30
$\begingroup$

This is a scenario I have encountered a few times when going through the edit suggestion review queue.

An asker posts a question and as context they provide some information that they used when trying to solve the problem. However, part of this information is just plainly wrong or perhaps misunderstood. It could be a simple typo or other errors. The error might not be critical for the question, but in some cases it is crucial. Then a user comes along and suggests an edit to correct the error.

One example, where the error made probably is not critical for the solution, is this question. The editor suggested to correct an error in the condition for convexity. I chose to reject the edit suggestion with the "deviates too much from the original intent" reason, since I thought it just might be a chance, although small, that the error was part of the reason for the OP to ask the question in the first place. Would you say that this was a correct assessment? For what it's worth, the edit was eventually approved by two other reviewers.

In general, how careful should we be when editing in corrections of the mathematics in a question, especially when it comes to the description of what has been tried by the OP? In my opinion, we should be very restrictive and post comments asking the OP for clarifications instead of just editing, even if it appears to be a glaring and blatant typo.

$\endgroup$
7
  • 9
    $\begingroup$ I don't have the time to check the details of the specific incident you link to, but IMHO you did the right thing. There are many occasions, where a misunderstanding of the context/condition is exactly why the question needed to be asked in the first place. However, it is a bit difficult to formulate an appropriate norm or policy. $\endgroup$
    – Jyrki Lahtonen
    Oct 7, 2016 at 12:35
  • 4
    $\begingroup$ I agree. Comments are the first response, and possibly the last. The assumption of some mathematical sense is often confounded by MathJax edits intended to give visual appeal. Ideally Readers will hold off until an OP weighs in with clarifications. $\endgroup$
    – hardmath
    Oct 7, 2016 at 14:21
  • 3
    $\begingroup$ You did the right thing. Even punctuation repair needs to considered carefully as the panda will attest to. $\endgroup$
    – copper.hat
    Oct 7, 2016 at 14:57
  • 9
    $\begingroup$ It's admittedly a judgment call, but if, to the best of your ability, find that an edit radically changes what is being asked, or sweeps away the misunderstanding that is at the root of the question, the edit should be certainly rejected. This is part of the more general rule that somebody who's about to approve edits should parse the edit thoroughly before accepting or rejecting, and that the lazy have no place in approving edits. $\endgroup$
    – J. M. ain't a mathematician
    Oct 8, 2016 at 4:32
  • 2
    $\begingroup$ For what it's worth, in my opinion, glaring and blatant typos should be corrected without comment. Anything subtler than that merits a request for clarification. Just my two cents. $\endgroup$
    – Mr. Brooks
    Oct 10, 2016 at 21:17
  • 3
    $\begingroup$ I've had similar experiences--rejecting an edit because it "fixed" something that might possibly have been the source of the OP's confusion, but two other editors approved it. In this case it seems the OP was having difficulty finding a counterexample even assuming a correct definition of convexity (judging from the answer they accepted), so I wouldn't roll this particular edit back at this time, but in other cases it seems this other Meta question is applicable: meta.math.stackexchange.com/questions/9079/… $\endgroup$
    – David K
    Oct 11, 2016 at 12:20
  • 2
    $\begingroup$ I noted the contrast between how that edit was handled, and the edit performed by a higher-rep user earlier, which did not change the meaning of the question at all but was still followed up by a comment asking OP to verify that the edit was correct. $\endgroup$
    – David K
    Oct 11, 2016 at 12:22

0

Reset to default

You must log in to answer this question.

Browse other questions tagged .