Duplicates and closing questions.

Question

This is in part a public service announcement and in part an invitation for discussion (see also the comments on Martin Tapankov's nomination).

First the announcement: everyone with at least 3000 rep can cast close votes on questions. In particular, if one thinks that a question is an exact duplicate, one can mark it as such by using the close button (it will then ask for a link to the duplicate question). This will not immediately close the question—it only closes when five people have cast close vote. It is not considered as an offense to the poster to cast a close vote (but you might want to add a comment in addition to the automatic one).

Now for the discussion: How aggressive should we be in closing posts as duplicates? I have noticed several times now that someone posts a comment pointing out a duplicate that gets upvoted several times, but (almost) nobody casts a close vote. Is this because people feel that we shouldn't close duplicates, or because they don't know that they have the power to do so?

Answer 1

Duplicates are often created by new users, naturally. Closing, which is a standard procedure, might feel harsh for users who yet don't know that site's practice of doing.

With regard to new users I would not close aggressively. It's nice to say "hey, welcome, have a look here and there, those answers might be a great help" perhaps additionally inform "and by the way, we usually close duplicate questions to keep the site tidy, don't be puzzled then" together with a vote to close. This way the user get's a warm welcome, we can close the duplicate soon.

Of course, if a duplicate happened to a well known regular user, it's perfectly fine to say "hey fellow TeX friend, it's a duplicate to this and that, let's close it".

The friendly atmosphere here is just more important to me than a well-oiled machine closing as quickly as possible.

Answer 2

Thanks for bringing this up; I've observed the same. In my opinion, posts that came up to 2 close votes should usually have been closed, but the problem is that for users with <10000 rep it is not always obvious that there are already close votes: It just says "(2)" after "close". If one of the first comments explains why the question should be closed as a duplicate, then this helps, and I hope that the question here helps some more.

EDIT 1: When I write "posts that came up to 2 close votes should usually have been closed", then I also mean that I don't think that many more than those should have been closed.

EDIT 2: Here's one of the problems (example post): There comes a question, not too hard to answer, and before you find the duplicates and post a comment, there's already an answer (or several answers). The more answers there are, the more the attention is distracted from the comment.

Answer 3

As I said in my nomination, I think we should get a bit more aggressive towards closing duplicates (not deleting them), and referring to single authoritative answers that are updated regularly. Now, since such answers could become quite long, some people might not want to read through them all, so a short CW answer (that the submitter can accept) and a link to the question should suffice. The submitter will hopefully take the time, if interested, to read the long version, and learn a new thing or two.

It's been the official policy to tolerate some amount of duplication, but I think this is mostly applicable to a monstrosity such as SO (we have around 500 times less questions than they do). For a smaller site like ours, people can usually read through most questions popping up in a day (at least the titles) without a problem, so pointing out duplicates (essentially, remembering that somebody asked the question earlier) is easier.

On your last question

Is this because people feel that we shouldn't close duplicates, or because they don't know that they have the power to do so?

I think there might be a third reason -- people wouldn't vote to close if they have already answered the question, or intend to do so, for the fear of losing (or not gaining enough) reputation. I'm not sure there's anything to do if that is indeed the case.

Answer 4

I agree entirely with Stefan, and I think maybe an example will illustrate.

Our top-voted question is LaTeX template for resume/curriculum vitae, and there are three question closed as being duplicates of it, all of which come above it in a search for "cv" ordered by relevance. The questions are:

1. Writing a CV in LaTeX
2. Are there any other packages defined to create a CV except the "moderncv" package?
3. Class for Curriculum Vitae

Looking over the askers of the first two questions, I'd say their overall experience was very similar, and while both got useful information from us: both asked their CV question, which was closed with no upvotes, asked sometime later another question which got one upvote and two useful but not welcoming answers, and then didn't interact with the site again.

I've given a late upvote to both these questions now, because I think that we should be generous upvoting any acceptable activity from very low reputation users, say up to reputation 50 (needed to leave comments). When closing such questions by new users, we should generally try to "compensate" them with an upvote to +1 and a friendly comment to the effect of, e.g., I think you'll find Herbert's answer useful, do ask another question if that's not quite what you were after (which Joseph did do with the first question).

Finally, I'll note that the first question asks for an example CV in Latex, which is not explicitly asked for in the "LaTeX template" question, and takes a bit of digging to find in the answers. Although the asker was happy for Joseph's comment, the question was not an exact duplicate, and an answer that pointed to the template question and either linked to an example CV or, better, pasted some Latex code in, would have made this site just that little bit better.