17
$\begingroup$

Recently, I answered a question on Math.stackexchange, but it turned out to be wrong. However, I do not know why it is wrong. Is it appropriate to create new questions asking about why the answer I wrote was wrong for the question?

$\endgroup$
  • 12
    $\begingroup$ I don't see anything wrong with a new question giving links to the previous one. I do wonder why you now think your answer is wrong; if it is just the OP indicating a lack of satisfaction, don't worry too much $\endgroup$ – Will Jagy Feb 18 '17 at 16:19
  • 1
    $\begingroup$ You may link up a new question, as it may have some very imp logic that was missed or not included. However as @WillJagy mention, if it just the owners choice not to select yours, then please don't bother. It had happen many times, that people will just copy paste your answer and get "selected/chosen" and you will miss that point(s). $\endgroup$ – Smit Feb 18 '17 at 18:57
  • 12
    $\begingroup$ If you post such a Question, explain the basis for knowing "it turned out to be wrong" while at the same time you "do not know why it is wrong". There are other Questions posted of this kind, in the sense that someone presents an argument or solution that is more or less clearly wrong (gives the wrong conclusion), but the gap or fallacy in argument is not yet found. $\endgroup$ – hardmath Feb 19 '17 at 0:11
  • 7
    $\begingroup$ Since no one else has said it yet, I recommend removing your answer (if you're sure your answer is actually incorrect) or at the very least putting a disclaimer at the top of it, so that future readers of the question and answers aren't misled. $\endgroup$ – tilper Feb 20 '17 at 18:56
  • $\begingroup$ It is appropriate to ask a new question, once you link the old one: however, if you like, we can go to the page containing that answer itself, and finish the matter then and there! $\endgroup$ – астон вілла олоф мэллбэрг Feb 21 '17 at 4:11
  • 1
    $\begingroup$ Would you give a link to the answer you mentioned? $\endgroup$ – Jack Feb 25 '17 at 2:36
  • 2
    $\begingroup$ If you have absolutely any doubt in your answer now then you should've never posted it to begin with. Clearly you thought about it some more and realized you made a mistake. Please do not blindly make posts like that in the future. $\endgroup$ – The Great Duck Feb 25 '17 at 23:25
  • 12
    $\begingroup$ @TheGreatDuck - I strongly disagree. It is one thing for a user to be flip or careless. However, it is easy to write a careful, thoughtful answer and then later realize, or have it pointed out, that one has made a mistake. This is part of normal mathematical life, not evidence that one should not have posted in the first place. The OP is in good form as long as they edit the original answer to acknowledge the uncertainty once it arises. $\endgroup$ – Ben Blum-Smith Feb 27 '17 at 14:23
  • $\begingroup$ @BenBlum-Smith true. what do you think the last sentence if my post was meaning? If the answer is flawed then it shouldnt have been posted. I never said they are in trouble for it. Im just telling them to be more cautious. If they are merely indecicive on the correctness (due to personal doubts in themselves) that is even worse as they shouldnt have unwarranted doubts if they trust their own abilities. $\endgroup$ – The Great Duck Feb 28 '17 at 19:31
  • $\begingroup$ Not an answer but perhaps useful. When I post a wrong answer and it's pointed out I try to leave the answer up, edited to flag it as wrong, because the way it was wrong may help someone in the future. $\endgroup$ – Ethan Bolker Mar 1 '17 at 19:25
  • 6
    $\begingroup$ "If you have absolutely any doubt in your answer now then you should've never posted it to begin with. " Are you sure you applied the same strict criterion to your comment? $\endgroup$ – Phira Mar 2 '17 at 14:11
13
$\begingroup$

(a) Yes it is totally appropriate to ask a new question inquiring about the old question. Describe the setting from scratch in the new question so the other users don't have to follow a link in order to understand what you're asking, but also definitely do provide a link to the old question you are asking about.

(b) But maybe in this case I can save you a step? Based on your posting history I assume you are thinking of this question:

Combinatorial proof for party-goers.

IMO, your answer is correct for the question as it is written. Your answer differs from the accepted answer because both the OP, and the accepted answer, are implicitly assuming (without stating it) that the group selected for the carpool has a fixed prespecified size $k$, while your answer is addressed to counting carpools of all sizes from $1$ to $n$.

$\endgroup$

You must log in to answer this question.

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