8

Review Low Quality Posts has just three buttons: Looks Good, Edit, Recommend Deletion.

I often see one-liner answers, sometimes link-only answers. Most of the times, such answers can be improved by the author and hopefully turned into great questions. However, what should a reviewer do?

  • Looks Good does not seem appropriate, as the answer doesn't look good.

  • Edit is not an option, because editing such answers would mean totally rewriting them. Also, in a few cases, the proposed solutions aren't valid solutions and I wouldn't know how to edit an invalid answer.

  • Recommend Deletion does not seem appropriate too, because there is room for improvement. Personally, I wouldn't remove even invalid answers:

    1. I may be wrong and the solution may be correct;
    2. I feel leaving a comment asking for more details is better both for clarity and to give the author an opportunity to learn;
    3. only the OP has the right to choose whether an answer contains a valid solution or not. He/she should be free to test and report back.
    4. AskUbuntu (as any other SE site) has a great voting system that lets the community choose which answers are helpful.

What I do in such cases is leaving a comment and clicking Looks Good. Rarely, when I strongly feel the answer has nothing to do with the question, I choose Recommend Deletion.

However I don't think this is a good practice, so I'm here to ask for suggestion. Personally, I think there's need for an another button that lets you leave a stock comment and go ahead.

2 Answers 2

9

"Looks Good" Means Good Enough to Exist

This is just my opinion, of course. (In particular, I am not one of the people who designed the review system.)

I disagree with the idea that the Low Quality Posts review is supposed to produce a whole bunch of great posts, plus a whole bunch of deleted posts. The purpose of the Low Quality Posts review is to delete posts that should be deleted, and leave the rest. More precisely, the purpose of this system is to determine when there is substantial agreement that a post should be deleted for quality reasons, and put that consensus into action.

  • Of course, editing is also part of the purpose of this system. But when you decide a post will not benefit from editing, there is rarely any good reason to put off deciding whether or not it should be deleted. (See Exceptional Edge Cases below for an explanation for why I believe this is rare.)

If a post should be deleted, use Recommend Deletion.

If it should be edited, such that after editing it shouldn't be deleted, use Edit.

If you don't know what to do because you don't have enough information, or because you wish to defer to the judgment of others, use Skip.

Otherwise, you think the post should remain. So click Looks Good.

It's OK to Click Skip, But it's Like You Didn't Review the Post

You should always feel free to click Skip. In review, never be afraid to defer to others when you think you're poorly suited to the particular review item in front of you.

But that's what Skip means. You're saying to the system, "Make it like I never reviewed this."

Skip doesn't say anything to the community. It's a way of remaining silent.

If you think a post isn't perfect but shouldn't be deleted, then you're not unsure. You're not of the position that others will do the job better than you. Posts that shouldn't be deleted should be marked Looks Good, making way for the posts that actually should be deleted to be seen and deleted.

Other Actions

Of course, you should not feel limited by the options available to you in review. You can always click to see the post in its "natural habitat," where you can perform a variety of actions, including some that can be done from the LQP review queue and many that cannot:

  • Upvote and downvote.
  • Flag as NAA or VLQ, custom flag, flag as spam or offensive.
  • Edit, without affecting your review history (occasionally it may make sense to edit a post in a small way and vote for deletion, for example, if an answer is really a new question, and editing helps clarify this by formatting code more properly).
  • Comment, upvote comments, and flag comments (also possible from review).

Going "above and beyond the call of duty" of review is often (though not always) a good thing to do. But this doesn't fundamentally change how the review queue works.

The Categorical Imperative (Sort Of)

Here's an additional, conceptual explanation for why I advise clicking Looks Good for any post you believe should not be deleted. (Unless you're editing it, of course, which is essentially a variant on Looks Good. The system assumes you approve of your own edited version.)

Not everyone uses review exactly the same way, and that diversity is a great asset. But let's consider what happens as the percentage of users who are reluctant to click Looks Good becomes large.

