Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 158 Stack Exchange communities.

What is meta?
Here's how it works:
  1. Any Stack Exchange user can ask a question
  2. The community provides support, votes on ideas, and reports bugs
  3. Your voice helps shape the way Stack Exchange operates

I'm beginning to notice an increase on the number of things that fail to meet basic SO expectations:

  • Zero research
  • No actual question, just a vague specification of what should happen, without any sample code
  • <50% acceptance rates from repeat questioners

Granted, newcomers to the site might be unaware of these problems, but that, in itself, is a problem.

I don't believe these expectations are well communicated to users. If I roll my mouse over the upvote and downvote arrows, it indicates why I might want to do so and refers explicitly to research.

However, when I ask a question, at no point am I informed that I should have

  • Done some research
  • Indicated what research I have done.

In addition, there is no explicit indication that basically asking SO to write your application for you is A Bad Thing.

As for the last part, since my acceptance rate isn't low, I don't know what occurs - do low acceptance users get a warning that they're unlikely to get responses if they have low acceptance percentages?

I guess what I'm saying is: can we do more to help newcomers write questions that aren't automatically downvoted and closed? It's a lot nicer to handle messages from the system while you're writing your question than handle comments, votes and closes in a public forum.

share|improve this question
Indeed, we could compile a list of common mistakes that result in newcomers being downvoted without them being the wiser, along with reasons why and advice against. We could call it something like What Stack Overflow is not. Looks like a good idea. On paper. – Frédéric Hamidi Jul 8 '12 at 14:56
Interesting post. I think the difference here is that this would be displayed to users in the privacy of their Question Asking screen, as opposed to the publicly visible forum of SO. – Dancrumb Jul 8 '12 at 15:06
Your third bullet certainly isn't an expectation of mine. I'd far rather have a good question asked by someone who happened never to accept any answers than a bad question from someone with a 100% accept rate. In fact, I make a point of trying to answer any reasonable question which has several comments telling the OP to improve their accept rate (as well as flagging those comments as not constructive). If the comment is a nice one, including a link to the answer acceptance page, that's fine - but all too often it's just "improve your accept rate". – Jon Skeet Jul 8 '12 at 18:54
try clicking "ask question" in Incognito/Anonymous/InPrivate mode in your browser. Just sayin' – Jeff Atwood Jul 9 '12 at 5:45
@JeffAtwood: Thanks - hadn't occurred to me to try that – Dancrumb Jul 9 '12 at 14:40
@JonSkeet: I agree that a good question for a low accepter is better than a bad question from a high accepter, but I think you're arguing from extremes there. I think there's a correlation between users who perform little to no research and users who don't engage in SO at the level of actually acknowledging the assistance they get by assessing and accepting the answers they receive. – Dancrumb Jul 9 '12 at 14:42
Related:… – Faust Aug 10 '12 at 13:01

New users are already shown the How to Ask page that instructs them to do research (and even has a search box!) along with making a few other suggestions.

This page is also linked from the sidebar for every user going to ask a question (the link is called "asking help").

The bottom line is that there are people who just see a textbox and will type stuff into it without regard for anything else. The people who would pay attention and read already do and are presented with the information they need. People who don't care... well, it doesn't matter how much text you put in front of them, they'll just click or scroll past it.

do low acceptance users get a warning that they're unlikely to get responses if they have low acceptance percentages?

No, but it has been proposed as a feature request.

share|improve this answer
As I mentioned in another comment here, I bet users don't even see the 'How To Ask' box. Sidebars are typically positioned on the right of a page so as not to detract from the main page content. In this case that UI design pattern is working against the site's goals. – Will Tower Jul 9 '12 at 5:44
@skinny she is not referring to the sidebar. There is a mandatory interstitial page, a full page, this page on ask for all new users. Try it yourself if you don't believe us. – Jeff Atwood Jul 9 '12 at 5:46
oh - my mistake. Guess I haven't seen that in a while. – Will Tower Jul 9 '12 at 5:48
The interstitial says nothing about the process that follows: votes, accepting answers. It's an unusual interaction model (you can up-vote and accept an answer??), and its not obvious to the new user. – Faust Aug 10 '12 at 13:05

enter image description here

See that yellow box at the right? All the information you are looking for is there, but I do agree that it's not really clear and I don't think most newcomers click either link at the bottom. However if you follow both links, I honestly don't see how you can get it wrong. The FAQ has a very nice summary of what you can and cannot ask about, and the "asking help" link takes you to the "How to Ask" page, that summarizes the basic requirements for questions:

