# Do we need a 'reject and improve' button?

Have a look at this simple example: http://stackoverflow.com/posts/5349764/revisions. There is a approved suggested edit, and an improvement was added on by a 10k user. The interesting thing is that the original edit should obviously be rejected - fixing one typo by introducing another is... not right.

However, its a small mistake, so the 10k user in question simply went for the path of least resistance, and clicked on the 'Improve' button. The problem here is that the suggested edit is implicitly accepted when you 'improve' a post, netting the editor 2 rep. Now the rep gain may be a token sum, but it does send the wrong message doesn't it?

Would it be useful to have a 'Reject and Improve' button for these sort of scenarios? Or would it be too rare to be useful?

-
Being able to explain in a short sentence why it is rejected would help the editor to understand how he could improve it. I accept/reject a lot (on P.SE) and sometimes I face situation in which I want to reject but don't do it because the editor wouldn't understand without a comment and therefore loose motivation to edit posts. –  user150926 Mar 18 '11 at 12:41
I wasn't aware that 'Improve' implicitly accepted a suggested edit. I have been treating it as a sort of "The question needs editing, but this edit sucks, screw it I'll fix it myself" button. –  Blorgbeard Mar 18 '11 at 12:56
Thanks for this question. I used to think that the site policy was to accept any improvement, no matter how small, but I can't find the post that made me think that. I've accepted several such micro-edits, all the while thinking to myself "this is silly, why did the editor fix that one blatant typo but not those five others?" –  Pops Apr 1 '11 at 15:18
+1 @popular precisely, and I think a lot of them are users thinking they need to actively remove salutations and hi's from posts! (Forgetting other more important things) –  gideon Apr 1 '11 at 15:55
I thought I could reject a weak suggestion and (through some URL twiddling) jump in and make a more substantial edit. Well, it turned out, I both rejected and implicitly accepted it as it was counted as an improvement. Lame... –  Jeff Mercado Jun 5 '11 at 22:03
Are we convinced it will change the behavior though? –  Adel Oct 23 '11 at 19:35
@Adel Yes, it will teach. And if not, it would be an improvement for the site and reviewers anyway. –  NGLN Oct 23 '11 at 21:10
Hmm, That's a very good point. I'm signed up then, Thank You So Much –  Adel Oct 23 '11 at 21:12
Here's a great example of this happening that I just ran into. –  animuson Dec 30 '11 at 20:43
Here's a proposal: provide an interface for the high-rep user to merge proposed edits (and possibly new ones), and allocate some token amount of reputation credit as s/he sees fit. –  Karl Knechtel Jan 3 '12 at 4:34

I changed the workflow for improve.

The edit screen now has an option to "deny" crappy editors the rep bounce when they suggest crappy edits and you improve it. The box is selected by default.

I do not want to introduce complex UIs on the improve "scenario", so you get no custom message for the decline. Community declines it, and that is it. The end user can learn how to edit properly by looking at the post.

Looking at the real data: http://stackoverflow.com/tools/suggested-edits?tab=improved&filter=day (10k only) I am struggling to find instances where the original suggestion was not helpful is some way.

On the other hand, it may empower users to improve posts that would have been left to rot.

-
Most cases would have something helpful, they just leave a lot left undone. –  Lance Roberts Oct 20 '11 at 4:56
So they helped, why deprive them of the 2 measly rep ... –  waffles Oct 20 '11 at 5:57
Well, I haven't been, but I think them making one edit, and leaving a bunch undone kinda sucks. They don't learn what they're doing wrong, so they will just keep making partial edits. –  Lance Roberts Oct 20 '11 at 5:59
There are lots of sub-6-character edits which are helpful in some way... :P –  Benjol Oct 24 '11 at 13:03
If I put a bounty of 500, will you fix it? –  BalusC Nov 10 '11 at 0:10
This doesn't address straight-up bad edits. It's often more work fixing or reverting those than the original post. Hitting "improve" sticks you with Mr.Editor's version rather than OP's. This is not common, but definitely exists - 15 minutes spent reviewing edits will turn up at least one. –  Wesley Murch Nov 22 '11 at 18:37
There is another annoying case: where you hit 'Improve', then the half-assed edit gets accepted, and you have to do your (better) edit again. I'm FOR Reject & Improve. –  Benjol Dec 22 '11 at 7:55
A checkbox saying 'suggested edit was hardly helpful' that would give the original suggester only +1 rep? That would send the message 'Thanks, you helped, but you didn't do too good a job at it, try a bit harder next time.' Sure, +1/+2 doesn't make much of a difference, but I would appreciate being able to send such a message. –  Daniel Fischer Jan 20 '12 at 11:58
My rejection rate on SO is about 1/4. Most rejects are to posts that would warrant editing. So I would use “reject and edit” in about 10% of the cases (say 15% if I felt up to/had the time to do the edit myself every time). That's not “ridiculously low”. –  Gilles Jan 20 '12 at 18:02
@Chichiray it is complete –  waffles Jan 30 '12 at 5:21
This is utterly fantastic! It's about time this finally got implemented. But I think you misspelled "helpful"... :-) –  Cody Gray Jan 30 '12 at 5:50
@CODY NNOOOOOOOO –  waffles Jan 30 '12 at 5:56
@waffles: thanks! This will gratly halp me making butter edits! –  Mat Jan 30 '12 at 6:26
@Madmartigan It's fairly easy to copy the original post prior to hitting Improve. But good point, otherwise I hadn't thought of this. –  NGLN Feb 1 '12 at 22:14

