What is meta? ×
Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 133 Stack Exchange communities.

In my answer to Should we require minimum reputation to continue asking questions? I came up with a possible solution.

The essential problem is that some people come to SO, ask lots of lame questions, and never give anything back. The proposals have included only allowing a certain fraction of one's rep to come from questions and only allowing users to ask questions if they have a certain amount of rep, but neither of those seemed particularly popular.

This suggestion is:

Require that unqualified users have their questions reviewed like edits before being posted.

What is the qualification? It could be rep that's a certain multiple of their questions (say 10 rep per question) or a certain ratio of questions to answers, or some other criteria. We could also allow a certain number of questions to not require review (say, 5-10) so that a user would have to lose their ability to ask unreviewed questions rather than have to earn it.

Who reviews these questions? It could be 5k, 10k, 20k users, or maybe anybody above 2k who also has the ability to post questions without review. It could even be people with sufficient points in the applicable tags.

How would this mitigate the flood of bad questions? First of all, users would be rate-limited by only being able to have a certain number (1?) of questions in review at once. If a user keeps getting their questions rejected, they could lose the ability to submit for a period of time ("We are no longer accepting questions from this IP"). When two reviewers reject a question, nobody else will ever see it (maybe not even as a deleted question). Of course, there would be an "Improve" button so that reviewers could fix minor problems before anybody else ever had to read the problematic post.

Furthermore, upon clicking the "Reject" button, there could be a dialog box with a list of checkboxes to select reasons for rejection ("Off topic", "Spammy links", "Lack of formatting", "Not a question", "Missing sample code", "Too much sample code", "Poor English", etc.) and maybe even a box for remarks. This would allow the person asking the question to see why their question was rejected, fix it and resubmit (if applicable). This should train users how to write acceptable questions. The review page could have the question's preview remarks and deltas from the last submission, a link to the user's most recent rejections, or something in between.

What are the downsides? Unfortunately, this would increase the burden on reviewers. Hopefully the "training" aspect should reduce the ongoing burden of any particular user and not having the bad questions posted in the first place should eliminate edit reviews (i.e. review the question before it's posted so you don't have to review several edits to the question after it's posted). In other words, this is like unit testing -- it seems like it would add too much of a burden initially, but it actually ends up saving time and helps deliver a higher-quality product in the end.

Some users could be insulted by having their question criticized. This should be minimized by having standard rejection reasons that aren't personal, while the remarks would hopefully be a polite nudge in the right direction (e.g. "Just specify the OS you're using"). Of course some users would go away entirely after being confronted by this, but it's not clear that those are users we want.

share|improve this question
    
I seriously doubt if the gains weight up to the burden of this formal approach. And with burden I refer to the downsides you mention yourself. –  Rogier Apr 7 '11 at 19:37
    
@Rogier I don't see any other solution than investing even more personal effort to educate users. I don't know of any automated system that can teach someone how to behave or how to communicate. –  Octavian Damiean Apr 7 '11 at 19:40
    
@Octavian Damiean I'm not saying that investing more personal effort is not a good idea per se. But does it have to be this formal? Furthermore, the fact that a rejection by (only) two reviewers could prevent a question from being visible at all feels a bit like 'government censorship'. –  Rogier Apr 7 '11 at 19:58
    
@Rogier I agree that this idea could be refined a bit but it doesn't sound bad to me as a start. –  Octavian Damiean Apr 7 '11 at 20:01
    
@Octavian Damiean All effort to increase SE/SO quality, like the above question, is to be respected. –  Rogier Apr 7 '11 at 20:07
5  
This is an awful, awful job. Wouldn't wish it on anybody. –  Uphill Luge Apr 7 '11 at 20:28
2  
@Hans: I agree. However, the downsides would be somewhat mitigated by the rule being that you can lose the ability to answer questions, rather than having to earn it. I can't see doing it any other way, in fact. As proposed, a user gets a number of questions "free" (though I'd say more like 25), and as long as those questions are mostly not closed and downvoted, and as long as the user has some modest rep gain, e never need be subject to the review process. The tricky bit, as always, is where to set the threshold. –  Josh Caswell Apr 7 '11 at 20:50
    
@Rogier: To me this seems a bit like unit testing. This proposal would stop bad questions before they can become a problem, which in theory should reduce the overall amount of work and end up with a better final product. –  Gabe Apr 8 '11 at 21:05
1  
Sounds like it just makes more work for the people who are already burdened with answering (or editing, or closing) these poor quality questions. Not sure who this is really helping... –  Cody Gray Apr 14 '11 at 7:59
    
@Cody: Let's say somebody posts a question with unformatted code. The idea is that rather than having to spend time editing a post, I could just click "Reject" (or "Fix your formatting" and "Reject") and let the OP fix it themselves. The OP learns how to format code (or never posts again), relieving the burden of everybody having to fix their crap, and everybody reading the site isn't burdened with their illegible post. –  Gabe Apr 14 '11 at 8:09

1 Answer 1

I don't really see the point of this.

Why not "review" questions as they are asked?

http://stackoverflow.com/review

The existing flagging and downvoting mechanisms should suffice without adding Yet Another Queue of Work for more of our users to have to deal with and think about.

Also, questions which fail to get any significant activity are already auto-culled over time, and downvoted questions with no significant activity auto-culled even more aggressively still.

share|improve this answer
    
I can't tell if your "why not" question is rhetorical or if you're actually asking it. –  Gabe Apr 14 '11 at 7:57
    
I would want to believe he is actually asking because the review section works. Maybe it is hard to see it on SO due to the sheer volume, but I am sure if you ask mods on the rest of SE network ... it works. –  phwd Apr 14 '11 at 8:16
    
@phwd: Ah, I see. After the link was added it makes sense. The difference is that the "review" page doesn't have a "reject" button, so the person posting the crappy question doesn't have any impetus to learn. –  Gabe Apr 14 '11 at 8:43

You must log in to answer this question.

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