The story behind this question is that a new user posted this question with a simultaneous answer. I posted a comment indicating that "answering own questions was extensively debated" and suggested to not post his answer immediately after the question. Then, Ken White replied to my comment with an opposite point of view and ended: "Don't propagate wrong information here, what you said is false". After a short discussion, all of the comments were deleted by a moderator.
The central point of this topic is this: Can users answer their own questions using the option of the question page or they may be downvoted just for this reason? In my opinion, they shouldn't use that option. These are my reasons:
It is a fact that many answers posted by the original question poster are downvoted mainly for this reason.
It is a fact that many users, mainly beginners, post answers to their own questions because they see the "Answer your own question" option in the question page.
I got these conclusions after read many questions/answers/comments in this meta with the filter "answer own question" (more than 5,500 results).
You must note that I am NOT proposing to eliminate the possibility of answer own questions. I just propose to eliminate the "Answer your own question - share your knowledge" option from the question page.
So it might be worthwhile to ask the question, wait a day or so to see if any new answers come up, and then add your own answer.
So often I have seen new (and unfortunately older) users saying that “you shouldn’t answer your own question” and sometimes they even downvote the answer for that very reason.
If you ask a question, then immediately answer it yourself, how is this encouraging participation and community development? Yes, it’s archiving the information for all the Googlers out there, but posting on your blog would do exactly the same thing.
I asked and answered my own question early on and got heavily voted down for doing so in the wrong format. The interface should be clearer about how to do this, if it really is encouraged. Someone new to StackOverflow who has an incredibly useful tidbit to share is surely going to violate this awkward format and experience the same fate – and that’s going to make it very unlikely that useful tidbit ever surfaces in a search.
In the meantime it would be nice if the community could help new users to how to do this instead of voting otherwise useful info down for procedure.
As a new user of the Stack Exchange site, thank you for this explanation. I found this while reading the faq and would not have considered self-answering a question otherwise.
No. I completely disagree.
I don’t think anyone should ask a question to which you think you already know and have the best answer.
If you are trying to get new ideas and think one particular way of thinking is the best, then you should include it in your question and ask whether that is best. That is okay.
The one place I think it is okay to answer your own question is when you really don’t know the best answer, and the answers given are not quite it either, but they lead you to thinking or further web research and then you discover the best answer. Then you can include it and state so.
But preferable to that, you should make sure you wait long enough, say a few weeks or even a month, for someone else to come up with the same answer you discovered. Then you can and should give them credit for the answer. And you can document any subtleties in comments to that answer or by editing your question with a followup.
And waiting the few weeks or month without accepting an answer will result in more answers, giving you more ideas, and maybe a better solution.
I see now that Stack Exchange is not only allowing answering your own question. They’re actually promoting it!
Click “Ask Question” and on the page just below the “Post Your Question” button, you’ll see a line that says:
“Answer your own question – share your knowledge, Q&A-style”
At the beginning of the line is a checkmark. If you select it, up pops a text box allowing you to immediately supply your answer.
The link on that line takes you to this blog post.
Well, rather than the one exception I stated above, I still very much disagree with doing this and think that promoting this when we ask the question is doubly bad. I still think one should only answer their own question after others have had a chance to do so – and especially should not do it immediately when they ask it.
I disagree on ask and answer our own answer. There are possibilities that one sided answer is there whether it is positive or negative.
EDIT: Second bunch of requested references added:
The longer until you post an answer to your own question, the more points the question gets, as people try to answer it. Most of the points are awarded within hours of posting. So waiting at-least a couple of hours will continue to be my recommendation. Unless you care nothing about points, or are a well respected user
The aim of SO is to produce an online resource consisting of questions matched with good answers chosen by the community, but it achieves this by having another matching function: to provide smart people with some newbies to provide help to.
Self-answered questions do not provide the second function, and so they make people angry, sometimes without them even knowing why.
10 NEVER post your own answer, to your own question, within 12+ hours, unless you want people to take it the wrong way. Most people will assume, that if you answered your own question, that you didn't really need to post the question in the first place. I believe this is just human nature, you can't really control it, you just have to deal with it, and move on.
@Brad Jeff said many times that it's encouraged to answer your own questions. In reality though, answering your own questions is likely to go -1
While I personally feel that this is OK (in moderation), the community may not agree. An example I recently saw...
