7

Historically our beta site has discouraged self-answered questions.

The site has been running for 6 years (3 questions/day, 98% answered, 3 ½ answers/question, and very few questions are closed). The topic of the site can be a bit "subjective" (like Parenting, or amateur History), sometimes opinionated, and not entirely dry and technical like Stack Overflow.

We have discouraged self-answered questions -- because, as one of the moderators says,

I'm strongly opposed to giving green light to quasi-questions posted with the sole intent to share an opinion, an unsolicited teaching, rambling, philosophical musings etc. [...] self-answering [might be] utilized to take advantage of the SE platform to push unsolicited info at the audience [...] I'm not convinced we should greenlight a random newbie getting on the proverbial soap box [...]. How are we to decide who is allowed to get on the box and who's not? I suppose the main issue here is community self-moderation or rather the lack thereof.

My experience of other SE sites has been that self-answering isn't an important feature -- that the large majority of topics, and of the most upvoted answers, are all people answering other people's questions. So IMO if the community votes it wants to avoid people using self-answered questions to preach on the topic of their own choice (i.e. by posting a rhetorical question and self-answering), then that's fair enough -- i.e. it's not a huge diversion from SE's core mission as a Q&A site.

I'm aware that SE also usually encourages people to self-answer and hopes its sites may become like Wikis (which IMO is fair enough when the subject matter is like SO's).

My fear is that self-answered questions may be:

  • Mostly summaries of or references to material already available online elsewhere
  • A bit spammy (e.g. a summary of something on the OP's blog or YouTube channel)
  • Someone might flood the site with unlimited self-answered topics
  • It can be preachy -- users who disagree in their answers within one topic might use further self-answered questions to continue their disagreement into other topics
  • People (or trolls) might post extremist, unorthodox, unwelcome, controversial views on topics of their own choosing.

Someone recently argued though that "experts" want to be able to post self-answered topics and won't frequent the site if they are not allowed to.


What's your experience (and on which site)?

  • Are self-answered topics a problem, a distinct category, difficult to control or to quality-control?

  • How do you (do you even) prevent people posting the kinds of low-quality or problematic content which I listed above)? What implements the prevention -- is it the community, the moderators, or some site-specific policies about what content is welcome?

  • Do self-answered questions seem a bit spammy sometimes? How to not be a spammer says,

    Always solve the asker's problem. A good answer should at minimum allow the person whose question you're answering to solve their problem. Not all questions can be answered this way, but if you don't think you can write up a complete solution then you're better off looking for a different question.

    ... but a problem with self-answered questions is that there isn't even an asker -- the post is unsolicited, it's not cooperative like answering someone else's question.


Historically when someone posts a question in order to self-answer it, a moderator says, "That's not really conventional on this site, this is a Q&A site which is preferably for answering other people's questions".

So we've had little experience with allowing self-answered questions -- and having to quality-control or moderate them.

How does a site cope with the various kinds of problem listed above, without deprecating all self-answered questions categorically?

