4

Disclaimer: Although I don't feel it has particular relevance, I am a new user here and haven't experienced the full scope of everything that's happened. These are just some of my own personal observations which may or may not be fallible.


Lately I've been noticing a trend of downvotes being issued on a few questions; a lot of which seem (to me, at least) to be thoughtful that fit well within the model of our site. This is not just a few random downvotes that appear on a question, but a massive ratio.

Example 1

Here's a particular question that's relatively recent. It has an up to down vote ratio of 0.857 (6 to 7).

Why is this question highly divided? If it were poorly written and just not a fit for this stack, the votes should reflect a much higher amount of DVs compared to UVs. Such feedback gives just about no information to the OP... how are they to take that input and alter their question so they can improve when the feedback is on both sides?

Example 2

Let's take a look at another example of this. This has an up to down vote ratio of 1.250 (15 to 12).

Again, this question is highly divided in its votes. The comments actually bring up the idea it might be too broad (feedback) which might better explain the reasoning for the downvotes. But then again, 15 people voted "Yes! This is a good question."

Example 3

Here's a third example asked by a new user. Ratio up/down votes is 0.667 (2 to 3).

In this question there are no comments describing why the user's question wasn't deemed 'good'. This is especially noteworthy given that it was a new user asking. They also haven't asked a question since (which may or may not be important).


Analysis

There are many more examples out there I haven't mentioned. The main takeaway I've gotten from this: it seems that voting has very loose definitions.

In several of these questions it seems that people have cast a downvote simply because they had personal misgivings for the OP and their situation.

Why do I say this? Because reading the comments and answers themselves seem to reflect this idea. There's conflict in these responses arguing over the OP and their goals. This is especially evident in questions which encourage frame-challenge answers (like the first and third example).

I've even personally received negative comments challenging my intentions simply because a user thought ABC thing I did or wanted was 'stupid'. Some contributions have little effort and thoughtfulness and aren't helpful at all. I think because many of these responses can't be given (they'll be removed) they default to using voting instead.

Voting should be indicative of how well the question was written, researched, and aligned to site goals. I feel there's too much of a personal involvement in how users vote and comment. The ratios in those examples greatly reflect this. Questions that encourage frame challenges especially; too many are downvoted or argued on because people think it might have an 'obvious' answer. I argue it doesn't matter, as long as it follows those three 'good' question guidelines.

For newer users this is especially concerning; we want to keep people around and involved in our community for time to come. Even more though, we want to be constructive, polite, and encouraging to all users asking questions.

Conclusion / Discussion

I personally feel we have (as a community) a few things to work on. I think that quite a bit of feedback has no substance to helping a user ask better questions. Frame challenges especially illicit more personally threatened and direct responses; which, I feel would discourage people from asking. They might think:

What if my question get's people riled up because I did something wrong in their eyes? Maybe I shouldn't ask (a potentially good question).

Here's some things to think about in essence:

  1. Is voting as loose as I've described? Do these severely symmetric voting ratios matter?
  2. Are users receiving 'good' feedback (especially for frame-challenges) they can use to improve their experience on the site? Are the comments for these types of questions often rude?
  3. How do you vote for a question?
| |
  • 1
    I just took a look at questions still open but with a negative score, I find it crazy. Current link (this link will stop working in the future since the number of questions will increase and that the page 150 will no longer be the last one) – Ael Feb 28 '19 at 9:46
  • @Ælis That doesn't look too bad to me. We only have 17 questions that have a negative score and are still open. We might want to put those under scrutiny individually (do they need closing/deleting?). But I can see at least 5 written by usernames that are familiar with regard to writing problematic questions, so I'm guessing these may be closed eventually. Looking at other exchanges, it's also not that uncommon to have questions scoring negatively but still being open. I'm guessing this isn't saying much, and we shouldn't be scared by having negatively scoring questions. – Tinkeringbell Feb 28 '19 at 11:55
  • @Tinkeringbell I took a second look and some of them are, indeed, close worthy (bad I don't know if they were at the time they had been asked). But, in any case, I still struggle with the idea that something can be open (so, is a good fit for the stack) and have a negative score (so, it's not really a good fit?) – Ael Feb 28 '19 at 12:00
  • 1
    @Ælis interpersonal.meta.stackexchange.com/a/1400/1599 (I linked that in my answer too). Sometimes people do the comment and downvote part, but don't vote to put on hold (or simply don't have the privilege of close-voting). Other times, questions are just 'unuseful'. It's a bit hard, but not every reason to downvote has a 1-on-1 relation with a reason to close-vote. 'Unuseful' doesn't really have a counterpart when voting to close. 1/2 – Tinkeringbell Feb 28 '19 at 12:07
  • So yeah, a question can be downvoted and still open. A close-vote is not a super-downvote. Open just means a question isn't too broad, opinion based, unclear, off-topic or a duplicate, that there isn't enough reason to close a question for any of those reasons, but there still is something about the question that doesn't make it a perfect fit for IPS, that makes it an example for other question writers, that makes it useful for future visitors of the site. 2/2 – Tinkeringbell Feb 28 '19 at 12:10
4