An employee of The MathWorks who is active on SO decided to post a couple of questions (here and here) that he knew were most commonly asked about their product (MATLAB). He felt that putting these questions (and their answers) in a highly visible place (i.e. SO) would help MATLAB programmers be able to easily find them. This seemed to me to be a perfectly good use of SO, and was in line with the FAQs and posts dealing with answering your own question. However, a few users responded with downvotes and a negative comment or two.
I guess answering your own question in this way is something of a roll of the dice. Sometimes the community likes it, sometimes they don't.
If it was a genuine case of posting a question because you had a real problem you needed help with and then posting an answer a day or so later when you solved it - with or without the help of any answers - then I can't see a problem with it. In fact I've done it with a couple of the questions I posted on SO.
Questions with an answer by the person asking the question, don't seem to get that many points. I would definitely wait 12+ hours before posting your answer, otherwise most people will just ignore it and move on. – Brad Gilbert
I disagree with answering your own questions or using StackOverflow as your personal blog.
The FAQ states: "You should only ask practical, answerable questions based on actual problems that you face." But the moment you ask a question you know the answer to already, you are no longer facing a problem. You already got the solution. There is no point in asking the question anymore, because you do not need help.
The only reason to add my own answer (for me) would be if I didn't get any useful answers and then managed to solve the problem on my own in the meantime. Sure, the resulting page would be the same as providing both in one go, but at least I was facing a problem then when I asked the question.
Now, I do understand that StackOverflow is a Q&A site. And by that definition, it should not matter whether you know the answer already. If you can provide both, it's still a Q&A site and maybe your question is helpful to someone with a similar problem in the future. So, all is well, right? Well. No. Because you are breaking the process.
When you provide Q and A in one go you effectively tell people: "I don't need your help. Go on. No problem here." That's broken. The community-driven problem solving process is a key element of this site. You are not supposed to take our role in this game. You ask your question to us and we will help you. We enjoy helping you. Why are you taking that away from us? We are the community. If you need to share some solution so badly, go find someone with that problem - or post it on your blog.
Furthermore by answering, you stole my chance to get the green tick. People love getting the green tick, right? Upvotes are nice, but only one person can have the green tick. It indicates a winner. Sure, the OP could decide that my answer is better, but how often do Accepted Answers get changed? Rarely. Once the OP has a working solution, the Accepted Answer stays. I have very little incentive to add my answer there.
Also, chances are "the problem" was already solved on StackOverflow anyway and I'd rather like to see good solutions added to existing Questions to have collections of quality answers instead of people spreading their quality answers into many individual posts. The latter cause duplication and make it much harder to find the good parts among them. If we want to encourage Canonicals, we need to encourage people to add them to existing questions and not each try to do their own.
Moreover, if we allow people to use StackOverflow as their personal blog, we implicitly also allow them to not care about duplicate or inappropriate content. I can write whatever I want on my personal blog. If I feel about writing a tutorial about How to format a timestamp into a specific date format I can do so. If I want to write a mini-series about what each letter in SOLID means, I can do so. If I want to write about how to use my project X on GitHub, I can do so.
But the first one would be a dupe and the second one would belong on programmers.SE and the third one is likely too localized. IMO, we already have more than enough to do fighting the dupes that come in on an daily basis for real questions (even with combined coordinated efforts). By encouraging people to abuse StackOverflow as their blog, we only make it harder for those trying keep this place tidy.
Finally, the idea that StackOverflow can substitute for my personal blog is flawed anyway. A personal blog is an individual effort. I have full control over the content. Not only can I decide the licensing terms but I can also delete content and comments at a whim. And people cannot edit my posts. This is fundamentally different on StackOverflow. I'll always have a community moderated blog on StackOverflow at best.
answered May 22 '12 at 18:18 Gordon♦
This has nothing to do with rephunting. It has to do with motivation. The green tick is a strong incentive for answering on this site. It's like a mini-award and a strong part of the gamification process that drives SO. The suggested insta-answers lower chances to gain that tick for anyone (the "winner" is determined before the game already. So why play?). Hence lowering chances that someone provides another good or even superior answer for you and future visitors to enjoy. – Gordon♦
There is a (de facto) accepted way to do this:
Post your question
Wait for the normal attention time span (question rolled off the main page - depends on traffic/site/time of day). Can be less than 5 minutes on SO, more so on the other sites.
