Why aren't people voting for questions?

Question

While at least some answers from every question seem to have been up-voted as a "helpful" answer:

• about 49% of questions have score of 0 or 1.
• another 21% of questions have a score of 2.

Most questions seem to have an answer or two with score +5. Is it that we have a lot of fluff questions and not useful ones? Or perhaps it just seems more natural to vote for answers than questions? If we want the most useful info on Stack Overflow to filter to the top, how can this disparity be fixed?

I think this may cause real problems for people getting the badges which require +25 or +100 votes on questions.

---

I like a lot of the feedback I'm seeing. Two problems are highlighted:

1. The work flow doesn't promote voting for questions the way it promotes voting for answers. One solution: when a user answers a question, the system should invite them to vote for the question. If someone cares enough to answer a question, then they care enough to vote for it as well.
2. Users with less than 15 reputation cannot vote for questions or answers. Is this warranted? Maybe these users could provisionally vote for questions and answers and those ratings would only take effect when the user reaches 15 reputation?

All I ask is this. As you browse Stack Overflow, when you see a good question: vote for it.

If everyone does, then this problem will disappear.

Answer 1

I would have up voted several things in the past 20 minutes if there wasn't a silly reputation restriction on up votes. I can understand down vote limits, and an can understand a probation period, but it shouldn't kick in automatically it should be applied when your reputation drops, not when you're first starting off.

Now the initial impression is "Why did I just register? I can't do anything." It will also lead to far more duplicate answers because I can't easily say "I agree with this existing answer." I either have to post my own 90% similar answer, or say "@johndoe I agree" which doesn't elevate the answer in the rankings, requires the reader to parse and scroll back to see what johndoe said, and gives johndoe no benefit from their good answer.

This post is a perfect example of this, lots of other people have said "Vote Up shouldn't be restricted" but I have no way to reinforce their point without you reading all this garbage I just wrote...

Answer 2

The answer ratings have a graphical interpretation. The little up-arrow pushes the answer up the score-scale, and thus up the page; and the down arrow pushes it down the score-scale, and thus down the page. That a question has a score of 20 vs 200 vs 2000 doesn't tell me anything, as there is no implied comparison with other questions. I have to think about the meaning of the question voting widget every time I see it, as it does not intuitively mean anything.

I would suggest that the question's rating (for whatever purposes that rating has) be inferred from the answer ratings. Maybe the question score could be the sum total of all answer votes.

Inferring from the number of answers is less attractive, as some answer sequences are actually conversation threads, and do not reflect on the quality of the question.

Answer 3

One of the fundamental issues that doesn't seem to have been addressed seems to me to be pretty fundamental: There seem to be no incentive for upvoting questions, so why would I bother?

In practice, I do go out of my way to try and upvote questions, but I seem to be in the minority.

At first, I did get some value from upvoting because I could then find interesting questions again through my personal page (that's how I found myself behaving: "upvote = mark this question so I can find it again later"). Now that "favourites" have been enabled, I no longer even get this benefit.

So basically, the only motivation I now see for upvoting is some nebulous "do unto others" kind of philosophy.

That may suit me, but by embarking with the whole reputation system, stackoverflow has marked itself as a different beast. Live by the rep, die by the rep.

That's not a bad thing, just the way things are.

SO is a rep-based economy. Want people to upvote questions? Easy: attach some incentive to it.

Answer 4

Give a separate badge for rating questions.

Answer 5

I see a down vote as indication that the question is a "bad" question, rather than a "stupid" one. There are no stupid questions, as has been stated. However, there are bad questions --- insufficient information, poorly formulated, no attempt to solve the problem self, no research into possible solution, please do my homework for me, and so on. These type of questions are distinguishable (I believe) from those indicating that the questioner is struggling with the concepts and does need some help. A down vote prompts the questioner to put some effort into finding a solution and researching the problem a little before asking the community to bail him/her out.

Answer 6

This is being overthought.

I call Benford's Law on the voting distribution.

Answer 7

Sorry about the sad truth: I skip the question entirely and scroll straight down to the top answers. More often than not, I hit jackpot on first try, as the answer seems to address my question perfectly fine*.

It's a TL;DR feeling. Do you reread your Google/Bing search queries before you start scrolling down the results?

*Yes, on rare occasions I might check the question for some details just to verify if indeed it matches up to my issue.

Answer 8

Couldn't it automatically infer the value of a question by other metrics like # of answers and # of up votes on answers, # of views, etc.

Answers that are voted up are usually deep answers, but they typically require a deep question. Don't they? Let's turn to the DB to find out.