17
  • 10
    Generally sites treat self answered questions the same as any other question. They are still subject to closure for all the same reasons for instance. Commented Jul 9, 2020 at 13:12
  • 1
    When I read my question above it seems to me that what the community may dislike is the self-answer, more than the question.
    – ChrisW
    Commented Jul 9, 2020 at 13:16
  • 1
    And the community don't want to engage with (answer) the question themselves -- because it seems as if the OP isn't interested, isn't asking a sincere question, already has an answer.
    – ChrisW
    Commented Jul 9, 2020 at 13:23
  • 1
    Again self-answers are normally treated like any other answer. If it's not useful downvote it and if you have the rep you can even vote to delete it. Commented Jul 9, 2020 at 13:24
  • 3
    I usually post a question/answer pair if I solved a tricky problem and want to spare others the pain it took me to get there. Is this not what you are supposed to do? Your post reads like there is a huge issue surrounding self-answered questions, but I've never seen it that way.
    – fuz
    Commented Jul 9, 2020 at 13:28
  • @RobertLongson Ah. On our site only moderators would delete an answer, and wouldn't do that without a good excuse (e.g. the answer's being hostile, or breaking some other site-specific rule). So I guess you're saying that the community (not moderators) can handles unwelcome answers on a case by case basis (i.e. without needing much of a formal rule), and can "handle" it by deleting it.
    – ChrisW
    Commented Jul 9, 2020 at 13:30
  • 5
    Maybe you don't have many users with answer deletion privileges. Moderators generally only delete non-answers. High rep SMEs delete bad answers. Commented Jul 9, 2020 at 13:32
  • Related Commented Jul 9, 2020 at 14:33
  • 3
    "Historically our beta site has" Which beta site?
    – Mast
    Commented Jul 9, 2020 at 14:33
  • @RobertColumbia Thank you for the reference, I reviewed those answers here (in this self-answer).
    – ChrisW
    Commented Jul 9, 2020 at 14:50
  • @DonThousand Thank you for the reference, I reviewed that too.
    – ChrisW
    Commented Jul 9, 2020 at 14:58
  • 3
    @Mast looking at OP's profile that would be Buddhism. The quoted part by a Mod can also be found on that Meta.
    – Luuklag
    Commented Jul 10, 2020 at 9:01
  • @Luuklag Yes I'm currently trying on that Meta -- What about self-answering questions? -- to reevaluate or improve our policy. And I posted here hoping for insight into whether other sites find that self-answered topics may be associated with low-quality content or disruptive behaviour -- and what special guidelines we might formulate for self-answered topic to avoid these drawbacks, if our policy were to welcome them in general.
    – ChrisW
    Commented Jul 10, 2020 at 9:14
  • @RobertLongson What did you mean by "High rep SMEs delete bad answers"? Is it only trusted users who can do that, or can users who have a gold badge for a subject-matter tag do that too?
    – ChrisW
    Commented Jul 10, 2020 at 19:06
  • 1
    That's one of the things that varies site by site. Non-answers and nonsense should be deleted everywhere, but some specific sites may have a policy to delete e.g. unsupported answers, even though they're only bad but still answers. Commented Jul 10, 2020 at 19:39

8 Answers 8

20

I'm a moderator on two sites, on both of which I've earned Reversal badges on meta for defending self-answered questions to OPs who suggested discouraging or disallowing them:

The long and the short of it is that self-answered questions are explicitly allowed and encouraged by Stack Exchange, and I've never seen any major problems with them on any of the sites I know.

Here's a very popular and well-researched self-answer on Movies & TV. Here's one on Sci-Fi & Fantasy. Here's one on Literature. Here's one on Puzzling.

Sure, there have been less successful self-answered questions too, but that's never been a major issue or concern on any of the sites I know about. Sometimes a user has posted self-answers which seem pointless or uninteresting to the community, and in that case they get hit hard with downvotes. But if a Q&A pair is useful and interesting, surely it should be rewarded regardless of who wrote it. To address your concerns one by one:

  • Mostly summaries of or references to material already available online elsewhere

    That can be a good thing, if the material is all compiled in one place for the first time. This self-answer contains a long list of information which was technically "already available online elsewhere", but only by trawling through dozens of different webpages to collate that list.

  • A bit spammy (e.g. a summary of something on the OP's blog or YouTube channel)

    Usual self-promotion rules apply. If someone is promoting their stuff without attribution, mods can take appropriate action whether they're answering their own question or someone else's. If someone is promoting their own stuff too much, mods may need to have a word with them, but that's not directly a consequence of being able to self-answer.

  • Someone might flood the site with unlimited self-answered topics

    I've seen this type of fear very often on SE, usually as an argument for making something off-topic: "but if we allow people to ask questions like this, someone could flood the site with hundreds of them!" My response is always that such floodings are almost unheard of and can be dealt with individually if they ever arise. Many question types on many sites are "mass-producible" but still on-topic. Most people don't flood the site with unlimited questions, and if someone does, then it's a problem with that user not with the type of question.

  • It can be preachy -- users who disagree in their answers within one topic might use further self-answered questions to continue their disagreement into other topics

    If people post bad answers, they should be downvoted regardless of whether they're self-answers or not. If people post good answers, it's a win-win situation. Voting will usually handle this kind of situation, in my experience.

  • People (or trolls) might post extremist, unorthodox, unwelcome, controversial views on topics of their own choosing.

    Again, if people are trolling or posting offensive content, they can be banned or the posts removed just as if those posts weren't self-answers. Nobody's advocating for allowing any and all content in self-answered questions, just for treating self-answered questions the same way as others.

TL;DR: your concerns aren't really about self-answered questions as such, but general site moderation issues. If people abuse the system using self-answered questions, that can be handled just like any other abuse of the system.

6

You ask for experience on other sites so I answer about the two where I have longest experience and on which I am not a moderator.

On CrossValidated my impression is that these are rare. One case which does occur is that following a discussion in the meta site about the need for a canonical question and answer about a topic some expert will post a question and then give a comprehensive answer. This can then be used as a target for closure as a duplicate. Not surprisingly this is seen as valuable and usually gets heavily up-voted.