In that case, most reviewers are choosing between two alternatives:

  1. Recommend deletion (or delete, if they have enough reputation).
  2. Hand the review item off to someone else, to either delete or hand off to someone else.

For a post that shouldn't be deleted, this has three possible outcomes:

  1. It is deleted even though it shouldn't be, because a low risk of mistaken deletion is multiplied by a very large number of reviewers who are unwilling to take the action that mitigates this risk.
  2. It is reviewed by a very large number of people, until someone like me who is more willing to click Looks Good does so. This just makes lots of people do a little bit of work before achieving the exact same result as if you clicked Looks Good yourself as soon as you decided it shouldn't be deleted.
  3. It is reviewed by a very large number of people, and everybody clicks Skip. This combines the effect of nobody reviewing it with the effort of everybody reviewing it.

Exceptional Edge Cases

On rare occasion, I agree with FEichinger's way.

If a post is highly volatile, such that you believe it is about to change extensively, it may make sense to click Skip out of the belief that nobody knows how to review it (but that people will know how soon).

However, this is rare. In particular:

  • If the post almost makes sense and can be improved, click Looks Good or--even better--click Edit. Remember that even if you don't feel comfortable making all the changes that might improve a post, but you can significantly improve it by applying some of them, you should do so.

    Whether the author improves it or not, a post that is fundamentally good can be improved by almost anyone.

  • If the post does not make sense, vote Recommend Deletion. The author can always post a totally different, better version separately (or request a moderator to undelete an edited version).

  • If a post is high quality but wrong, it should not be deleted unless it is malicious or unusually dangerous. Comment, downvote, and click Looks Good. This is the "low quality" review queue, not a "not awesome yet" review queue.

  • Link-only answers are occasionally expanded by their authors, but this is actually pretty uncommon. (What's somewhat more common is for them to update them when the link breaks, if they're still active on the site and someone comments about it.) If you think it's valuable as it is, click Looks Good. If you think it should be deleted--and you may want to lean toward this, especially if there are other good answers that cover the same ground--click Recommend Deletion.

That's an incomplete list of normal scenarios that are not edge cases, where you don't have to click Skip.

I would give an example of a volatile post where I think everyone should (for a short while) click Skip, but:

  • I cannot recall any specific one, because this is not all that common.
  • The post wouldn't exist (in that form) by the time most people read this.

Posts should not be made to grow stale in the review queue with the hope that someday, someone will have perfect moral clarity. If a post circulates in the queue for an extended time and shouldn't yet be deleted, click Looks Good.

Remember, the LQP review queue isn't the only way to get posts deleted, if it turns out that's the best course.

1

Looks Good

Don't do that. As you said, it doesn't look good, it is not sufficient, it is not elaborate enough. Looks Good should be used for posts that, well, look good. Posts that are fine without an edit, are on-topic and don't need any other action taken.

Edit

This is exactly what you should do. If the post is salvagable through an edit, edit it. There may well be posts that have to be rewritten entirely - go ahead. As long as the meaning of the post remains the same, go for it. This is the very core functionality of our edit system - edit to improve.

Recommend Deletion

This should be used for posts that don't look salvagable. They're off-topic, or they consist of information that was already stated on the Q&A, for example. Posts that are entirely unsalvagable, because they are simply insufficient to the core, these should be deleted.

Skip / Not Sure

This is what you should use if you are not sure how to deal with it, or the result of your action will not lead to a conclusion to it. Leaving a stock comment does not give a conclusion, it gives a chance of a conclusion. Perhaps the user acts on it, perhaps not.

In most other cases, a stock comment is accompanied by a close vote or a flag, which initiate the community's decision - this is not the case here. As such, I disagree with giving a "Done."-ish action for the Low Quality Posts. Other than Late Answers or First Posts, which focus more (obviously not solely, however) on it being on-topic and appropriate, Low Quality Posts focuses directly on the quality.

The result of a run in the LQP should be either a deletion, or a great post.

You must log in to answer this question.

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