I'm a big fan of continuing to research a problem after a question has been asked. It shows initiative on the part of the asker. At the same time, I find it incredibly annoying when people ask a question only to answer it themselves less than 10 minutes later. This shows that they really weren't looking hard enough and, in my opinion, fills the site with unnecessary spam.

A good example is this question from yesterday. The author asked the question, then answered it 6 minutes later.

Ironically, I know this question would have probably been answered with a link to let me Google that for you if asked on the regular WP forums - a practice that's been degrading the quality of the site lately.

So here's the question: what do we want to do with this kind of q&a on the site? Should we encourage people to search a minimum amount of time on Google before asking? Should we discourage answering your own question the same day you've asked it?

My concern is that this author won't accept another as "the answer" since he's already answered it himself with the correct solution. Anyone else (i.e. John Bloch's answer 1 minute later) will end up appearing as a copy-cat response.

share
add comment

4 Answers

up vote 3 down vote accepted

I agree with your completely @EAMann and sorry to differ @tnorthcutt. I don't think it's a problem if we send them to Google first because it we follow a somewhat aggressive policy of policing really high quality answers then over time (and not too long a time) this site will be one of the top search results for practically every WordPress related question. And if my experience with Googled questions leading to StackOverflow is any experience, we'll also have the best and most relevant answers too. So if we focus on creating really quality it will be rewarded.

So yes, I think the FAQ should seriously discourage people from answering their own questions and that if they do we should close the question (can we delete them?)

In the case they do identify the answer on their own we should require the ONLY do so in the comments after their question and further they should encourage and accept others to answer it more completely for others who many have the same question in the future.

On the surface I like the idea of a waiting period but it might encourage hit-and-run questions where the user never follows up to close the loop. But if the time period is reasonable it might make sense. At least be worth a try.

I would like to add (maybe this should be another topic of discussion) that our FAQ state that by asking the question commit to following up and closing the loop to make sure the answer has as much value for others in the future as it did for them and if they can't commit to doing and they show evidence of not following up they will be notified and possibly banned (for example, without calling anyone out I noticed there is someone who has asked 5 questions recently but hasn't followed up on any of them.)

UPDATE: I'm going to add another context I was not thinking of when I wrote the above and that's when one of us knows of good question, especially one asked elsewhere on the web, and wants to post it here for everyone's benefit. In that case I'd suggest posting it as a community wiki and stating that you plan to answer the question yourself. That ways other's aren't busy doing research to help answer a question you already know the answer to. This is basically what @EAMann said in response to @tnorthcutt's answer below.

share
    
Someone downvoted this but did not provide a comment as to why. Whoever downvoted please do me the courtesy of explaining the downvote. –  MikeSchinkel Aug 23 '10 at 23:24
    
Hi Mike - that was me. It was a little harsh, given that it was a response based purely on the fact that I disagree. As my comment below implies, I'm quite a fan of self-answering, because one of the original aims of SO was to create a wiki to store a huge amount of knowledge, and self-answering is a good way of achieving that. Since your update deals with that, I've removed the downvote. –  Bobby Jack Aug 24 '10 at 0:05
    
I can see the benefits of this approach, but I do have some concerns. For instance, how would this approach handle Adam's scenario, where he posted a question, got a part-of-the-way-there answer, was then able to work out the solution himself, and returned to post a complete, "best" answer? I would be against that situation resulting in a lock or delete of the question. I also think if it's the best answer, it should be listed as an answer, not as a comment. –  tnorthcutt Aug 24 '10 at 3:47
    
@Bobby Jack - Thanks. What I wrote originally wasn't contemplating questions posted because the poster already knows the answer and want to shares (which I have done) but instead @EAMann's original question which is someone who posts a poorly considered question then answers it 5 minutes later because they figured it out; they former adds great value but the latter adds very little (i.e. "the blind leading the blind") It was my bad to not realize there were multiple use cases when I first posted. –  MikeSchinkel Aug 25 '10 at 18:33
    
@tnorthcutt - Yes, I'm moderating my views on this. One thing I'm not finding with StackExchange, and that's disappointing, is a way to segment out the great archetypical questions & answers and the random "I have this obscure problem" and "here's how to solve it." I can find nothing in the UI/mechanism that allows someone to browse the best answers. Eric's question is just a canary in a coal mine regarding this issue. –  MikeSchinkel Aug 25 '10 at 18:36
    
@mike - Kind of sounds like we need a "best of WP answers" tag reserved for the best questions and answers. Maybe one that only mods can apply. Not sure how that would work, though. And if it were a tag, I guess it could only be applied to questions, not answers. –  tnorthcutt Aug 26 '10 at 13:24
add comment

Should we encourage people to search a minimum amount of time on Google before asking?before asking?

