11

We have a large amount of low quality, one paragraph questions:

That's just a sample. There are many, many more.

In some cases, whole question is only the title. While I understand that sometimes there simply isn't much to say - this and this question being good examples - that is not the case for the above questions. The asker could and should have elaborated more, like explaining the situation that prompted the question or who the hell is Andrew Wakefield and what is paper was about.

It's better both from an SEO perspective, for the community - it's far more interesting to read questions that thoroughly explain a problem - and it helps to understand what exactly the asker is looking for.

Many of those questions have a link for "more information." That's not enough. For one, in the future, the page you linked to might go down. Secondly, there's no reason for you to not extract the most relevant quotes from that page and no reason for me to open another tab.

So, how can we get users to write more elaborate questions?

Downvoting seems a bit harsh, although it is sometimes tempting. I realize that most of those questions are users trying to seed the site with questions in order to avoid the empty restaurant problem but I'm scared that it sets a bad standard for the site.

  • 3
    +1 A valid concern! – David Hedlund Mar 3 '11 at 19:22
  • 1
    Can you provide examples of questions you believe are well-written? – Casebash Mar 6 '11 at 3:01
10

I think we need to be pretty straightforward here and educate people, and set the tone ourselves.

What makes Stack Overflow (my constant reference point) so great, in part, is that it's a fine balance between a Q&A mentality and a knowledge base mentality. It's not a knowledge base in that we shouldn't try to actively seed the site with questions. But it's not simply Yahoo Answers either, in that the case is closed as soon as the person asking the question is happy with it.

While I don't think actively seeding the site is the way to go at all, I do think we need to go to some length to let people know that the reason we're putting in all the effort is not only to help them, but because the answer can potentially help a lot of people, a lot later. I think it's important that we communicate this clearly. A simple comment "Hey, you're question's too short/not very clear" won't do. It's likely to be perceived as hostile, and quite off-putting to new visitors. They'll figure that "oh, come on, you understand what I mean," and you probably do. So they'll only think you're not interested in helping. This is where we need to show that we do want to help, but only on the premise that we can create a helpful answer for everybody. There are several aspects to this:

  • Simply answering "Nevermind, I found the answer myself" is an absolute no-no, of course.
  • Providing a link to an external site as a key part of understanding the question will make the question prone to become useless as soon as that site changes or disappears.
  • Relying too much on external context clues can be a problem as soon as context changes. Everybody right now might understand what you're talking about, because of world-known current events, but will anyone a year from now understand the question?

When people do any of these mistakes, we shouldn't just ask them to expand the question, but we should point to exactly what the problem is, and explain why it is a problem, because there might be no problem at all from the mindset that all you want is an answer to your question.

I think right now, in the upstart of things, is the time to be very polite, and treat these matters delicately. We should encourage and help people to create better questions, with helpful comments and concrete suggestions as to what they might add.

It is not obvious that this should always be the case. The SO community is not always very polite. A mindset have evolved which is similar to that of, say, a torrent community. "Hey, we're all making an effort to help here, people." In file sharing, there's the concept of hit and run (it's got many other names), for those who download stuff from others, and then immediately shuts down their client, without offering others the opportunity to download from them. Such behavior is openly fought, and those who engage in it are considered the lowest of the low. I think you'll see this in any climate where people are putting in a voluntary effort. Those who leech off the system entirely are treated with hostility. I see this in SO as well. Everybody is working hard on good answers, and then suddenly somebody drops in with a one-line question that completely reveals that he's asking here because he couldn't be bothered to google it. In those cases, well, we can't be bothered to help. The effort of those answering will always be greater than that of those who are asking the questions (for any given question, obviously not always on a site-wide level), so I think it's entirely reasonable to expect that those asking the question put in as much effort as possible.

I'm not very fond of the hostility I sometimes see on SO, towards those who write such questions, but I think it's a natural reaction to something that's working very well. It reveals that those people actually feel abused when somebody is trying to use up all their effort without putting in any of their own. If the attitude on Skeptics gradually fades into this, I think it is a by-product of something else we're doing very right, but at this moment, I think we should definitely try to be as polite as we can.

The issue of what's OK to edit is a very delicate matter. Edits should of course always be done with the highest respect for the author. In SO, it is very commonly accepted that you edit anybody's question or answer to correct grammar or code formatting. I think we can get away with that, too, and also edit titles so that they better describe what the question is about. When somebody is referencing a link, I'd say it's OK to pull in the relevant quotes, while keeping the link. Other than that, I would suggest that the author edits the text, rather than edit it myself, I think.

  • +9000 if I could. Absolutely amazing answer. – Borror0 Mar 3 '11 at 20:20
  • About your last sentence: Joel makes a good case that it becomes permissible to completely improve some questions after a while in The Wikipedia of Long Tail Programming Questions. – Borror0 Mar 3 '11 at 20:22
  • @Borror0: Thanks for the link, the Spolsky posts are always very agreeable. I got stuck on a phrase there, that I think encapsulates the entire mentality that makes the stack sites so exceptionally useful: "don’t just answer it to get a few points. That doesn’t make the Internet any better." Really. Making the Internet better. That is the emphasis. Not "helping a dude sort out his sql select clause" – David Hedlund Mar 3 '11 at 20:41
3

Don't be afraid to edit questions, or answers. That's the way you can MAKE them ask nice answers if they aren't doing it originally the way you like. One very easy thing anybody can do is add hyperlinks to Wikipedia for technical terms, and edit for spelling and grammaticality. Adding links to official sources that answer the question, even when it is easily Googeable, inside the question is also a good idea. Renaming titles to be more informative to future searchers demonstrates you are trying to build a knowledge base, and not just answer the question and set the question-asker off. In general, be a busy beaver, because that conveys that you are taking pride in the site and edifice that is being constructed here.

2

I fully agree that this is a problem and that we should try to improve the situation. I always try to put enough information into my questions that someone who had never heard of the subject still can get a rough overview what my question is about. I think we owe it to the users answering our questions to put some effort into them and not just dump one sentence in front of them and expect an answer.

But I don't think we can really do more than ask those users nicely to expand their questions and provide more context. Those users that visit meta regularly will see this question and hopefully improve their future questions. For those that don't visit meta we can only leave a comment on their question and ask them nicely to put more details into their questions.

A more drastic solution is to edit those single-sentence questions ourselves into something better. Such edits are a bit on the edge of what is acceptable, as we shouldn't change the substance of a question. I still think that even such substantial edits can be justified if the user is not responding to requests to improve his question.

You must log in to answer this question.

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