Optional: wait some more
Post your own answer by editing the question and also incorporate some information from any answers (if appropriate)
And nobody will know if:
you already knew the answer or
you found it in the meantime helped by the incoming answers or
you found it in the meantime despite completely useless answers
You will not be accused of reputation farming as questions usually receives much fewer votes than answers.
answered Jul 28 '09 at 17:59 Peter Mortensen
I think it would be ok for a few occasions. But piggybacking on the system would be frowned upon I believe.
Of course the users of whatever system would love to hear how you solved problem X but when a user gets to be known as the self answerer, we would see a lot of negative reactions towards that due to various factors.
Reputation "exploit", unfair badge earning (even if it is a CW) etc etc.
So, do you suggest that it will be more fair, that, if I'm answering my own question, it will be more polite to mark it as a community wiki ? Never tried to do this, but does "community wiki" post could be marked as accepted answer?
Although it is possible to answer your own question and set yourself as the accepted answer, it is generally not a very well-accepted practice.
I asked a question recently, looking for examples of an online tool. Per the FAQ, I answered my own question with a few examples that I had already found. I did this so that these examples would be included in the answers, along with the rest.
The question has received only down votes and insults saying that this shouldn't be done.
Should I not answer my own questions?
posting a question for which you obviously have an already prepared answer is often seen as 'begging' for reputation, and therefore may not get as many upvotes.
Joel, sadly, you're correct about the unjustified (IMHO) negative sentiments to Q+A posted together. I just posted a question stackoverflow.com/questions/15178903 to which I couldn't find a good answer anywhere on the web at all. I spent days coming up with a good answer, and it seemed beneficial to share it with others. The fact that I spent many days thinking and solving the problem is obviously masked by me posting the question and answer together. Is there anything you think can be done to prevent the negative reactions? Thank you. – arielf
It's a tricky situation. I have questions in my head where I can think of a few answers, but they aren't the only answers.
Posting my own answers looks like I'm fishing for points; not posting makes everyone repeat work I've already done. Marking my answers as Community Wiki helps avoid that stigma
I've even seen people post a question, and then post a fully thought out answer to it about 30 minutes later. It was clear they knew the answer - they just felt it was an unanswered question that they could answer for the community
I've done this, but explicitly stated in my question that I would give my answer separately. In your case, yours is the only answer to the question and may give the impression that you are just seeking votes to increase your reputation. I did not take it that way, though.
I also agree with @ScottieT812. This should be a community wiki question.
I think that the main reason that people bagged on you was because you posted an answer only seconds after posting the question.
If your question was one where a solution could be worked out from the question (eg math questions, what’s wrong with this line of code, etc.), then it could have been explained in that typing out the question in a formal way caused you to view the problem in a new way that led to a Eureka! moment, allowing you to answer it immediately.
However, your question was one whose answer could not be derived from the problem (you asked for examples of sandbox webapp sites). Therefore, the fact that you posted an answer in less than a minute of the question, people (in my opinion rightly) found it to be quite suspicious, and that’s why you got hit.
If you’re going to do that sort of thing, at least wait an hour or so before posting an answer. That way it looks like you continued researching the problem while waiting for someone to post an answer, and found one on your own (ie, the whole point of having a self-learner badge).
answered Sep 7 '09 at 21:37
There is at least one situation where I find answering your own question mildly annoying.
If I'm perusing the list of questions, looking for something I can answer, then see a question which I know the answer to, I like to jump on it. Help a fellow out.
Then I'm halfway done writing an answer and the little "Another answer has been posted" banner pops up. I refresh the page to see that the asker has immediately answered his own question, meaning that there was no actual question.
No one wanted to know the answer, because the asker wasn't actually having a problem when he wrote the question. Even if I beat him to his own answer, it wouldn't matter because I wouldn't be helping him, he already knows a working answer.
This situation just wastes my time.
I don't mind answering your own question, but I would like it if one would only post a question if they (at the time) want an answer that they don't have. It's fine to answer your own question later if you discover the answer, be it trial-and-error, research, or a combination of other people's answers.
The difference would be akin to a man asking for the time.
The man asks if anyone has the time. No one answers right away, but someone does mention that it's sometime after 10 o'clock. He realizes that there is a clock on the wall and mentions aloud "Oh, never mind, I see that it is 10:40.
The man asks if anyone has the time. One person speaks up "It's ten fo..." but is interrupted by the first man who, while checking his watch, states "It's 10:40, just in case anyone wanted to know."
For the sake of argument, let's assume that it will always be 10:40 in whatever room this man is in.
edited Sep 8 '09 at 1:28