On Spanish.se we get a few of these and they are usually from respected users and so are received in the same way as normal questions and answers. There is sometimes a time gap between the question being posted and the answer which may imply that the OP had to do more research to find the answer.

5
  • So you don't get many but when you do they're never a problem -- you've never had a user flooding the site, or spamming their own books or blogs, etc.
    – ChrisW
    Commented Jul 9, 2020 at 13:39
  • 1
    In Spanish.SE we also have a bunch of questions following the 'canonical´ aspect of CrossValidated. Namely relevant aspects of grammar that seem to be common and we want to post as a way to have some kind of 'language documentation'. They tend to work well and gather quite a lot of votes and views. Commented Jul 9, 2020 at 13:39
  • The problem with canonical questions normally lies on 'How much research should you show when you are self-answering on the spot?'. In general, in Spanish.SE we tend to loosen the show your effort principle on these cases. Commented Jul 9, 2020 at 13:40
  • @fedorqui'SOstopharming' Yes by far the most frequent use-case for self-answered questions, that I've seen, has been for faqs (on meta).
    – ChrisW
    Commented Jul 9, 2020 at 13:41
  • @ChrisW yes in Meta it is normal to have such posts. Commented Jul 9, 2020 at 14:29
6

I find that it's usually pretty easy to distinguish between a question that is merely serving to provide a venue for an answer, and a genuine question that happens to also have been answered by its asker. If a question is well-written, contextualises itself in the broader field of knowledge through references or quotes, and clearly shows why someone might be left wondering something by the sources of knowledge that are out there or why they might think there is a contradiction between two facts, etc, then that's a good question. Self-answers aren't relevant. If they don't do that, which is semi-common for the self-answers I see, then that's a poor question. Depending on the site there might be reason to close it (perhaps as Unclear, because it's unclear why anyone would be wondering what the question asks), but at the very least downvotes would be warranted. And maybe a comment to tell the poster to get a blog if they just have thoughts they want to publish.

1
  • That sounds like what we do, i.e. users downvote, moderators may close, and tell the OP that this is meant to be a Q&A site not their blog. The important difference is, you do that if-and-only-if the question is low-quality, e.g. not one which other people would want to answer (or which they like but don't answer because the self-answer is already good).
    – ChrisW
    Commented Jul 9, 2020 at 15:09
4

I think your site can have self-answers just as well as any other. Of course not every question is suited for a self-answer, but as others before me have answered there are already mechanisms to deal with that.

I took a look at your site, and I found that these questions for example could perfectly be self answered. The first question is formulated rather broad, but the most upvoted answer cites a great number of sources that could also been given by the OP as a self-answer.

If a question tends to draw opinionated (preaching) answers, as you fear, it should be closed as "Opinion-based". Albeit that being a hard thing with such a subjective topic as religion.

3

I agree with Rand al'Thor's answer completely (in short, moderate them exactly like you would do anything else), but I wanted to add an additional point:

I think you are being way too cautious here. There's no reason to try to anticipate and prevent every problem that can possibly occur on a site. There are an infinite number of potential problems, but a very finite set of problems that ever actually occur. If you keep trying, you are going to drive yourself crazy and not get very much done!

It looks like you may have partially realized this already from one of your comments:

Part of my problem might be that I've trying to protect the community (from disruptive behavior) and the site (from unasked-for answers), by preemptively discouraging people's self-answering

Self-answering can be a very useful tool and there are good reasons why they are not only allowed, but encouraged across the network. Disallowing them out of fear that they will be abused (especially in the face of evidence from other site that show it hasn't been an issue) is just shooting yourself in the foot for no good reason.

What I would recommend is to keep thinking about this stuff and all the advice given in the answers here, but meanwhile allow self-answered questions on the site. Then, deal with problems as they come up. That way you are solving real issues on your site and not potential imaginary problems. Your site won't crash and burn the first time it encounters a small bump. In fact, that's how communities grow and learn how to manage themselves: by encountering and overcoming issues.

2

A comment suggested this might answer the question -- What can be done to improve moderation of self-answered questions? -- the question is relevant to this topic, but its answers aren't very useful to me.

Jeff's answer says to moderate like any other question -- ensure it's a high-quality question, which may attract other answers too.

Jon's answer seems especially about how to make a good self-answered question better:

  • If it's duplicate then close as duplicate
  • If it doesn't seem a useful question then close as "Not a Real Question"
  • Make the answer encyclopedic (or "canonical"?), answer it two or three ways
  • Cooperate with the community in improving it.

