We've talked a lot about closure lately, but at the end of the day, active voting is the most important assumption behind how the Stack Exchange network works. SE sites breathe votes and reputation: Good questions and answers are supposed to be upvoted, while poor questions and answers are supposed to be downvoted, closed, or deleted. A healthy SE site has a healthy "voting economy."

What's our voting economy like?

Here on CogSci, I've noticed that we tend to have a fairly flat distribution of positive votes, and that we are rather conservative about downvotes. The last time a question hit a very high score (20+) was in 2013. In all of 2014, only 14 answers earned 10 or more upvotes and 9 answers earned -3 or more downvotes (the level at which they become greyed out). In the nearly four months of 2015, only 4 answers have earned 10 or more votes, and only 9 answers have earned -3 or more downvotes.

Of course, this isn't and isn't supposed to be hard evidence of the problem I propose, but I think it's enough to start a discussion about CogSci voting. We may not have that many users, but we don't have that few, either!

Do we have this distribution because we simply have very few voting users, or our voting users don't vote, or they vote within a relatively narrow set of categories? I feel like good questions and answers should be hitting 10 votes (Bronze "Nice Question" / "Nice Answer" badges) quickly and regularly, as should bad answers should hit -3 (greyed out), but my impression is that few do in practice.

How and how often do you vote?

  • Nice queries ... but where does the 'expected' amount of votes come from? Perhaps worthwhile comparing to queries on other sites instead (normalizing based on user base).
    – Steven Jeuris Mod
    Apr 23 '15 at 21:08
  • @StevenJeuris The expected amount of votes come from SE features and badges (bronze question/answer and greying-out). Apr 24 '15 at 5:09
  • 1
    I don't think inter-site comparisons are all that relevant here, since SE features are what they are regardless of how we might compare. Also, they're not practical (i.e., I see lots of suggestions for such comparisons on old Meta posts, but such suggestions appear to be discussion-killers more than anything; the people who propose them never actually go do it.) It's not about how we measure up, but about conveying the right information to readers within the SE framework. Apr 24 '15 at 5:20

I vote up if question/answer contains new information, or makes an effort to support its assertions instead of simply assuming them. Particularly if it shows clear and traceable evidence of prior research.

I vote down if something is entirely anecdotal/speculative or begs the question with large-scale or multiple unsupported assumptions. I find myself irrationally reluctant to downvote things when my own rep ends in 0, but a bad enough question/answer can always overrule that.


I try to use all of my votes every day, or as many votes as time allows for. I use the 'front page' to locate new questions, and search filters for new answers with no votes to catch any stragglers.

On topics where I consider myself reasonably knowledgeable, I tend to up- and downvote based on a detailed evaluation of the post I am voting on. For questions, it matters to me whether its context is well-defined, and whether or not the variables have or can be operationalized in the broadest possible sense of those unfortunately technical terms. For answers, it matters whether I consider it to have sufficiently answered the question, of course.

Since the set of topics where I consider myself reasonably knowledgeable is relatively narrow, however, I use what I call the learning criterion outside of those topics. For both questions and answers, I ask myself whether I understand what the question was about, and whether I am more knowledgeable about its topic(s) than before reading the question and its answer(s).

In either case, I almost always vote, abstaining only when the post is both well-formulated and entirely beyond my understanding (typically nuts-and-bolts neuroscience questions).

  • 1
    How many votes do we get a day?
    – Krysta
    Apr 24 '15 at 12:56
  • 2
    @Krysta 30 votes per 24 hour period
    – Josh
    Apr 25 '15 at 3:10

You must log in to answer this question.

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