-10

A good reason for downvoting a question is a clear lack of research, asking more or less somebody else to solve your problem without having done any effort. Now, let's consider the particular case of a question posted together with a self-answer. I think it is a reasonable approach to put the research part basically into the answer. Somebody can disagree, that's not my point.

However what can happen is that a user might see the question and quickly downvote it for lack of research, not noticing that it had been submitted already with an answer. I am not able to point to any demonstrated case, I just want to point out that it can theoretically happen.

So my feature request is: Let's automatically mark question posted together with an answer with a clear information box like:

This question was intended for documentation purpose. The author has already provided an answer to the problem. Read about self-answered questions. You are welcome to submit alternative answers!

To avoid ambiguity, I am not speaking about a question posted and only later on self-answered. Also, I am not saying that an immediately self-answered question shouldn't be downvoted in any case, only that the downvoter should take an informed decision. I also want to mention the possibility that if you have never experienced this it might also been that you have been shielded by a good reputation score, which induced the possible downvoter to take a closer look to a "weird" situation (Why is this highly experienced user asking questions in this way?)

Not to hide (completely :-) ) anything, the question is inspired by another discussion which I prefer not to link here (if you are really curious you can try to find it, but I think it won't help to judge this feature request).

Edit: I now see that this idea has been previously brought up. In that case the proposed banner was:

This is a self-answered question, a behavior encouraged by stackoverflow. Vote on this answer as you would any other. You may offer an alternative to the original poster's answer.

The top answer suggested:

You might try including something like this at the top of your self-answers:

I encountered this problem earlier and managed to find the answer myself, so I thought I'd post it here in case anyone else is searching for the solution to this issue.

15
  • Self-answered questions do not get a free pass
    – random
    Mar 28, 2015 at 0:39
  • 4
    @random Why is a disclaimer a free pass?
    – Antonio
    Mar 28, 2015 at 0:40
  • 3
    Why do you need the disclaimer? Because the question is junk unless you tell people it's deliberately deficient?
    – random
    Mar 28, 2015 at 0:41
  • 3
    @random To motivate the apparent lack of research. The research has gone in the answer.
    – Antonio
    Mar 28, 2015 at 0:42
  • 1
    @random It seems to me that you've presumed that all questions that lack research are bad, but I disagree. Zero's example of a self answered question is very good, despite the lack of research. Mar 28, 2015 at 17:48
  • If it lacks research that's a reason to downvote. This proposal is to have a big display saying to ignore that and vote based on the answer. That already happens when you see crap questions with solid answers. @way
    – random
    Mar 28, 2015 at 17:51
  • 3
    @random I disagree that it's always a reason to downvote. Zero's example question is a perfect example of why. When a rule applied perfectly causes people to downvote a good question, then it is the rule that is in error, not the question. Mar 28, 2015 at 17:55
  • Never said it was always a reason to downvote. It's one of the reasons you can downvote. meta.stackexchange.com/questions/183847/…
    – random
    Mar 28, 2015 at 18:04
  • 2
    @random I think I understand what you're saying now. You are saying that some self-answered questions that have no research are bad (trivial ones, for example), and you worry that the banner would seem to give those questions a pass. Have I got that right? Mar 28, 2015 at 18:11
  • @random This question got a reversed upvote after your edit. Thanks a lot.
    – Antonio
    Mar 29, 2015 at 0:26
  • No problem. You're welcome
    – random
    Mar 29, 2015 at 19:30
  • @random Your edit is disrespectful of my line of reasoning. My line of reasoning started from preventing the risk an immediately self-answered question would be misjudged for apparently lacking of research. That's the only thing I am trying to prevent with this feature request.
    – Antonio
    Mar 29, 2015 at 20:58
  • If you're getting downvotes for an apparent lack of research because you put it in the answer instead of the question, rather than trying to change the process, how about trying to actually put the research into the question? It solves the problems, without having to modify the process.
    – fbueckert
    Mar 30, 2015 at 18:10
  • 1
    @fbueckert That's also an option, I sometimes go for that. Zero has shown in his answer some examples in which a "dry" question might be more useful to be used as reference documentation.
    – Antonio
    Mar 30, 2015 at 19:06

3 Answers 3

9

One of the projects of the community that's developed around the Stack Overflow Python chatroom is to build a collection of canonical answers to common questions, for use as duplicate targets.