It is important to remember that downvotes do not necessarily correlate with whether a question is a good fit for the site. Consider that if downvotes were simply a measure of fitness for the site then there would be no point of them because they would be the same thing as close votes.

In order for a question to be considered a good fit for the site it doesn't have to achieve that much. First it has to pass the regular Stack Exchange bar, which includes:

  • The question is not rude/abusive.
  • The question is not spam.
  • It is clear what the question is asking.
  • The question is limited enough that it is answerable in an answer post.
  • The question is actually one question, not multiple questions.
  • The question is objective enough that answers can be evaluated.

The above criteria apply to all the sites on the Stack Exchange network. If a question passes that threshold it still has to fulfill the criteria for on-topicness of the specific site. In this case that information can be found in the Help Center.

However, even if a question fulfills all of the network criteria and specific-site criteria it is not necessarily a good question. The information provided for when to upvote and downvote is notoriously vague. The information in the tooltip says:

This question does not show any research effort; it is unclear or not useful.

This itself already provides three different reasons for downvotes, each of which is open to interpretation. Someone might downvote a question because it doesn't mention any research that was done. Someone might downvote a question because the answer is obvious (thus showing a lack of research). Someone might downvote a question because it is not well-written (loosely falling under “unclear”). And someone might downvote an answer for any number of reasons that could be subsumed under "not useful".

Question downvotes are also important for some internal algorithms, such as blocking questions from certain accounts and auto-delete processes.

Additionally, downvotes on questions alert other readers to the fact that the question might not be something that they want to spend their time reading. The downvotes also alert the questioner to the fact that something about the question is not finding favor in the eyes of the community.

In short, a question can be downvoted for just about any reason that the voter doesn't like it. The only enforced rule when it comes to downvotes is that you don't target specific users. If you think that questions are being downvoted by certain users in a targeted manner, and it hasn't been caught automatically, you can raise a flag for the issue to be addressed. Otherwise, there is not much to do about downvotes. You can try to help the questioner improve the question (or do so yourself), but even that does not guarantee that the downvoters will reverse their votes.

As for why a question would be getting both downvotes and upvotes, that simply indicates that the community opinion is divided. Some users like the question and others don't. This is to be expected when there is a large userbase consisting of many different types of people from all walks of life.

Of course, it would be nice if downvoters would leave comments explaining why they downvoted the post and what they would like to see improved. However, comments are not required, and users sometimes don't leave them for various reasons (anonymity, time, effort, etc.) The network-wide consensus has long been that it is better to downvote questions that should be downvoted without leaving a comment than to not downvote them at all.

Essentially, then, I don't think this is the biggest problem. Downvotes are not meant as a personal attack, and they should not be taken as such. Of course, this may need to be explained better to new users, but the underlying premises remain the same. Every site on the network has question downvotes, and presumably no one likes it when their questions get downvoted. But that is part of the Stack Exchange system.