Just wanted to add another voice in full support of this idea.

I've recently hit 10k on SO, and often spend down time reviewing the edit queue. It's extremely common to see trivial, weak, or terrible edit suggestions, but the post itself does need an edit badly.

What are my options?

• Hit "Reject" and get locked out of editing the post, even though it sorely needs an edit and I am more than happy to do it myself. The idea that I'm going to monitor the post and wait for another user to reject the first edit, so I can get in there and clean up myself, is ridiculous.

• Ignore the suggested edit.

• Hit the "Improve" button, do 95% of the work myself, and give the user who suggested the original "bad" edit +2 rep, which also sends the message that their edit was a good one - so keep going.

The "bad" edit suggestion can still be useful in a sense: It brought a post to my attention that needs editing (although a proper edit suggestion would be much better of course).

There's plenty of room for it in the UI, it's function will be self explanatory, and it will clear up the edit queue a lot faster and get sloppy posts the edits they need.

I've come back to visit this post many times in the past few months to see if there has been any progress. I think we really need a "Reject and Improve" button.

-
+1 For your first point. Indeed, when an edit is proposed, we cannot modify until it has 2 votes; ridiculous. No one will wait for the second (or third) vote. –  NGLN Oct 19 '11 at 19:37

As I wrote in my recently closed duplicate question I think the lack of a Reject and Improve button hinders good edits, while promoting accepting pointless ones (as described by @random people tend to accept edits more often than rejecting them). I don't think the use case is rare, I run into this problem almost daily.

There should be a better way around a poor edit suggestion than accepting it (because honestly, in most cases it will not be rejected even if you vote for it to be), as awarding rep for poor behaviour will likely only cause the offender to repeat it.

-

People need to harden up and hit that Reject button more.

If they added a "Reject and Improve" button you would still get throngs who just feel so sorry and pitiful for people trying their weakest to improve a question that they'll ignore that button and use Improve anyway.

Hitting Improve when you should be hitting Reject is definitely sending the wrong message. Token reputation gain or not.

People like showing others that they give plenty of backrubs. They don't want to be seen in anything close to a light of negativity or a downer.

-
I don't think people aren't rejecting because they're soft, I think they're not rejecting because improving actually rejects the edit faster. I can cast a reject vote and watch the question be wrong for the next 10 minutes, or I can improve it now and make it right. That's not being soft. It's a pretty unfounded observation to make anyway, given the efficiency with which we close bad newbie questions. –  Mark Peters Mar 18 '11 at 13:37
Plus someone else is likely to hit that Approve button @mar –  random Mar 18 '11 at 13:38
Regarding weak accepts - Jeff has indicated in a comment that he may add guidelines to the page to encourage better decisions. –  ire_and_curses Mar 18 '11 at 15:23
I'm getting so tired of seeing all the poor, crappy edits, but don't really want to reject and leave the post undedited. It's and OCD's worst nightmare. –  Lance Roberts Oct 15 '11 at 5:14

Yes, we highly need it!

It's even worse: if I'm already editing a wrongly formatted question, I don't get the orange bar “this question is already edited by someone else”, no, I am not allowed to send my edit because there's a pending edit waiting for approval.

This wouldn't be a problem if all those edits were useful, but I encounter lots of minor edits that don't fix formatting issues and the like.

With the orange bar message, I could save my edit, look at the previous edit(s) and rollback to (one of) theirs if it was more useful. Now I have to discard my edits (or copy all input fields manually, which is also quite cumbersome) and go back to the question only to see that there was an unnecessary edit waiting. Now, if I click ‘Improve’, I'll have to go all over the question again.

Obviously, this is a rather frustrating process to me.