A problem we face in that endeavour is that, when someone asks a question they don't know the answer to, they often don't know exactly what information is relevant, either. As a result, the closure of a question as a duplicate is frequently challenged based on differences in the two questions (or in the existing answer) that aren't pertinent to the underlying problem.

When we can't find a clear, focused candidate for a dupe target, we sometimes write our own. Here's one of mine:

Why do backslashes appear twice?

When I create a string containing backslashes, they get duplicated:

>>> my_string = "why\does\it\happen?"
>>> my_string
'why\\does\\it\\happen?'

Why?

As you can see, it's very short. That's deliberate – I tried to reduce the problem to its essence, so that it would be obviously relevant to whoever came across it. Without reading the accompanying answer, though, a user might well decide to downvote it for lack of research effort (in fact, someone did), or even vote to close it as a duplicate (which would defeat the purpose of creating a canonical Q/A pair in the first place).

It's pretty clear that Stack Overflow wants to encourage these kind of instant self-answers. There's even a bit of UI to make it easier:

"Answer your own question" UI

... but I know, both from my own experience and from that of other users I've talked to, that there's currently a disincentive to write these kind of "share your knowledge" Q/A pairs, precisely because of the fear that the question will be downvoted by someone who didn't notice the accompanying self-answer.

So, Yes, I think a banner to highlight "instant self-answer" questions is appropriate. I might change the wording slightly:

This is a self-answered question, intended to share knowledge with the community. Please read the accompanying answer before deciding how to vote. You are welcome to submit alternative answers!

1
  • 1
    That backslash question is written the same as any of the thousands already on the site. The only difference is who answered it, not how they answered, because if you look back the answers on older questions are as wordy. So there is no difference all the same.
    – random
    Mar 28, 2015 at 17:20
4

Self-answered questions aren't a glad-hand to free passing because the asker is leaving all/any detail in the answer and just using the question body as a subheading.

The reason you need such a disclaimer is that the question is deficient.

If it's deficient, you downvote.

You don't need such a warning message like street signs you see in residential neighbourhoods because the children are slow and playing in the area.

You need to write a better question.

5
  • 1
    I am not speaking about lack of detail. The question should contain all details to be well answered. But about the research efforts to find an answer.
    – Antonio
    Mar 28, 2015 at 0:51
  • 1
    Again, why do you need to special case a question just because the answer was posted at the same time, when if they waited a day it would be seen as lacking? Lobbing yourself a curveball isn't something to shield.
    – random
    Mar 28, 2015 at 0:53
  • 3
    What is the harm of a question that looks like "Please do my homework" if you post it with a working solution? A self-answered question is not a normal question. A self-answered question can be much shorter (without distraction from "I tried this and that and it did not work"), easy to retrieve and quickly understand by future users. I am not saying all self-answered questions should be like that.
    – Antonio
    Mar 28, 2015 at 1:03
  • 2
    A self-answered question is a normal question. If it wasn't then it would not belong on the site
    – random
    Mar 28, 2015 at 2:26
4

I do not support this feature request.

Every now and then someone comes on meta (this meta or SO's meta) to complain that their self-answered question was downvoted. They wonder if somehow self-answered questions are held to a different standard than other questions. Every time, the answer they get is that self-answered questions are no different from other questions and are held to the same standards. Well, at least that's the "party line", because some people seem to have a problem with self-answered questions. While we cannot read the mind of voters I've seen instances where the downvote was difficult to explain by anything else than the voter somehow thinking self-answered questions are cheating. (And indeed every now and then someone comes on meta asking whether it is at all okay to self-answer. So some people clearly do not know that it is okay. Perhaps they downvote thinking that it is not okay.)

At any rate, again, the party line is such questions are to be evaluated the same as other questions. The feature request here is in effect suggesting that voters in fact should treat self-answered questions differently.

Also, I am not saying that an immediately self-answered question shouldn't be downvoted in any case, only that the downvoter should take an informed decision.

(Emphasis mine.)

The implication here is that the fact that a question is self-answered should be a factor in how the voter votes, which flies in the face of the party line.

I would say in the case that Zero Piraeus brought up, the issue is not that the question is self-answered but that it is meant to be a canonical question that is addressing a problem that a lot of users consider to be trivial, and thus that can be answered with minimal research effort. Some users are going to consider that no research effort was expended and vote accordingly. Whatever benefit there could be in protecting such question from voting, the solution would not be to change how people approach self-answered questions in general but how they would approach canonical questions.

0

You must log in to answer this question.

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