Finally, here is a bit of information as to how Interpersonal Skills compares to other sites on the network in this regard. I looked at the statistics for negative-score open questions on every site that I have posted on, and compared them to Interpersonal Skills. This information will tell us about the amount of questions that are considered a good fit for the site yet still get heavily downvoted:

  • Interpersonal Skills

    • 3,000 questions
    • 12 negative-score open questions
    • 0.4% of questions are negative-score open questions
  • English Language Learners

    • 60,000 questions
    • 1,059 negative-score open questions
    • 1.8% of questions are negative-score open questions
    • Percentage of such questions is 4.5 times more than Interpersonal Skills
  • Biblical Hermeneutics

    • 6,400 questions
    • 181 negative-score open questions
    • 2.8% of questions are negative-score open questions
    • Percentage of such questions is 7 times more than Interpersonal Skills
  • Mi Yodeya

    • 28,000 questions
    • 656 negative-score open questions
    • 2.3% of questions are negative-score open questions
    • Percentage of such questions is 5.75 times more than Interpersonal Skills
  • Science Fiction & Fantasy

    • 53,000 questions
    • 899 negative-score open questions
    • 1.7% of questions are negative-score open questions
    • Percentage of such questions is 4.3 times more than Interpersonal Skills
  • Literature

    • 2,500 questions
    • 28 negative-score open questions
    • 1.1% of questions are negative-score open questions
    • Percentage of such questions is 2.7 times more than Interpersonal Skills
  • History

    • 10,000 questions
    • 159 negative-score open questions
    • 1.6% of questions are negative-score open questions
    • Percentage of such questions is 4 times more than Interpersonal Skills
  • Mythology & Folklore

    • 1,400 questions
    • 8 negative-score open questions
    • 0.6% of questions are negative-score open questions
    • Percentage of such questions is 1.5 times more than Interpersonal Skills
  • Seasoned Advice

    • 21,000 questions
    • 286 negative-score open questions
    • 1.4% of questions are negative-score open questions
    • Percentage of such questions is 3.5 times more than Interpersonal Skills
  • English Language & Usage

    • 105,000 questions
    • 3,297 negative-score open questions
    • 3.1% of questions are negative-score open questions
    • Percentage of such questions is 7.8 times more than Interpersonal Skills

As you can see, Interpersonal Skills is actually better than all the other nine sites in this regard. And except for one other site, it is much better. (These statistics only cover questions where there are more downvotes than upvotes, but not all questions with any downvotes at all.)

| |
2

How do you vote for a question?

Let's start with the simplest question first. How do you vote? Well, hover over the voting button, and it says something about research effort, being clear (or not) and useful (or not). I usually try to stick to that, and refrain from voting when in doubt.

So, based on that I feel people should vote for a question based on:

  • How useful it is, determine whether this is a good, well-written and on-topic question (in other words: Is it about the behaviors people use when interacting with others to achieve a goal?).
  • Whether the question is clearly written, with enough detail but not so much that the actual question becomes unclear because it is drowned out, a clear statement of your 'research effort' (more on that in the next point) and a clearly defined goal.
  • Research effort: what you've tried, usually do, or were thinking about doing and your doubts on why that would work.

Is voting as loose as I've described? Do these severely symmetric voting ratios matter?

Yes, because people can (and will) vote to their personal preferences, voting is as loose as you see. If people feel questions asking 'How to do bad idea X' are flooding IPS and are not the useful questions they'd like to see to build up a good site, they can vote for them being unuseful. The voting ratios matter, as they can also be used to build up the scope of the site.

I can't see into people's heads, so I can't determine if they downvoted because a user has a history on IPS and they're getting tired of seeing their content, if they are trying to show that questions like the one you used as examples here are a bad fit for IPS or if they are seeing defects in the questions of the kinds I described above. If I can find a match to one of those reasons though, I'm going to go with good intentions and say it the voting is done for quality, instead of making stuff personal.


Are users receiving 'good' feedback (especially for frame-challenges) they can use to improve their experience on the site? Are the comments for these types of questions often rude?

I'd like to link you to this (really, read that one) and this answer to an old question. It's about a (now deleted) question that's basically about 'asking whether someone is pregnant without offending them if they're not'. The first, Catija's answer, gives a good example of what feedback should look like, while the second, Shog9's, mentions an important thing about these questions: that it's often not clear there is a problem to be solved at all.