That's not quite the same as my question, which is apparently more how to squelch low-quality self-answers.

I note that "Not a Real Question" sounds appropriate but is no longer one of the available close options -- it was replaced by "too broad" or "unclear" -- not really the kinds of problems I fear, e.g. "spammy", "preachy", "insincere", "argumentative", "controversial", "too many", etc.


Another comment suggested this: Asking questions to add to the general knowledge pool

The answers to that describe beneficial use cases, i.e. why it can be useful -- not why it can be harmful and how to handle it if it is.

One highly upvoted comment (more popular than any answer) suggests,

Just stop making judgement based on who is posting a question and who is answering it. The value of a question/answer pair doesn't really depend on who is the poster.


RandalThor's answer references this topic which says self-answered questions might be downvoted because they're self-answered -- and says that people who understand that self-answered questions are allowed should focus instead on whether the post is "good content".


There are some posts on the Christianity Meta, here and here.

One suggests that if you self-answer then you might prefer to accept someone else's:

As a matter of habit, I prefer not to accept my own answers and am eager to reward competing ones.

The other is interesting as suggests that self-answering should be used for genuinely difficult questions which required research. Also make sure it's not a duplicate, and spend time on the question, including your prior research (which IMO will help to make it a genuine question)

Make sure you spend almost as much effort on the question as you did on the answer. Much like on Stackoverflow we want people who ask questions to tell us what they've tried attempting to work out their problem. Tell us what you thought about, where you looked before you found the answer etc. If you can't write more than a sentence or two about the problem, seriously consider whether you're the right person to be asking this question.

As a side-effect IMO this would keeps the OP from seeming to be conceited -- because they depict themselves as a fellow-student, not a teacher. The effort also makes the question and its self-answer more accessible, to any fellow-students and to teachers.


The Hinduism Meta encourages people to self-answer even if the answer is easily Googleable (and people do because they feel enthusiastic).

Shog9 suggested there that poeople shouldn't (he posted this after the site had been open for only a week):

There are a lot of questions here that are... Pretty basic. "Could be answered by a Wikipedia entry" basic. Questions that require no expertise aren't a particularly good way to attract experts. You gotta expect a few of these, but when every other question can be answered by a Google search...

...And worse yet, an awful lot of them are answered that way: by quoting external sources. A site that just regurgitates things that others have written isn't much good; answers like that just clog up search results without adding anything. It's important to cite your sources, but just as important to demonstrate knowledge of both the question being answered and the topic itself - this is where we can actually make the Internet a better place, demonstrating how to use this knowledge to answer specific questions. Answers that are primarily copied from Wikipedia or other forums / blogs - even when properly attributed - are just speed-bumps, annoyances to folks searching for good information.

I'd really like to see more questions that require some serious research and experience to answer... I know some of you have been studying these topics for years, and are well past the beginner stages of your education - what sort of problems do you encounter?

So I take it we should bear in mind that answers and self-answered questions should not be little more than a reference to or a copy from, a summary of, some external source..