-

+1 For this question which I was about to duplicate a few minutes ago (with thanks to the new question similarity filter).

Yesterday I reviewed a couple of revisions which had good grammar, case, salutation and thanks edits, but the proposer strangely enough also added the edit summary to the post body... Well, that needs no encouragement, so I had to reject.

One great advantage of a Reject & Improve button would be that possibly good parts of the proposed edit remain intact so you don't have to rewrite those parts, as you would if you solely reject.

And secondly, if that rejection is the final closing rejection which deletes that edit from the revision queue, and if I want to make the edit myself again, then... "Oh, which post was it again?"

Now, as this question from march is still open, and the votes and reps of the answerers speak for themselves, me wonders why the button is not added yet. If the multiplication of priority index and implementation hardness prevents it from coming up in the top of the todo list, then I suggest to (temporarily) implicitly reject (instead of implicitly accept) the proposed edit when the Improve button is chosen. That would slightly make more sense.

-

I've been doing a lot of suggested edits, and it really sucks when an editor fixes one part and leaves a bunch of crap in the post. I end up having to do an Improve, since if I Reject it, then it's still not fixed, and that doesn't really solve anything. Those poor editors shouldn't get 2 rep for doing such a poor job.

Please give us a Reject & Improve button.

-
basically you are asking for one-click reject ... something that was declined already a few times. Improve is already a back-door... reject and improve is a bigger one. –  waffles Oct 19 '11 at 22:52
No, you missed my point. If I one-click reject it, then nothing gets fixed. Many times they have some good edits, but may have some bad ones, or just left crap around, and still need a lot more help, and I don't want to have to redo everything. Right now I'm just improving it, which gets my job done, but still gives him 2 points (which I'd rather he had, than not get the post edited). They also don't learn if they don't get the reject. –  Lance Roberts Oct 20 '11 at 4:54
The key point for me is the learning - certain editors are so close to being very helpful, 90% of their edits are good but that last 10% would be so easy to improve. Sadly I think those editors (who seem to be very prolific!) get approved more than rejects so don't know about that 10% improvement they could make. –  David Hall Jun 17 '12 at 10:44

I whole-heartedly agree having came across this today, normally I don't bother worrying about the implicit approve, as the edits are tiny anyway

however in this post, it already had one approve and it would have been a lot easier for me to hit the improve and implicitly accept the title correction, and carry on straight away - but I decided to be good and follow the "correct way" and rejected rather than improve.

I then had to wait for a second approve or reject before I could do the rest of the edits I thought were right - as it was the second vote was an approve so I feel I wasted my time completely ;).. well not quite, I still did what I thought need doing, but the point is a "reject and improve" would have saved me a wee bit of time! and might have taught the quick edit, rep collectors, something?

-

Hmm I'd just rather have "Reject" and "Accept" because both Reject and Improve or Improve(as currently implemented) is just syntactic sugar for editing the post yourself.

1. In the case of Improve its like Accepting(+2) and then clicking edit on the post yourself.
2. For Reject and Improve again as good as clicking Reject and clicking Edit itself.

Its just a shortcut? No?

One thing IMO, I dislike is, half-a** edits. I've seen some lower rep users add edits like "removed salutation" without correcting spelling/tone/ or formatting the code. There should be a way to notify/prevent users from doing this.

+1 for Pierres suggestion on adding a line to tell the user telling him/her what to fix, this way if they're edit is rejected, they will know that the same kind of edit will be rejected again. (They'll know what they're doing wrong and improve if they want to or just NOT suggest edits)

-
The problem is that hitting "Reject" will not reject the edit, it will simply add a vote for rejecting the edit, which means that you'll need to wait for others to express the same opinion before you actually can get around to editing the post yourself. The reject and improve-button would mean instant rejection, and a new suggested edit in place of the rejected one. –  Mia Clarke Apr 1 '11 at 16:25
oh! lol Didn't realize that, I've always just tried accepting and if I see something bad I accept and improve because there like I said, there is no way to tell the user what he/she is doing wrong. It is annoying that rejects also require peer review, I mean I'm sure most 2k+ users will not reject a suggested edit unless we really have to, heck that it what we are doing right now. –  gideon Apr 2 '11 at 9:45
It is kind of frustrating that I could have edited a post directly, with no voting, if some bozo hadn't come along before me and edited it badly. yes. –  Wooble May 23 '11 at 19:08

I think Improve and Reject Button is for taking second opinion. It must be there as any Approval or Rejection must belong more than one person's decision. Any single user can not decide what is to be Approve or what is to be Reject. So as per my Answer There Must be Approve and Reject button in case of Editing.

-