Answer 9

From a design standpoint I reckon you should have the question rating on the browsing pages. It I don't like a question I'm not going to bother with opening it and then rating it. I'd rather have a "bury/promote" function in the question listings themselves.

Answer 10

Who comes to Stack Overflow with the idea that "I really need to know something, but I'm not sure what it is. I'll look at the top-voted questions and find out what question I need answered"?

No one.

You search to find questions you need. You tag topics you are involved with and watch for your tags on the front page.

When you've found the question you need, you read the answers, paying special attention, perhaps, to the accepted answer, to answers that have lots of upvotes, and to answers from people you recognize, or who have high reputations.

To me, it's silly to compare stats of the questions and answers, because they are apples and oranges.

Answer 11

Perhaps by the time you've finished reading a really good question, the voting controls have been scrolled off the top of the page.

My rule of thumb is, if I found a question with an answer that is perfect for what I needed to know, they both get a vote up, unless the answer is already accepted (I know), AND if I am answering a question, I definitely vote the question up. It's worth something to have a clear, focused, and answerable question. Anything I can actually write an answer to falls into that description.

Answer 12

Upvoting questions should be reflecting that people who ask questions which lead to good answers are contributing to the knowledge base of the site. This currently is not working well.

This seems to be a problem with all such sites. I was just over on a site based on a model very much like SO. This site is still very new so it is not being flooded with questions. I asked two very basic questions (the type that should be in any wiki about the subject) neither had been asked on the website (and before someone comments they happened to be questions I didn't know the answer too). Both questions received answers and those answers received votes.

Someone wrote that often the question is not useful to them so they don't upvote it, but that the answer to that same question is useful, so they upvote the answer to that same quetion. Is that really the system working as intended? How wasn't the question helpful in that it was what led to the helpful answer?

Reasonable solutions I can think of:

Encourage those answering questions to upvote with a reminder if they have not already. Perhaps even a suggested list on how to evaluate questions (question is of wide interest, question adds to the technical knowledge base of SO, etc).

Alternatively, all answers should give some value (even if less than an actual upvote). Given that questions can be closed (which should remove points given for that question) and people can use comments (which probably should be used far more then downvoting).

Another approach would be when someone upvotes an answer to have the fly out ask about and allow voting on the question.

Yet another that answers who do not rate the question do not receive full points for their answers.

Answer 13

I try vote everyday on questions of the topics I find interesting (opencv tag) so people get encouraged to keep asking. Also I think it is good to vote (as the person who asks) up for those answers that try to help and are correct even if they are not really helpful or not the answer you would consider perfect.

I think that people in general is kind of greedy with voting and there are several good questions/ansawers with 0 votes that probably not many people will read.

Answer 14

I know this might sounds a bit odd to some of you, but I didn't really see a place to vote.

Yeah, there are big honkin arrows, but they didn't jump out at me.

I kinda expected to vote on the question on the main browse page too, as done on other sites.

My suggestion, add + and - to the arrows. Yes, the Department of Redundancy Department.

Answer 15

It might be because everyone's too busy using their votes on "threads" like this one. I think maybe we're trying to express two things with the question voting system, both that a question was a good one that everyone can learn from, and that a question needs attention from the community. Toss in the notion that lots of people are using votes to recognize the fun factor in a subjective question or to reinforce their opinion while "answering" such questions, and you can see why questions maybe aren't getting their due.

It is instructive to note that at the time this was written, the only two gold-badge level questions involve how to use this site (shouldn't that be in the faq?), and a "what do you think?" that's taken place on nearly every active board in the world. Two of the "Questions" filters are already drowning in noise.

Answer 16

I like the reputation restrictions. They keep me from making strange misinterpretations of concepts like 'down' and 'up' until I've had a chance to assimilate more of the culture here. Spending more time reading, rather than clicking, improves my feel for what each click would mean.

Rating based on 'useful' makes sense to me for answers, but not so much for questions. A thread might be useful, but the question itself? Well, the most I could say about one ordinarily is that it was interesting or thought-provoking.

Answer 17

Just posted a dupe question to this one (accidentally) so here is my question text as I am going to delete the duplicate:

I have noticed that when going into questions there can regularly be several answers to a legitimate question. The question is thoughtful, useful, well presented and yet has garnered no up-votes.

Surely if you are answering a question, then your first thought should be to upvoting it. I mean its in all our best interests surely, because the natural SO process will ensure that good questions feature prominently.

Nobody seems to think twice about upvoting 'Subjective' questions, which in my opinion (although I am guilty of some too) should not be applicable to upvoting, only to favourite<-ing> (Have I shot myself in the foot here with this highly subjective question?)