No.

I can't find the reference right now, but Jeff Atwood (IIRC) made the point a while back on the SO blog that doing so would defeat the purpose of a Stack Exchange site. We want this site to be the best resource for finding answers to WordPress questions. Ideally, when people Google a WordPress question, they'd find a link to this site at the top of the results page. We don't want to tell people to go to Google, which essentially means they'll find the information somewhere else.

share
    
Valid point ... so what can we do to avoid people immediately answering their own questions? That defeats the purpose of an answers exchange. –  EAMann Aug 19 '10 at 15:39
2  
You're right. That is a tricky one... What if answers you post to your own question earn you no reputation? That would be a (small) start. –  tnorthcutt Aug 19 '10 at 16:42
2  
A small start yes, but a start nonetheless. Beyond trying to farm reputation, my concern is the actual usefulness of the solution. Answering your own question effectively shuts down communication/discussion of the question - meaning it will score lower in search results and not really be useful for people with similar questions in the future. I think you should be able to answer your own question if you do find an answer, but there should be a definite waiting period ... –  EAMann Aug 19 '10 at 17:05
    
Interesting, I hadn't thought of a waiting period. I can see how that would help, and that seems to be a common mechanism on SO sites. However, what if that led to people abandoning their question, and not returning to it with the correct answer after the requisite delay? –  tnorthcutt Aug 19 '10 at 18:04
    
That's a valid risk. If we implement some kind of timeout period, it should be one that makes sense - more than a day would be too long. Less than 10 minutes would be too short. –  EAMann Aug 19 '10 at 18:45
    
"It's also perfectly fine to ask and answer your own question, as long as you pretend you're on Jeopardy: phrase it in the form of a question." - FAQ. Jeff has also said several times on the blog that it's acceptable to ask a question and answer it yourself. –  Thomas McDonald Aug 19 '10 at 21:36
    
This has been dealt WRT StackOverflow. I'm not sure what the resolution on rep was (other than "rep isn't significant enough to worry about this kind of 'positive' gaming") but, IIRC, the decision was that self-answering was perfectly acceptable. There's even at least one badge for it! –  Bobby Jack Aug 20 '10 at 11:22
add comment

What if an asker's answer wasn't grouped with other answers, but was tucked up right under the question and easily identifiable as a self-answered question? That way someone else can consider the asker's answer and other answers as two different things. Plus, if some one else has a different way to solve the problem, they can post their answer, which might get higher votes than the asker's answer, and become a resource for the long tail.

This tucked-up answer is intended to be a badge of sorts that the asker didn't really research their question or make much of an effort before asking. A stigma, if you will.

I think asker answers should be grouped outside of the normal list of expert answers.

share
add comment

As another sample of this category of questions, I could use some advice on this question as I don't trust my own objectivity. I asked the question, and an answer was provided. Even though the answer pretty much duplicated code I already had, it did get me looking for "featured images" which made everything that followed possible. There was a ~2 hour delay between asking the question, and posting the solution.

So, what would you do? Mark my final code summary as the accepted answer? Vote to close? Leave it alone?

This doesn't exactly fit as an "answer" to EAMann's question, but it didn't warrant its own question, either.

share
    
The kind of ask-and-answer question that I'm concerned about is the "Oh, nevermind, I figured it out" variety. Your example, on the other hand, is a perfect example of the way things should happen. You asked a question, solicited answers from the community, and put together a superior answer to your own question. IMO your answer should be marked as "the" answer because it does the best job at actually answering the question based on your own research and the input of others. –  EAMann Aug 20 '10 at 19:13
add comment

You must log in to answer this question.

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