13
  • 1
    I find the Unclear reason an appropriate modern substitute for Not A Real Question, in that for questions like these it's often unclear why anyone would be wondering what the question asks. Poor questions like that need more contextualisation, and if it's not provided, I'd vote to close as Unclear. Commented Jul 9, 2020 at 15:02
  • @curiousdannii Unclear itself is no longer modern, having now been replaced by Needs Details Or Clarity. Commented Jul 10, 2020 at 13:12
  • @Randal'Thor One of the things I worry about is "conceit" -- e.g. lecturing people (by posting a self-answered question) when nobody was asking. It (i.e. self-answering) can be done well, be interesting, etc. -- but I've also seen a couple of users do it in a way that seemed "disruptive" (i.e. "a broad sense of community resentment over this user’s behavior" and "a dark storm cloud of moderator flags" etc.). It's to that extent it seems not a real question -- an organic question, a question that anybody else wanted other people to answer.
    – ChrisW
    Commented Jul 10, 2020 at 13:26
  • @ChrisW I considered adding a paragraph to my answer emphasising that such things may vary considerably across sites. I know that on the sites where I'm active, self-answering to improve the knowledge base makes sense. I imagine it would make less sense on sites like The Workplace or Interpersonal Skills where questions are often based on a specific real-life problem that the OP needs help with. I'm not familiar with your site's topic/scope at all. But your question specifically asked for experience from other/various sites, so I guess that goes without saying. Commented Jul 10, 2020 at 13:29
  • 2
    Also, it seems you agree that self-answering can be done well, with just a couple of bad eggs turning it into a bad thing. If your community at large also agrees with that, and if you can get a description/consensus of what makes a bad self-answer on your site, then you should be able to create a site-specific policy around that. Commented Jul 10, 2020 at 13:34
  • @Randal'Thor Part of my problem might be that I've trying to protect the community (from disruptive behaviour) and the site (from unasked-for answers), by preemptively discouraging people's self-answering -- instead of requiring users to vote on it etc.
    – ChrisW
    Commented Jul 10, 2020 at 13:34
  • @Randal'Thor what makes a bad self-answer on your site Yes that's why I'm asking here, thank you, i.e. to see how other people see it, what characteristics are considered good and bad and so on. Eventually I may be able to formulate/propose some guidelines about what's good or bad in a self-answer, and whether it's the users or the moderators who are supposed to vote on it.
    – ChrisW
    Commented Jul 10, 2020 at 13:38
  • @Randal'Thor Oops. Shows you that in my mind I just map the close reason names onto their old ones automatically! Commented Jul 10, 2020 at 13:41
  • 1
    In my experience (mostly from Science Fiction & Fantasy), a good SAQ is one where many people would be curious about that Q: someone is sharing knowledge that's useful for the community at large. A bad SAQ is one that doesn't seem interesting to people, seems like the user posted it just to have a Q&A for rep. (We once had a user asking questions like "did this superhero ever do this thing?" and then self-answering with a comic panel image showing that superhero doing that thing. People downvoted those a lot.) The community can downvote bad SAQs on the same basis as any other uninteresting Qs. Commented Jul 10, 2020 at 13:44
  • @Randal'Thor Is a topic that's downvoted like that also closed or deleted, or is it left to sit around with a negative score? If it is closed and/or deleted then by whom (user, moderator, or SME), and for which reason? What about people who downvote SAQs on principle -- do you agree there are people who do, and does that affect how/whether you moderate SAQs?
    – ChrisW
    Commented Jul 10, 2020 at 13:53
  • 2
    @ChrisW Not closed or deleted unless there are other problems with it. There's no close reason for being "uninteresting"; a question might be perfectly on-topic but just get downvoted because people don't like it. Closure and deletion are for things which objectively violate community standards (e.g. off-topic, opinion-based, etc.) There may be a few people who downvote SAQs on principle, but usually not enough to affect the main flow of votes. Commented Jul 10, 2020 at 13:57
  • I see, thank you: there are several negatively-scored Qs remaining on SFF -- which all seem to have upvoted answers. What about negatively-scored answers (which aren't as easy as questions for me to search for) -- I guess those too may remain on the site, or be deleted by "trusted users" -- even then according to SE perhaps an answer shouldn't be deleted just for being wrong.
    – ChrisW
    Commented Jul 10, 2020 at 14:13
  • @ChrisW "Part of my problem might be that I've trying to protect the community (from disruptive behavior) and the site (from unasked-for answers), by preemptively discouraging people's self-answering" This is a good realization and I'd agree with it. Generally, I find there is a danger for people to anticipate problems that will never occur and would be easily handled if they do. I think this is such a case. Don't try to solve problems that don't exist yet would be my advice. Allow self-answers and prepare yourself for potential issues, but deal with them as they come up over time. Commented Jul 10, 2020 at 15:44
1

A possible partial answer might be that self-answering shouldn't be used for any self-promotion.

  • If you want to reference your own blog, you can -- but only in your answers to other people's questions as described in How to not be a spammer
  • You can post questions to self-answer them, but these answers should be self-contained -- i.e. instead of a blog, and not in order to summarise and reference your own blog or book.

In theory this policy too might depend on the quality of the post -- if it's good then it's welcome -- however perhaps even if the OP were a famous author, a respected subject matter expert and a good writer, their posting "Question + Short Answer + Reference to a publication of mine" would seem a bit spammy.

2
  • This might be a good rule but impossible to apply to the author's fans and enthusiasts, who (in theory) might do this to promote the author's work.
    – ChrisW
    Commented Jul 9, 2020 at 15:56
  • Another guideline might be that the OP's own publication shouldn't be the answer's only reference -- if a good answer should be "encyclopedic" then it might have references to several sources.
    – ChrisW
    Commented Jul 10, 2020 at 14:30
1

Thank you for all your answers.

I've tried to propose a new FAQ here -- Proposal: Self-answering questions is welcome

2

You must log in to answer this question.

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