What is meta? ×
Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 133 Stack Exchange communities.

I was wondering whether any attempts have been made to measure the time a user takes to write a post. It could be used to warn users (possibly only low rep users) that the question they want to post might be badly received if they don't show enough effort in their question.

My guess is this could eliminate some of the low effort questions of new users which still show up regularly. Instead of expecting users to read the faq before posting, a warning message can be useful for two reasons.

  • The user can reconsider posting the question. Hopefully adjusting it before posting.
  • The user gets an indication his question might not be received very well, preparing him for possible down votes. Nobody has to leave a comment redirecting him to the faq or anything, he already got the message.

As a clarification: as Joachim mentioned, this shouldn't prevent the user from posting. It should only serve as a warning message for new (low rep) users. If they feel their question is good enough as it is all that is needed is one extra click.

share|improve this question
4  
I'd like this, but it should not prevent them from posting. Requiring a second button press might be sufficient. –  Joachim Sauer Oct 19 '11 at 8:27

3 Answers 3

I find measuring the time for other purposes (like finding out how much time I spent writing stuff on SO, outside any kind of ranking or competition) interesting, but it's not fit to be a low quality filter - what if the user composed their question elsewhere, and pastes it into SO?

share|improve this answer
    
You already have sufficient rep. so you would never get the message. –  Steven Jeuris Oct 19 '11 at 8:31
4  
@Steven I don't mean "I" literally here - changed. It's an interesting idea but I think it would yield too many false positives. Maybe it might be worth informally looking at the time taken, though, to see whether there is a pattern that could be used? –  Pëkka Oct 19 '11 at 8:32
    
I don't consider a false positive to be a bad thing per se. The question can still be posted, only a confirmation message would pop up. Once the user gained enough rep (he knows how SE works) he will never see it again. –  Steven Jeuris Oct 19 '11 at 8:35
7  
@Steven yeah. If it can detect paste events (and disregard the time if there were any), I agree it might be worth a shot. But I'm not convinced that most terrible questions take considerably less time than many half-way decent ones. It might be worth informally "timing" a sample set of questions first to find out whether the data suggests there is a strong correlation –  Pëkka Oct 19 '11 at 8:45

I'm with Pekka -- while I'd find it vaguely interesting how much time is spent on questions / answers, I feel it has no bearing on the quality of the question/answer. There's far too many factors at play:

  • The poster may have composed their response elsewhere and copy/pasted
  • The question may have been simple and not required a long time to ask/answer
  • The poster may have had to search for relevant information
  • The poster may have stepped AFK
  • Some people type faster than others
  • The post may have initially been posted quickly and then edited with more detail
  • The post may be modified a copy/paste from another source like a forum question
  • Etc etc etc

Basically, it might be interesting but I strongly don't think it's useful information to be used for showing an sort of warning. And I don't think it's worth the time for the team to build this.

share|improve this answer
    
"The poster may have had to search for relevant information" ... isn't that what we want to promote? –  Steven Jeuris Oct 19 '11 at 12:26
    
Not always. What I mean is, they may have had the code open and they could copy and paste it. Or maybe they didn't have the code with them and had to SFTP it from somewhere. Or, they may have had to locate some reference to cite on the internet. My point is, all questions are different and a delay doesn't mean the same thing all the time. –  The Unhandled Exception Oct 19 '11 at 12:33
    
It's not the delay which is interesting, but the lack of one. The fact that a delay can have different causes isn't a counter argument for the fact that the lack of a delay might indicate a low effort question. –  Steven Jeuris Oct 19 '11 at 12:39
    
I don't think the lack of a delay indicates a poor question. Besides we already have a number of means of identifying poor questions. –  The Unhandled Exception Oct 19 '11 at 13:58

I agree with the Unhandled Exception on this question, and want to generalize it: There is no functional identity between size of question and its quality.

If you argue that there is a correlation, I would answer: prove it! There might be a correlation, but with a high standard deviation, I guess.

A lot of posts include needless introduction

Hello! My name is Ben, and I'm new to ruby.
I hope I don't bore your with my question, which I looked up at google, facebook and Youtube ...

(question core)

Thanks for helping me out. This is really urgent for me.
Ben
Tycoon of Yoyodyne Inc.

share|improve this answer
    
In order to prove it, it would have to be measured first. P.s.: I don't recall reading that many awful questions which also have a needless introduction. It's the likes of questions like this one I hope this would address. –  Steven Jeuris Oct 19 '11 at 12:45
1  
a) You see introductions rarely, because I edited them. b) A single question with a single attribute doesn't show much. The question is unclear - that's the main problem, imho. A longer version needn't automatically be better - think of introduction and good-bye-greetings, not to mention comment-vodoo <!--do you voodoo?--> –  user unknown Oct 19 '11 at 13:11
    
a) Seems like you missed one. b) O K, here you go. –  Steven Jeuris Oct 19 '11 at 13:24
1  
To clarify, my point is that users who actually bother to write introductions, in my experience usually pose a valid question. They just don't know SE conventions yet. As Unhandled Exception's answer, this isn't a counter argument that the lack of a delay might indicate a low effort question. –  Steven Jeuris Oct 19 '11 at 13:31
    
We don't have a measurement for the time taken, but for message length, which could be more appropriate, since inserted code can and should come from an editor, and might have been pre-build. While interruption by phone, twitter or coffee pot don't improve the code at all. The second measure could be the number of up-votes after some time, but only unedited posts could count here. Maybe early edits are a better measurement, or close votes, closed questions. –  user unknown Oct 19 '11 at 15:34

You must log in to answer this question.

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