In this question, there is a factual error about the documentation in R. I believe the actual question being asked is a useful one.

Should I correct the factual error, or leave the correction as a comment, as I have already done?

  • 2
    $\begingroup$ I'm interested in hearing what the community thinks. In the meantime, if you have commented about an inaccuracy in a question or an answer and the writer will not correct it (or does not respond within a reasonable time, which should be at least a day), then you should downvote it: that strengthens the message and provides more incentive to fix the problem. Just be really sure you're right and they're wrong! $\endgroup$
    – whuber Mod
    Jul 2, 2012 at 16:39
  • 6
    $\begingroup$ My personal opinion is that a comment to seek clarification (or to make a clarification) should be preferable if your edit would potentially materially change how the question is likely to be understood. $\endgroup$
    – cardinal
    Jul 2, 2012 at 18:17
  • 1
    $\begingroup$ @whuber I really only mostly lurk here, but I added my 2 cents below, strongly agreeing with cardinal. $\endgroup$
    – joran
    Jul 8, 2012 at 0:48

2 Answers 2


I generally only lurk here at CrossValidated, but based on my experience at StackOverflow, I believe that the crucial point here is whether the factual error is in a question or an answer.

One will frequently see something that seems obviously wrong or mistaken in a question. If the asker wasn't confused about something they probably wouldn't be asking a question, would they?

The rule I follow is that I will never "fix" mistakes in questions unilaterally. The mistake in the question may be the very source of their confusion and hence would be essential to answering the question. Instead, I always leave a comment pointing out the error and asking for clarification. But even then, I would let the OP fix any errors. Even if they fail to respond after a week, there is too much danger that editing to "fix" the error would radically alter the meaning and intent of the question. My comment would remain visible, though, to alert others readers of the issue.

On the other hand, if I spot an error in an answer I would adopt an approach more like what is sketched out in @atiretoo's answer: leave a comment, possibly downvote, and wait. If there's no response in a suitable amount of time, perform the edit myself.

Finally, this obviously applies primarily to errors in content, rather than style.

  • 3
    $\begingroup$ (+1) Very nice. In particular, I've found that factual errors in questions tend to be directly related to the reason the question is being asked, even though that may not be apparent at first to either the OP or those of us reading the question. $\endgroup$
    – cardinal
    Jul 9, 2012 at 2:33

So based on 2 comments worth of community input, here's a proposed "best practice".

  1. Point out the inaccuracy in a comment.
  2. If the writer has not corrected after 24 hours, downvote the question, OR
  3. Edit the question to correct the error as long as it does not materially change the question.
  • 3
    $\begingroup$ The 24-hour rule would need some adjustment to take into account weekend, public holidays, etc., and users that don't necessarily check the website on a regular basis. Also, keep in mind that unregistered users don't get system-wide notifications. $\endgroup$
    – chl
    Jul 4, 2012 at 19:55

You must log in to answer this question.

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