Vote For This Question or The Kitten Gets It
When the wordpress.stackexchange.com community asked Why are questions not being voted on …
I have noticed a trend that questions (even good ones) that have multiple answers are not being voted on.
Out of our 5,550 questions only 41% have at least 1 vote which leaves around 3,000 with 0 votes and a few hundred with negative votes.
I had a strong sense of déjà vu all over again.
One of the longest running concerns in Stack Overflow and Stack Exchange history is Why aren’t people voting for questions? a question originally posed on Stack Overflow on August 5, 2008 — long, long before we used UserVoice for this sort of thing. At that point, meta.stackoverflow wasn’t even a glint in anyone’s eye, much less Area 51 or the WordPress Stack Exchange.
So, yes, we’ve known basically forever that questions don’t get voted on nearly as much as answers.
Personally, I’m not convinced this problem is necessarily solvable, because it might represent the natural “market value” of questions and answers. Users intuit that answers are the real unit of work in any Q&A; system and tend to favor answers in their voting. After all, the world is awash in endless questions, but answers — great answers — are a precious and rare commodity indeed.
There’s also a serious workflow problem. Consider what happens when you open a question page:
-
Start at the top by reading the question.
-
Scroll down. Begin reading answers.
-
Consider the relative merit of each answer as you read it, and possibly vote on it.
- Reach the bottom, where the form invites you to provide your own answer.
By the time you get to the bottom, you’ve probably spent so much time mentally processing the existing answers and deciding whether or not you want to add an answer yourself that you’ve forgotten the question even exists! That’s a shame, because the quality of the answers and the quality of the question are often related. In both positive and negative directions, I mean. If a question is worth answering, isn’t it at least worth considering whether you should upvote it? Assuming you can remember to scroll all the way back up to get there, that is.
So how do we encourage people to remember the questions when voting? Perhaps we should institute a new policy: every time you forget to vote a great question up, or a bad question down — a kitten gets it!
Just kidding. Mostly.
Because we love kittens, we decided to make basic voting statistics a bit more visible for every user. First, in your user drop-down, you can see how many votes you’ve cast.
Second, on your user page, where we’ve broken out your voting in a similar public way.
The daily vote limit used to be 30 votes per day; we’ve increased that to a maximum of 40 votes per day — but only if you vote on a combination of answers and questions. This isn’t as significant as you might think, since it is exceedingly rare for users to even hit the 30 vote daily cap.
Most importantly, we have added a gentle reminder to the voting process itself.
That is, if you haven’t voted on at least one question in the last 15 votes you cast — you’ll now get the “you haven’t voted on questions in a while; questions need votes too!” reminder every time you vote until you do.
We also added a voters tab to the users page, so you can get an idea which of your fellow community members are truly exercising their democratic right to vote early and often.
I realize we probably won’t solve a basic problem we’ve had since inception of the network overnight. And I still believe that answers are fundamentally more valuable than questions and thus will always naturally garner more votes. But there’s no reason we can’t put our thumb on the scale to help rebalance things a tad. We’ve already seen a big increase in question voting with these latest changes, so I am … cautiously optimistic.
So please do try to keep questions in mind as you’re voting. Either up or down.
You know, for the kittens.
2 Comments
I am convinced that there is a very simple solution to the observed effect that you seem to overlook, in my opinion. It is the notion of the term “There are no bad questions, only bad answers”. Basically, this means that any human being on earth and beyond is allowed to ask any question he or she can think of. It is never “wrong” to ask. It is never wrong to speak someones mind. It is only “wrong” to give a bad answer.
A bad answer is an answer that is clearly not trying to explain something, or does not bring in new information from which an improvement or increase of value can occur. A bad answer is a personal, emotional reaction to the question which needs to be relinquished or it results in a direct personal insult towards the question author. A bad answer is a destructive answer that has nothing to do with the topic but is targeted towards a judgement to decide if a question is “valid” or not. In terms of StackOverflow this is applicable for answers to questions and comments to answers alike.
Your statistics show that most people seem to get that it would not be nice to down vote on someone who has an equal right or opportunity on posting a question as they have. But StackOverflow does not handle this right in my opinion, because the voting does not match the desired effect of measuring.
A down vote on a question should never be encouraged. This is “inhuman”. The only valid reaction on a question which the reader thinks is eligible for a down vote is leaving the question’s website, because he or she is clearly not interested in helping on solving that question. If someone comes to a question page and the only reaction there is would be “this is nonsense, makes me angry, I have to let them know to feel better” (maybe this thought is just subconsciously present), then this is proof enough that the reader is not even remotely willing to answer the question. The reader does not even have the intention on working on the topic if he uses a down vote, but the reader is only interested on working on the “question’s author”. This is plain wrong.
To my mind, the fundamental problem is the intention to try to measure “relevance” or “value” of a question with a voting system (emphasis on “of a question”, of an answer is a different thing). There is no question, whether “the question” is valuable, as I wrote above. The consequence is, that the voting on a question is futile. The prerequisite for a vote is not present. Therefore the vote should not be present. And in fact, therefore it is not used, I think.
There is however a fairly simple tweak that would put this skewed scheme into more useful lanes. It goes like this. As I stated, not the question is of question, but the reader’s intention or reaction is of question. There needs to be a tool which allows the reader to react suitably to his perception. The basic difference would be, is the reader willing to help/contribute on the topic or is he not, for whatever reason. If he is, then he should be encouraged to post an answer. If he is not, he should be encouraged to not interfere with those who are willing to work at that topic.
The tool could be: Show the question with a set of large buttons beneath (colored the same, not green and red or something, sized the same, because they are all equally valid). Pressing the one that reads “I want to help on that” opens up the section where a new answer can be posted, as known. Hitting on “I am interested in that” opens the answers in read only mode. And hitting “I am not interested in that” puts the page on the personal ignore list and the reader is not bothered by this question ever again. There could also be a “I have a question on that” and this brings up what is known as posting a comment. There could be a button which reads “I know better than the author”, which would open up the edit mode for the question itself. Any changes there, are presented to the questioner to be edited, improved and ultimately accepted or declined.
The attitude “I am not interested in helping, I am not interested in contributing, even if I could or know better” has no other valve than to cast a down vote or worse, right now. The attitude “I am willing to help” does not need an up vote, because the reader starts on writing an answer and thus already has the needed valve. An up vote would be a waste of time, double effort.
The questioner has the attitude “I want help, I want an answer”, which is satisfied if an answer comes in. There is no interest in knowing if there is someone who does not approve that question. If this is of concern, then this should be part of the question and that what was formerly known as down vote would become an answer with a value.
But what if we hate kittens?
Also, I would have thought good questions would result in good answers (all other things being equal) and so you could make a decision based on that.