Reviewing edits that substantively modify an answer

Question

What is the appropriate way to handle these kinds of edit reviews?

1. The edit claims that some facet of an answer is wrong and proposes replacing the "incorrect" part with the correct information. The change isn't enough to warrant an additional answer and a comment is transient. Is the appropriate edit to replace the information or add the correction as supplemental information (if it should be done at all)? Also, does it make a difference in how the edit should be done whether the edit includes a citation validating the correction? Does it matter whether the answer has been accepted and/or upvoted?

2. An old answer. The edit claims that the answer is no longer correct (for example, because the current version of the software is different). Should the edit be in the form of supplementary information that refers to the newer version of the software? Should some form of citation be required?

3. An edit like 1 or 2 is proposed by an anonymous user who provides no citation and the reviewer cannot readily validate the information. Should unattributed substantive edits be accepted?

4. Should the reviewer be in the position of "ruling on" substantive changes of this nature? Answers get the review and scrutiny of the community and that process is bypassed when these kinds of changes are done as edits.

5. Is there a mechanism to provide a comment to the user proposing an edit to explain why an edit is rejected or to request additional information, substantiation, or changes to the edit that only they would have the knowledge to make?

Edit: There seems to be a grey area. On the Tour page and the Help pages, edits are described as: "to correct minor mistakes or add addendums / updates as the post ages" (in addition to grammar and spelling mistakes). For community wiki posts: "...others will edit them to keep them up to date, to add useful information,...". So it appears that community wiki posts actively invite at least updates to keep them current and expansion, while regular posts have a more restrained policy.

However, it looks like addendums and updates for currency, even on regular posts, are part of the plan, and correcting "minor mistakes" may be somewhat of a judgement call. Can anyone elaborate on the scope of what is appropriate for "correcting minor mistakes" within this context?

I assume, as Psycogeek suggests, that the default position on any proposed edit of this nature on a non-community wiki post is to reject or skip unless the person proposing the edit substantiates the validity of the change (or you can verify it), and that addendum is preferable to modification. And I agree with him on the importance of keeping visible on non-community wiki posts what are the original author's words and what are not when these kinds of edits are made.

Answer 1

This all being my opinion:

1) Don't put words in peoples mouths. If you have a better answer then create a better answer, create it in full using all available information, which can include SU. Half of everything is wrong on the web :-) citation to what? Is it testable, or provable? can you prove it yourself?
There is a comment feild and a downvote button, if you (this person editing) have large corrections or complete modifications of parts of an answer, then use the comments section and do tell what the problem is. People reading the answer can just as easily read the comment. On the side many of the best least known information comes from the obscure and different.

As you can tell I am not ok with it.
If the information was fully verified, I would still want to know that it was changed or ammended, added pieces to the answer (or even questions) Should be marked as such "edited By:" "Added By:" or whatever it takes so it is known that the information did not come from the person whos name is shown.

Even "edited:" or "added:" which are important headers to an altered question are REJECTED here at super user, they are needlessly edited out, along with thanks and other things that make us all human.
That information still exists if you want to view the history. IMO again this is just wrong, people going to (say) the Wikipedia do not weed through 50 pages of corrections and discussions, they just accept what is there. Here at SU the Q&A does not get enough people viewing it (like the wiki) to insure this slopped in added information is correct.

On the other hand . . . most of the editing that has been done on my own posts (after being reviewed and accepted) has been acceptable, and usually improves it. As the original questioner or answeree they can roll back any time they like, or they can re-edit. Reviewers are for the most part doing that job correctaly. (I have seen some that ehh not so much :-)

As a Reviewer you Must Be Very carefull with this crap added to a post as an edit, it has been 100% wrong before, it could be an opinion, or it could be malicious. Enough edits to a post and the original meaning goes out the door, and your left with . . . well the same crap written up on the rest of the web, which does not make it right (all the time).

2) Updates are great (yay!) still the information must be verified. Spammers will change links to suit thier web site, idiots will change things that look right and good, but do things like point to wrong web sites. Should the original information be kept? IMO yes again, not everyone uses the latest versions, I use software still created in 1996 :-)
If things have changed then it would even be ok to move the old junk to the bottom and put the new information at the top, Mark what has been done.
Again if the person has much better information, there is a box for that. The person can make comments, or they can make a whole new answer with all the new information in it.

3) No, you could "Skip" , you can skip anything you cannot verify. then some other person will run an accept on it, But it woult be Your Fault that the question or answer is destroyed by editing, with wrong informations. A person more qualified to verify , or who has tested it, or who will take the time to fully verify the information would be qualified to review it. (I have a story I could tell here about a judge unqualified to judge a case put in front of them, but there is not enough room).
Again, it could easily be a comment, or "If you want it done right Do It Yourself!" the editor could create a much better answer. Are they too freaking lazy to make a whole answer ? that I could understand.

4) If you have the rep to be a reviewer , then you have the knowlege (lol) to properly review the edits people are making. You know by now that there is a lot of crap, if you see it, "Reject" it takes 2 people to fully reject. If you do not know, then skip. (and some other person unqualified to know will accept it :-) Hey it wont be you, and that is all that counts.

5) Yes, I very often use the box where you get to make a comment on why you are rejecting an edit.

I have also moved "comment editing" to the comments box where it belongs. the system is not designed for a simple reviewer to do such a thing, so it works, and it doesn't work, but that is one possibility to put the data in a User3948 Said: comment. Best of both worlds, the information is shown, and the original persons work is not damaged.

Many times corrections , opinions, verifications, and adjustments to an answer are made in the comments. The original answeree then will see these comments , and make changes they want to make. The viewing public will also see all of that, and decide or verify for themselves. (at the least) Some drive-by anon changing the content who has zero rep, is the suspect :-) not the authority on the answer the person wrote. That is why it takes 2 reviewers to accept those changes. 2 reviewers who actually review it.

End Discussion.