6

Questions about supporting new users and my question about trivial questions raised an issue. Briefly stated, H:SE quality requires that we close questions quickly, but our record at revising and re-opening questions indicates that closure is a trapdoor function. It is easy to close a question, but hard to re-open.

IIRC, close and open require the same number of votes, but they don't require the same amount of effort. (can anyone substantiate that; I lack the skill to use the statistics to demonstrate what @sempaiscuba asserts. My gut says he is right, but statistics are better than my gut at estimating the size of the problem).

I believe that the site would be better if we were quick to close low quality unanswered questions and then to edit & re-open.

(IMHO, we should never cast a close vote on a question with a passable answer).

  • 1
    I copied and pasted the 59-page Review history of reopen votes into and Excel spreadsheet. I can produce some stats from that, and - superficially - it does seem to reflect my experience. One major problem is that it also identifies who voted and how. Another is that there doesn't seem to be an easy way to identify why people vote to leave closed (no reason is required). – sempaiscuba Oct 3 '17 at 13:50
  • Hash the usernames. If I knew the close reason, I think the data is useful. Are we more likely to reopen unclear questions, or trivial? (Example) – Mark C. Wallace Oct 3 '17 at 14:50
  • Taking just the most recent 1000 votes (which takes us back to April 2016, and so probably includes most active current users), 806 voted to leave closed, 178 to re-open, and 16 to edit and reopen. The close reason isn't shown on those pages (probably because a particular question can be closed for multiple reasons). Let me see if I can find a fairly quick & simple way of extracting that info and adding it to the data. – sempaiscuba Oct 3 '17 at 18:54
  • 1
    There doesn't seem to be a straightforward way to extract the close reason, and most closed questions seem to have more than one reason chosen. I don't think I'm going to have the time to extract that info & add it to the list of re-open votes. – sempaiscuba Oct 13 '17 at 10:12
  • 1
    Thank you for the effort. Alas, you've described the situation every time I use those tools - I feel like I'm on the verge of discovering something useful, only to find that I just can't get there from here. – Mark C. Wallace Oct 13 '17 at 10:18
  • 1
    I just did a little research, and we may not be doing as bad as it seems. If you consider this process the equivalent to an appeals court, our 17.8% reversal rate almost doubles that of the US court system – justCal Oct 13 '17 at 17:52
6

For me, one of the problems with the re-open queue is that it seems to be triggered by superficial edits. That is, a question gets closed because there's a problem, someone corrects a minor typo (which in no way alters the meaning of the question) and it appears in the re-open queue. Since the question hasn't been improved and the reason for closure still stands, my reaction is to vote to leave it closed.

As far as I've seen, when people make a real effort to improve their closed question, it's usually re-opened (although that might take a little while to get the required number of votes via review). However, in most cases where questions remain closed it's because the effort to improve them and bring them on-topic isn't made.

  • 1
    I'm not sure that it's the edit that triggers the queue. I think a question has to be manually nominated for re-opening. (Although you are right that the edits that precede a question being nominated are sometimes trivial). – sempaiscuba Oct 3 '17 at 13:31
  • @sempaiscuba Editing the body of a question within five days of closure will add it to the reopen review queue. – yannis Oct 4 '17 at 12:02
  • 1
    @sempaiscuba Actually, it's already there. Last sentence of the first paragraph. I completely missed it when I followed your link, I guess my eyes went straight to the bullet points. Took a search for "queue" in the article to find it. – yannis Oct 4 '17 at 12:19
  • @yannis My eyesight must be failing. I've visited that page more than half-a-dozen times now and hadn't spotted it! Chalk it up to old age. :) – sempaiscuba Oct 4 '17 at 12:41
4

What are some probable obstacles to re-open (factors that raise the effort of a re-open vote over that over close vote)?

My bias is to vote yes on re-open, but:

  • When I review the reopen queue, I can't tell if the question has been changed since it was closed. The edit history is a clue, but not always a good clue. I want to see why it was closed and what change has been made to justify a re-open. That requires effort. Is there a convention for how we can document the closure reason and the changes? Is there a way to make the work evident and demonstrate the relevance?

  • Spewing comment streams affect the effort needed to re-open. We should always avoid comment spewage and we should flag for obsolescence when the comments cease to be useful. But I confess than when I see more than 5 comments, my enthusiasm for re-open diminishes.

  • Some questions are ... difficult to salvage. I think examples would violate "be nice", but when the question is incoherent or prima facia trivial, I need more convincing to re-open.

1

Let me describe one question that was closed for being "too broad" that might support the idea that closure is a trapdoor function.

The question is Why did the United States want to “make Europe prosperous once more” with the Marshall Plan?

As originally posted, the question certainly did appear to reflect a certain bias, and this was noted in the comments. However, based on the OP's later comments in response to my answer, I realised that it was actually probably just a question of language, rather than bias. It is all-too-easy for someone whose first language isn't English to equate the terms "self-interest" and "selfishness". I edited the question accordingly.

I have been meaning to add a link to the appropriate volume of the Congressional Record, so that the OP (and any others who are interested) to read the official record of the debates on the Act in Congress (especially since, in the context of self-interest, it is illustrative to note how often senators mention specific benefits to their home states when discussing the Marshall Plan).

Because of other distractions (and the fact that the search tools on archive.org aren't really geared to multi-volume publications like the Congressional Record), it took me a while to find a publicly accessible link to the right volume. When I did eventually get around to it, the question had a re-open vote, so I thought that I would wait and see what happened.

Now, as the question currently stands, I would say that it is clearly "a specific problem with enough detail to identify an adequate answer" (as required in the help centre). Despite this, two-thirds of those who voted said that it should remain closed.

Now, it may be that they feel that the question should remain closed for other reasons. Those reasons may be entirely valid, but they are not recorded. In a case like this, how can an OP be expected to improve their question if they don't know why it was closed?

So yes, based on this, and my experience of trying to get a question that was incorrectly closed as a duplicate reopened, I'd say that closure does indeed seem to be a trapdoor function.

  • I do think you may have hit on something with the 'should remain closed for other reasons' issue. There have been times where questions received close votes for one reason, but definitely qualified to be closed under several others. One aspect may be corrected by edits, but the others remain. Maybe we need to change the close reason selection to a check all that apply type selection. I wonder if the new top bar review format will have an impact? – justCal Oct 13 '17 at 16:23
  • @user2448131 I certainly noticed that a lot of questions attract close votes for more than one reason (in fact most of those I looked at when trying to analyse the re-open data attracted votes for more than one reason). The reason that I chose to watch this particular question was partly because _all_the close votes gave the reason "too broad" on the original VtC. – sempaiscuba Oct 13 '17 at 16:33
  • Personally, and I added the wiki link to this question, I considered it trivial(wiki covers the extent of the question), and showed no research. That's why I added the link, hoping the user might get his answer before it got closed. The trivial nature is why I wouldn't reopen it. As mentioned, usually multiple reasons. We used to be fairly strict on the if it can be answered by a quick read of wikipedia, it doesn't belong on SE. – justCal Oct 13 '17 at 16:44
  • @user2448131 I thought that may have been the case for some. Interestingly, I agree with the OP that the reasons aren't that clear in the Wikipedia article (perhaps especially for those who don't have English as a first language). The direct benefits to the US are implicit, rather than being explicitly stated. There is a lot of good content on what Truman wanted, and what was achieved by the plan (leaving aside some of what was done with Marshall Aid by the UK gov't), but rather less on why Congress would support it in the first place. – sempaiscuba Oct 13 '17 at 16:57

You must log in to answer this question.

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