Looking at the questions you linked, the comments on the first question are actually pretty good. From that answer from Catija I linked above:

We can not respond in a helpful way if we do not know why the OP feels it's absolutely necessary to broach this subject.

The comments do exactly this: ask for clarification, ask why this is a problem, why OP needs to talk about this. They ask for goals, they ask what the problem to be solved here is. None of these comments were particularly rude, though 1 (now deleted) of them shows an impatience with OP asking these types of questions more often, that might be labeled as personal and thus a tad unkind.

The deleted comments on example nr. 2 show no rudeness either. And the last two comments (about the question being really broad and the one asking for details on which site) have 9 upvotes together. So I'd say these downvotes are (at least to a certain extent) accompanied by useful feedback.

As for example nr. 3, yes. I see the feedback lacking there a bit. The question is lacking research effort. It doesn't mention how OP thought of approaching their friends about this and how they are thinking this isn't sufficient. The downvotes are explainable to me, so I wouldn't be so hasty as to say this is because the question is asking something where the answer is an obvious 'don't'.

| |
2

All of the examples you've given have something in common: they could be seen as controversial, as they're tackling delicate issues on which people are very likely to disagree/they'll try everything to challenge your intentions. I think this is the main reason why they get downvoted so heavily.

Voting guidelines on Stack Exchange

Ideally, downvoting should be used like this:

  • On main sites: expressing the fact that the contribution isn't a good fit for the stack / does not meet (yet) the sites requirements. From the help center, on why voting is important:

Voting up a question or answer signals to the rest of the community that a post is interesting, well-researched, and useful, while voting down a post signals the opposite: that the post contains wrong information, is poorly researched, or fails to communicate information.

  • On meta sites, the rules differ. Downvoting is used to express your disagreement with the Q/As. For instance, here is one of my latest meta questions on which I'm challenging the site functioning. The ratio is of 5 UV for 3 DV, and that's perfectly normal: this is meta. On the meta sites, you're encouraged to use your downvotes to express your disagreement.

Most of us use our votes like this, but the issue is that some others choose to use their DVs differently. This is not how Stack Exchange sites work. This is how you end up with downvotes on a question that people didn't like, either because it was tackling delicate issues (e.g. LGBTQIA+ cause, autism, feminism ...) or they didn't like what OP was asking help for. One other interesting pattern too is on the questions that are asking for help to force someone to do as OP wants. Again, bad using of downvotes, yet ... not so much, as questions asking how to force things, although not officially banned from the scope (AFAIK), are usually edited out (see edit #6 of this question of mine that initially sounded like a forcing request).

What could we do about it?

Erh ... IMO, not so much. The rules are expressed in the help center and are valid across the whole SE network. I have the chance I can see the votes ratio by clicking on it so when I see a big difference between the votes I try to leave a comment to ask users to explain their choice, but it's rarely useful, as it almost never gets an answer (plus you could see such comment as chatty, and that therefore should go).

You say:

For newer users this is especially concerning; we want to keep people around and involved in our community for time to come. Even more though, we want to be constructive, polite, and encouraging to all users asking questions.

Well, here's the thing: most of the regular users know those voting rules and respect them. Maybe our help center is too unintuitive and contains too much information. Maybe it has to do with the fact that interpersonal skills are more of a "soft science" and that there's never only one truth. Most of the people who leave comments for asking to improve the Q/As are very kind and trying to be helpful in their response to the posts, and usually provide a link for further reading. I don't know if there's much more we could do without deeply rethinking the site (and that would imply rethinking SE as a whole). Interesting to note that sometimes, people know about the rules and choose not to follow them. How could we solve this?

Conclusion (TL;DR)

There are two ways of voting depending on whether it's about the main or the meta site. Some people swap the voting rules. Those rules are clearly expressed in our site guidelines and we try to be as helpful as possible when helping someone to improve their contribution. Despite of writing a community wiki meta post on how should people vote on the main site, I'm unsure what else we could do.

| |

You must log in to answer this question.

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