a fact of SE: there are grounds for downvoting and anyone with enough reputation is at liberty to exercise downvotes. That's not to say that all downvotes are appropriate, but who decides and how do they know? Political allusions are tempting here, but anyone reading this will be smart and well-informed enough to think of their own.

Your general case doesn't stand or fall on your examples, but that's your specific evidence, so I will add some comments.

At the time of viewing each question I see this. Clearly each thread may change, not least because of the Meta effect of being mentioned here.

PCA 1 vote up, 1 vote down. 2 answers, reasoned comments. What's to remark on?

Regression for trend 1 vote up, 1 vote down. Reasoned comments in response; the OP has yet to improve their question. The question is not obviously doomed, but could go either way.

Uber question 1 vote up, 1 vote down. Put on hold. @DJohnson was among those voting to put on hold. I didn't participate, but I have to agree with the decision, although I am puzzled by the stated grounds. The question is in my view off-topic and/or too broad. But it is often true that a question that won't work well here gets close votes on several different grounds. The OP did receive some reasoned comments.

Harsh, like rude or sarcastic, is what the other guy is said to be, and virtually always unnecessarily so. Everyone knows irregular verbs of the form: I make fair comments; you are sometimes contentious; he and she are often downright obnoxious. But I have no axe to grind as defending any votes, comments or answers on the stated threads. I can't see anything here there that is harsh, or uncivil, or inappropriate.

The downside of CV is the downside of any technical forum I know, a long tail of questions that don't seem especially clear or interesting or answerable to many of those active in answering questions. Old story this, but no one is paying us to do this or obliging us to give our time this way, or to answer what we don't want to. Moreover, there is not much value for the forum in a large fraction of threads that do not or will not get anywhere far for the benefit of others. I don't disparage sympathy for those who might be discouraged by zero or negative responses, but a forum like this is not a help line. No one is entitled to an answer.

Socratic dialogue worked really well in the Socratic dialogues. It doesn't usually work well here. CV is in a mature phase where it's actually quite hard to ask a good new and specific question. It's also vulnerable to the postings of those who will not read or do not understand the generic advice displayed (e.g. on software-specific questions). I wish it weren't so.

  • 6
    $\begingroup$ More than 10% of the questions here have the tumbleweed badge, mainly given for low views (<20 in a week or so). That shows the main problem: Not enough eyes to really follow up on all the questions. $\endgroup$ Commented Apr 6, 2017 at 21:44
  • 7
    $\begingroup$ @kjetilbhalvorsen Good point. I have no idea how to persuade anyone to contribute here if they don't independently think that it will be (in no particular order) fun; educational, as helping you think through a question and learn from other answers; ideally of some help to others. Otherwise put, whatever satisfaction people get by contributing is essentially personal, visible only here, and not of (say) career benefit. $\endgroup$
    – Nick Cox
    Commented Apr 6, 2017 at 21:51
  • $\begingroup$ It seems to me that some of the other StackExchange have downvoting problems much worse than ours, exemplis gratis, and by comparison most (but not all) cross-validated participants are gentle by comparison. Nevertheless, downvoting is a negative, and occasionally I review my downvotes and retract some. Interestingly, the higher my reputation has become, the lesser downvoting seems to be problematic. Is there an ugly patch of reputation around 1k-5k where it matters more? Why? $\endgroup$
    – Carl
    Commented Oct 21, 2018 at 3:58
  • $\begingroup$ Could be that with some or even most of us, the longer you stick around, the less surprised you are by lousy questions or answers. I've shifted a little to more emphasis on explaining what's wrong; and to voting to close as a reaction to the content not the poster. I am occasionally happy to pile on with a "Me too!" downvote if and only if the poster is spamming or being very rude or otherwise obnoxious. Also, asking numerous different versions of the same question is a personal no-no. That wastes too much time and effort from people who answer. $\endgroup$
    – Nick Cox
    Commented Oct 21, 2018 at 9:11
  • $\begingroup$ This thread remains striking. The OP answered and accepted their own answer, which is always allowed, but the voting went the other way. $\endgroup$
    – Nick Cox
    Commented Oct 21, 2018 at 10:36

Thanks for commenting. It's always nice to see @nickcox stepping up to field and adjudicate queries on behalf of other senior participants with his usual considered and, for the most part, fair points.

To his point about "irregular verbs," I would only add that the syllogism, "I make fair comments, you are sometimes contentious," is an illocutionary statement representing a ubiquitous human failing: to wit, everybody thinks they are correct, right, even-handed, objective, fair, whatever -- everybody. This is true even for the most vicious and psychotic trolls that inhabit the universe of online social media. The obvious, implicit assumption here is that one is speaking from an authoritative position -- a highly dubious assumption.

And to the point about "the long tail of questions," a moderator once noted that as much as 30% of questions posted to CV go unresponded to and/or unanswered. There are many reasons for this but that's not my concern. In my view, the real "long tail" here revolves around nursing and rehearsing hoary grudges, the revelatory descent into which does little or nothing to ameliorate fraught relationships. Just so, baiting and/or goading participants into arguments serving no purpose other than scoring passive-aggressive points may be "what the other guy is said to be," but such comments absolve their speaker from any personal responsibility or involvement in perpetuating undesireable communications. While a few participants get sanctioned, "tarred and feathered" or run out on a rail for extreme behaviors along these lines, this is little more than scapegoating -- a true "bandwagon" effect. Not to mention that it's well known (i.e., readily observable) that more than a few senior CV participants stoop to openly, even frequently, venting their impatience and irritability with other, less senior participants without sanction. As noted, such are the downside hazards of any forum, technical or otherwise.

Coming back to the main point, the concern expressed in my query has to do with knee-jerk votes for closure simply due to the fact that a question may not be up to some participant's high standards of communication wrt statistical issues. The examples provided were an effort at illustrating the ubiquity of this phenomenon. The Uber simulation question may be one of the best examples as the back-and-forth in the range and diversity of opinions for closure is quite explicit. @Tim does the best job of articulating a view in favor of retention.

My (probably unrealistic) hope is that, by shedding light on such ill-considered behaviors, some blunting of their momentum may take place.


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