I recently made a drastic edit to an accepted answer because it was wrong. The edit was rejected for the following reason:

This edit deviates from the original intent of the post. Even edits that must make drastic changes should strive to preserve the goals of the post's owner.

One of the suggested questions even discusses this topic,

If the answer is older, and maybe not likely to get engagement from the original author, you can opt to just edit the answer to make it correct. However, you run the risk of changing the answer's meaning. That path invites a possible revision war between you and the original author and/or community readers who disagree with the scope of your edits. So unless the edit is just making minor corrections (such as fixing typos, or easily-supportable technical fixes that are not opinion-based), I strongly do not recommend this route.

~ How to discuss the technical accuracy of an answer?

The suggested route is to downvote the answer and make a new one. I would argue that if there is an accepted answer to an old question with, in my particular case, 12 upvotes, a new answer is astronomically unlikely to usurp the accepted one. This policy of nonediting (and/or nonredacting) of incorrect accepted answers leads to the proliferation of misinformation without recourse from the community.

Why is the current policy the way it is? Are the moderators open to a policy change that will improve the information content of the site?


This really isn't likely to change. SE is clear that an answer isn't marked as "correct" but rather "accepted". If it helps the OP, marking it as the accepted answer is appropriate regardless of if it is correct.

Fundamentally, a crowd sourced platform is not an ideal place to establish correctness. This is the same reason why academia does not accept Wikipedia as a reference. Places like SE and Wikipedia are fantastic resources, but they draw on community knowledge, not expert knowledge.

In order to materially alter a question, two things would have to be dealt with, first, you are miss-attributing a viewpoint. The author's view is not that which is being presented under their name, so there are speech implications here that are non-trivial. Converting a post to be a community wiki answer is an option, however this comes to the second issue.

Stack Exchange has no system of review for establishing correctness. Who is to say which post is more correct? The OP may not believe you are correct and believe that the answer should remain unaltered. We have no mechanism for making these determinations and moderators are specifically not supposed to make these kinds of determinations, so effectively, community wiki would just be a) stealing the answer they started with and b) hijacking it to something we can't establish as being more correct.

The proper response for an incorrect answer is to a) down vote the incorrect answer, b) add a comment to the incorrect answer pointing out what is wrong with it and how it can be improved and if possible, c) write a correct answer or upvote existing correct answers. If it's particularly distressing, it may be worth bringing up in chat to get some additional eyes on it to help with down voting and upvoting if they agree with you.

I know that feels non-ideal, but for what SE is, there isn't really a better alternative that doesn't either invariably degenerate to feuds and petty conflicts or unacceptably delay the answering of questions while an academic peer review process of known experts occurs (and which SE doesn't even have the beginnings of a foundation for).

  • 1
    Another minor point about "hijacking" a marked-as-accepted and/or high-voted answer: not only is it misattributing the answer author's viewpoint, but it misattributes the voters' votes and OQ's acceptance as well.
    – scottbb Mod
    Nov 6 '17 at 3:28
  • @scottbb - true. Personally I do find it quite a bit less distasteful to misatrribute anonymous things like votes if we were able to academically rigorously verify the "correct" answer compared to saying a particular person said "xyz" when they didn't. There's a huge difference between "oh no, my vote count got used wrong" (which is still allowed to happen if the original answerer changes their answer) vs "you are saying I believe something I don't believe".
    – AJ Henderson Mod
    Nov 6 '17 at 3:40
  • 100% agreed. I should have emphasized minor point. =)
    – scottbb Mod
    Nov 6 '17 at 3:43

Why is the current policy the way it is?

Because your version of the answer might not be quite as correct as you think it is. The author of the answer you "corrected" might well see things quite differently than you do, and when you edit his/her answer to make it "better," you've materially changed the meaning without permission. How would you feel if someone "improved" an answer you'd written in a way that completely changes the meaning in a way that you disagree with?

This policy of nonediting (and/or nonredacting) of incorrect accepted answers leads to the proliferation of misinformation without recourse from the community.

Again, how do you know that the answer in question is incorrect? More important, why are you reluctant to offer your own answer and let the community judge it on its merits? If your answer is better than the accepted one, it will get more up votes over time. More importantly, if you make a compelling argument for the correctness of your answer, you will have effectively stopped the "proliferation of misinformation" that you're worried about. That is your recourse.

  • Re: the second part. I "know" my answer is correct because it is well researched with citeable sources, and the answers I'm discussing are often completely out on space (e.g. drawing light coming to a focus in the aperture of a lens). As I discussed in my post, new answers to old questions are extraordinarily unlikely to gain any traction vs a highly upvoted and accepted answer. Nov 13 '17 at 17:37
  • 1
    New answers to old questions don't get a lot of attention because people tend not to read old questions as often as new ones. But if people don't read old questions that much, the problem you're trying to solve isn't a big problem after all. But since you've written a new answer and claim that it's more correct because it is well researched with citeable sources, then you should explain the research you did and cite your sources. As it stands, your answer contains two links, and neither one does much to support your assertions.
    – Caleb
    Nov 13 '17 at 18:26
  • A problem not being the biggest issue doesn't mean it shouldn't be addressed. Asking for citations, etc, is beyond the norm of SE and unreasonable for this platform. Nov 13 '17 at 20:26
  • 1
    It's not unreasonable if you want to convince people that yours is the correct answer. Having read your answer, I'm far from convinced that it's better than the accepted one. Citations don't always have to be in MLA or APA format -- links to persuasive and relevant supporting information would help, and doing that much is absolutely the norm for good answers. Otherwise, you're asking us to just take your word that you're right.
    – Caleb
    Nov 13 '17 at 21:52
  • Do you disagree that the red and blue elements of the image I linked to are darker than the green or neutral colored ones? Nov 13 '17 at 22:27
  • @Caleb Old answers to old questions will tend to rate higher in search results than new answers. That's one of the problems with time dependent answers on StackExchange sites. This highly rated answer considers 20mp to be extremely high. True when written. The question is top for two of its three tags and second on the other.
    – user50888
    Nov 17 '17 at 18:42

You must log in to answer this question.

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