  • Do your homework
  • Be specific
  • Make it relevant to others
  • Be on-topic
  • Keep an open mind

Back when I was a new user, I did click on both links before asking my first question, and whenever I'm visiting a new Stack Exchange site, the first thing I check is the FAQ. Granted, I'm very familiar with how Stack Exchange works, and still have struck out once or twice (it happens ;). I'm starting to think that we are at a point were we have too many rules & guidelines, and newcomers just take the easy way out: Ask the question, and if it's get closed, it's closed. Trial and error is what we do, we are programmers after all ;)

It's a problem, a problem I don't have a good solution for. As I already mentioned the information is there, the documentation is adequate (imho), what seems to fail is that people just don't read it, or when they do, it's a bit much and they either skim quickly through it or just read the first couple of paragraphs.

Furthermore each Stack Exchange site has its own culture and its own set of secondary guidelines, for example Stack Overflow users the [homework] tag to denote homework questions and treats them slightly differently than other questions, and Programmers has eradicated the tag and treats each question by the same standards, homework or not.

I guess one solution would be to require people to read both pages, the FAQ and How to Ask, before letting them post their first question. It'd be quite obnoxious, but it's something we already do for answers, there's a big overlay covering the answer textarea when you try to answer your first question. I don't know how effective that is, on Programmers were I'm a moderator I see people skipping it all the time (which raises an automated flag), Stack Exchange probably has or can easily get hard data to tell us if the answer overlay works or not - combining whether the user skipped it or not with the eventual fate of the answer (upvoted / downvoted / flagged / deleted).

At the end of the day, though, the problem is more of a social problem than a technical problem. People can choose not to read any kind of documentation we throw at them, however in your face it is. Polite guiding comments and edits from more familiar members of the community is what works best for questions that are a bit ..., but salvageable.

System messages can be ignored, similarly to how the FAQ is ignored a thousand times per day, but I don't think a comment along the lines of "Hey, welcome to Stack Overflow! Can you tell us what exactly the error message you're getting is?" can be easily ignored. And if it is, well, just down vote and go on with your day...

share|improve this answer
To be honest, considering how much I ignore the right side of the SE network (re: Moderator Election notice I missed), I could understand how nobody would read it prior to asking. – user7116 Jul 8 '12 at 15:21
Feel free to borrow my free hand circles that illustrate the entire process. – Tim Post Jul 8 '12 at 15:35
I think people simply focus on the question fields - selective blindness. If that box were on the left instead of the right (since English reads left to right) the chances of it at least being scanned would I think increase. – Will Tower Jul 8 '12 at 16:22
@skinnyTOD That'd make for an interesting feature request, but I'm not really familiar with UX to tell you if it's a solid request or not. Quite interesting though. – Yannis Jul 8 '12 at 16:24
@TimPost, maybe literally borrowing your free hand circles and arrows and using them where you've put them might be quite a good idea :-). – ben is uǝq backwards Jul 8 '12 at 20:51

However, when I ask a question, at no point am I informed that I should have

  • Done some research
  • Indicated what research I have done.

Yes, it may be someone's first time on Stack Overflow. Is it also their first time

  • doing research?
  • using Google?
  • solving a problem?
  • communicating with others?

No, I'm afraid that attitude is something that some people just bring with them when they come here. I have also seen great questions from people who seem to understand that being "new" doesn't really mean anything.

I don't understand why newcomers are given such leeway. "I'm new here" seems to be something like a hall pass. Newcomer means they have never visited this particular forum before, not that they are brand new to the world. It is each person's responsibility to avail themselves of the FAQ and the rules of the forum before they begin blasting out messages like you describe:

  • Zero research
  • No actual question, just a vague specification of what should happen, without any sample code
  • <50% acceptance rates from repeat questioners
share|improve this answer
You raise some interesting points, but I think you misrepresent the crux of the problem. A drowning swimmer will splash and gasp, trying with all their primal instincts to get oxygen, even if it means pushing a nearby lifeguard under the water. In the same way, an inexperienced programmer will struggle and start gasping for some relief. They are more likely to rush to a forum (often a successful one like SO) where they can get (from their point of view) quick help tailored to their needs, rather than digging through Google. They will do this even if it means dragging the forum down with them. – gobernador Jul 9 '12 at 16:21
My point is that asking on Stack Overflow is far easier for an inexperienced programmer looking for a mentor than reading a blog post. The attitude you reference is not a conscious decision, but rather a primal reaction to a problem. – gobernador Jul 9 '12 at 16:23
@gobernador Many so-called "new" users are able to follow forum rules. You are confusing "inexperienced programmer" with "inexperienced forum visitor". I may not know how to write C#, but I know how to ask a C# question. Chances are good that someone new to Stack Overflow isn't also new to forums in general. Nor are they new to searching Google. At what point is there an expectation that people new to a given forum be responsible for learning its rules? – JimmyPena Jul 9 '12 at 16:29
Going back to drowning analogy, notice the difference between an experienced user's question and an inexperienced user's question. One of these askers is calm, collected, and fully understands his problem. This is good. The other is drowning in a problem he is having. While he's drowning, he's unlikely to sit down, read the FAQ and how to ask. He's going to throw as much information out there as he can and hope for relief. – gobernador Jul 9 '12 at 16:48
For people like him, you can expect him to learn the rules, but he won't, because to him, getting out of the water is the only thing that matters. – gobernador Jul 9 '12 at 16:48
@gobernador Going back to drowning analogy -- Sorry, I still don't get it. I don't know how to swim, but I know how to ask someone how I should do it. And it is still my responsibility to learn how, before I try. – JimmyPena Jul 9 '12 at 16:58
I confess, it was probably a poor analogy. I agree with what you say, I just want to point out that while it is reasonable and responsible to expect everybody to learn how to use a tool before using it, desperate people don't think in a logical, level-headed fashion. – gobernador Jul 10 '12 at 5:12

One thought I had to help improve the quality of questions would be a simple UI nudge in the form of some shadow prompt text (as appears in the title and tag bars). As large and obvious as the sidebar instructions are, a user is more likely to simply be focusing on the question fields.

This prompt text could list a few questions -

1. Have you done any research before asking your question?

2. Do you have code...

3. etc

Keeping the text short and to the point. A numbered list rather than a block of text would be more eye catching and indicate a sort of 'pre-question' checklist. And of course the text would disappear once the user selected the field and started typing.

Might not help but couldn't hurt.

share|improve this answer

The problem at heart here (as I see it, and in simplified form) is that the noble goal of the site is to create a great repository of knowledge, for the benefit of the greater community.

However, the vast majority of new users don't come here to improve the Universe's body of coding-related knowledge, they come to solve a problem they are fighting with at the very moment they are typing the question. Which is a very understandable position, btw.

This relative incompatibility between the philosophy of the site and the mindset of a large proportion of new users is what makes the problem unsolvable as such. There are flags to cover all this, but as can be seen from the discussions here on meta, it is not really deemed desirable to use them on new users, unless it is a really blatant case.

Personally, I believe that the work-around is simply to provide a filter, based on some sort of soft flagging or voting. That way you can configure your home page and pages of tagged questions to show or hide these questions that actually are not a good fit for the site but tolerated anyway.

It is a bit like closing the windows if you don't like the noise in the street - simple, easy and doesn't involve the cops. The problem right now is that the house has been built with windows that can't be closed.

share|improve this answer
they come to solve a problem they are fighting with at the very moment they are typing the question Actually, that's a good thing, that's exactly what we want, Stack Overflow works best when questions are about solving a specific problem and it fails horribly for anything else. We can improve sub par questions that are about specific problems, it's all the discussion oriented questions, do my homework for me, give me a tutorial because Google is broken for me, etc that we can't do much about (other than close aggressively). – Yannis Jul 8 '12 at 16:15
Yes, I agree, but I am also convinced that part of what we are after is for instance to narrow the problem down to an isolated and reproducible problem. Otherwise it won't help anyone. I guess what I am after are the questions consisting of a page of code and a blatant request to the community to debug that code. – Monolo Jul 8 '12 at 16:19
I don't think of this as an "incompatibility" -- in fact, to put it crudely, those who are interested in the repository aspect are using the need of askers to get their individual problems solved -- that need generates content for the repository. Both parties' goals can be achieved much of the time, although certainly there can occasionally be tension. – Josh Caswell Jul 8 '12 at 19:23
I think my experience is different. I'm seeing a lot of people coming to SO without a specific problem, beyond the fact that they're trying to create some application and don't really want to learn how to do it. There are a seemingly increasing number of questions that can be boiled down to "How do I create the following webapp", which show a complete lack of research - indeed these questions are opting to put a request on SO in lieu of research. – Dancrumb Jul 8 '12 at 20:40

You must log in to answer this question.

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