While this may sound like a duplicate to this question: Is there a way to improve the quality of edit reviewing? the answer to that question did not reply to the fundamental problem. Nor did I find another appropriate answer.

A short situation description: I am working on an electric problem I have in my household. I found a relevant problem solved here: How to throttle power consumption of a heating coil? While calculating my numbers according to the solution provided in the answer, I found out, that the answer is missing some brackets, that are fundamental to the correct solution.

I had no reputation on that Stack Exchange site, so I could not comment on that problem. While people suggest to quickly get 50 reputation points and then comment on the post, I am not a professional in electronics nor do I plan to become one. So I am not going to gather 50 rep points to correct the error. Still I would like to save other users from spending time looking for the error in the calculation. So I posted an edit. The edit got rejected, as the person stated that adding the brackets changes the result and the formula itself. Yes, the brackets changed the formula, to the correct one.

I think a reviewer should pay more attention to the edit. Here it seems that a quick reject is the easiest way. I posted the edit once more and then it got accepted as it was correct for obvious reasons.

I wonder how many errors are there on the many Stack Exchange sites, which someone tried to correct, got rejected and left it the way it is. Does someone who wrongly rejects an edit get some reputation reduction? I think he should, so he would be encouraged to make quality reviews.

  • Actually I'd dare say that's paying more attention than the average (robo-)reviewer. -__- The reviewer doesn't necessarily have expertise in the matter in question. – M.A.R. Feb 19 '16 at 12:59
  • I can see that you got another edit approved. So what's the fuss about? – M.A.R. Feb 19 '16 at 13:05
  • @IͶΔ Thanks. I don't know what you don't see, but CL has approved my edit and afterwards edited himself my explanation to my edit away. Approved 1 hour ago: CL. reviewed this 1 hour ago: Edit Community♦ reviewed this 1 hour ago: Approve (more) – AndyZ Feb 19 '16 at 13:20
  • @IͶΔ I added the source for the correct formula to my edit: C = 1/(2pi f Xc) = 1/(6.28 * 50Hz * 18.3R )= 174 microfarads . (The forumla for capacitive reactance is discussed here: en.wikipedia.org/wiki/Electrical_reactance) but the source was then edited away...that's ok for me. – AndyZ Feb 19 '16 at 13:21
  • Well, you should discuss that in their meta; it's not very clear to me why he would do that, but the link might have seemed redundant. – M.A.R. Feb 19 '16 at 13:32
  • @IͶΔ The problem is not the link, the problem is that two people peer reviewed and rejected a correct solution while thinking that the correct solution is in the original post. Such bad reviews should be punished :-) Never mind, I'm just angry that I have to spend time to proove my edit, when it was obvious that it is correct. So the question here, how does SE support the idea of quality reviews. – AndyZ Feb 19 '16 at 13:43
  • There is some point where you have to trust other people's judgment instead of relying on an algorithm. – Deer Hunter Feb 19 '16 at 14:15
  • @DeerHunter Thank you, I thought about a conceptual algorithm not a software one – AndyZ Feb 19 '16 at 14:20

Well, this is my take on it.

The review queues were up for the most dedicated users to just "help". Ultimately there shouldn't have been any incentives for reviewing stuff – not even the typical gamification incentives such as rep points – other than the sense of responsibility and caring for the community.

Reviews are needed a lot though, so we should've had some badges for this; so we'd give people more reasons to watch random often problematic posts for seemingly no benefit. Three badges were introduced for each queue, and we have some other ones such as proofreader that are also earned by reviewing.

Apparently for some people the badges became the end and the reviewing became the means, while it was intended to be vice versa. So came robo-reviewers — the problem of "robotic" reviewing has been noticed long ago.

To counter, right now most of the higher-traffic SE sites where the problematic behavior is more destructive have audit systems enabled. Furthermore, moderators or too many failed audits can punish the offending user with a review ban. The problem persists and hasn't been vanished yet, but if you have a better idea that has a very very low chance of getting people with good intentions into trouble 1, we're all ears.

That said, I don't think the action taken on your previous edit was robo-y. People can and do err and it's with errors that life goes on. You're making it seem like a big deal while it's not. Up until this time of the day, 129 review tasks has been done by EE.SE guys. If only two of them are erroneous, that's a pretty good review team they get.

1: After all, the best users the community can offer are the ones that pay the review queues a visit.

  • 1
    Thanks a lot, this gave me some insight in the topic behind my question. So, I will learn to relax :) , thank you for your effort. – AndyZ Feb 19 '16 at 14:49

Suggested edits are not for correcting errors in answers. They are for correcting spelling and grammar mistakes, improving formatting, and so on. This means the reviewers were correct to reject your suggested edit. Since suggested edits are also shown to the author, sometimes corrections reach the author's attention this way and get made, but it's not the preferred mechanism.

What is the preferred mechanism when there is an error in an answer?

  • comment. As you have said, you don't have the rep and don't intend to earn the rep to be able to comment
  • add your own answer based on the flawed one, with the error corrected. (Should the other answer get fixed, you could always delete yours.)
  • breathe deeply and hope that others notice the error. After all, you did.

Suggesting an edit is a bad approach to the problem of a technical error in an answer. Reviewers who reject these suggestions are doing as they have been asked to do. Fixing errors of fact is not the role of the suggested edit.

On the title question, is there a way to make reviewers better, yes there is, with audits and automatic bans for those who fail them. It's working a little bit, but could be better. Generally the issue with suggested-edit reviews is over-approving, because it takes only one click to approve, but several clicks to reject (you have to choose a reason) so those who are just grinding for a badge click Approve, Approve, Approve all day. There is not currently a perceived problem with over-rejecting. Your rejection may sting, but it was in fact the correct action. Even if it had been the incorrect action, the SE network as a whole does not consider over-rejecting to be an issue at the moment.

  • Thank you for your reply. I think, the edit function should become avaliable only at the same time as the comment function. Then I would see that there is nothing I can do about an error before I get 50 reputation points and I would leave it there. Else I spend time correcting an error in a wrong way - result is waste of my time and the time of others. – AndyZ Feb 19 '16 at 15:20
  • No @Andy; there's a good reason people gain the commenting privilege much later. – M.A.R. Feb 19 '16 at 15:27
  • @IͶΔ Exactly, that's why new users should get the edit privilege (edit posts from other users) also only after getting 50 reputation points. – AndyZ Feb 19 '16 at 16:33

You must log in to answer this question.

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