Its costs nothing to up-vote a legitimate, well-asked and useful question so why are people so reluctant?

I guess that this is the best place to put this...

Answer 18

Maybe instead of up/down voting, there could be a ranking for how general the question is? For example, the likelihood of others finding it useful.

I never vote on questions because it doesn't make sense to me. Which question is "better?"

• How do I handle mouse clicks on a QTableView in Qt?
• Why do I have to set my static variables outside of my class in C++?

I just see them as two different people having a frustrating time with whatever their current language is.

Answer 19

I will add in my own experience using Stack Exchange sites. Many sites (unpopular ones in particular, I primarily use the one related to natural sciences) have declined in both popularity and in terms of creating an encouraging environment because it has become too "strict" in terms of rules and regulations.

While some questions clearly deserve a downvote/closing (like homework questions). More often, a lot of curiosity-driven questions that are exploratory in nature get discouraged. Many users (including moderators) under the "influence of rules" do not seem to understand if a question addresses a "new" or "unusual" area, it is meant to be a bit broad and exploratory in nature. It also does not mean that it will be harder to answer, as being "open-ended" can attract very diverse and interesting answers. I saw a number of questions that would have the potential to receive tens of upvotes 5 years ago would now gone ignored/closed.

The end result is that it creates an environment where questions are strictly limited to a narrow focus and is well-researched already. Being specific means it is less likely for people outside the field of the narrow focus of the question to be hooked and therefore voted. Myself included, I have accustomed to such culture to the point that I simply "forget" to vote many times. Worse is that, it actually drives away users hence the lowering of popularity.

Answer 20

I upvote questions when I see that someone made his own thoughts about it, and he/she was searching before. I can see it if relevant information is present in the question. Whether or not the problem affects me doesn't make any difference to me.

And I take the chance to learn something that is new for me.

I sometimes have the feeling, if English is not your native language, and you cannot express yourself well, you are lost.

Answer 21

I hate to be Mr. Pessimist here, but I don't think voting will ever become a key component of Stack Overflow. The nature of the site is one in which voting is a filtering mechanism, and a way to help "face" content. However, I have no problem with that, and I actually prefer it. The goal of this site in my eyes, is instead to be a quick (but accurate) resource where programmers can ask questions, and get questions answered. It should also have a robust search feature, and be highly optimized for SEO so that programmers can use it as a long-standing and reliable resource. I really liked Yahoo! Answers for that reason. It was Q&A, plain and simple.

I really do hope that developers come here and have fun and participate because they enjoy the community, and enjoy helping others learn. If the elitism is absent, and a general notion of common sense and friendliness is apparent, this site will be a success.

Answer 22

Maybe a question should not be voted per se, but have a kind of ranking based on how many and how high scorings answers it has.

Answer 23

The answer is easy for me - the site lets me vote on an answer in the same "place" that I read it, but I have to navigate to a new page to vote on a question. When I read down the long lists of questions, I see plenty that are obviously good questions, but since they aren't in my area of expertise, I don't bother opening the link, and I don't bother voting for the question. On the other hand, when I see a good answer, one click votes on it, and without taking me out of my "reading flow".

Add the "vote up/down" affordances to the question list pages, and I bet you'll see question voting take a substantial jump.

Answer 24

As a possible solution, what about giving reputation for the number of views of a question to the question poster? Then again, bad questions could get a lot of views too. A multiplier on the current score, perhaps? But then this would make the system dynamic. Although you could just take into account the score at the current time when a view is made, and use fractional reputation scores which are rounded up / down. It's a tough problem.

Another solution might be upvotes / downvotes carrying more weight from users with more points.

Answer 25

I don't know why people don't vote, but one way to get people to vote more would be to provide a badge incentive that requires a lot of involvement. One suggested would be to extend Vox Populi - Used the maximum 40 votes in a day with

Max Vox Populi - Used the maximum 40 votes in a day 100 times.

or a somewhat better yet even worse suggestion:

Vote Distribution - Used the maximum 40 votes in a day 100 times on non-consecutive days.

Answer 26

My idea is to remove the up/down arrows for questions completely, and completely replace the functionality provided by the 'upvote' arrow with the already-existing 'favourite' mechanism that allows you to keep track of posts, but would now give upvotes to the user (some kind of multiple perhaps), providing the question is not flagged and removed.

Unique SO user view count may also contribute to score. The flag mechanism would provide the 'downvoting' functionality, and removed questions would contribute negatively to score.