10

This is a narrowly-scoped question; I'm just curious if this has been investigated quantitatively.

Background only:

I spent some substantial time and effort and love and coffee writing this answer to a question that had some close votes and was subsequently closed of an older and valid duplicate.

That all that happened is all okay with me generally; generating posts is of itself already rewarding, I'm not in it for the rep. But...

Begin actual question:

I've sometimes felt a pang of concern when a question with a good answer is closed as duplicate, my concern has been that fewer future eyes will reach that good answer due to the closed status. I currently have no basis for that concern, it is just a hunch.

Since Stack Exchange has it's data exploring SEDE tool and there are potentially outside ways to count traffic, hits, etc., I'm curious about the following:

Question: Have there been analyses to see if views of well-received answers are reduced by closure as duplicate where the analysis tries to control for other factors?

I am no expert in data analysis other than I know that it can be really hard to control for other factors, but someone skilled in the art may have tried.

Please: It's not necessary to post an answer explaining why this is hard or impossible therefore there couldn't ever have been any QED. I'm scoping this question to focus only on what has been tried. Thanks!


Advice from 2012:

14
  • 1
    Unfortunately, for some reason, the earlier question, despite having almost the same wording and tags, does not come up as a potential duplicate when asking the question that was closed. I have no idea why.
    – ProfRob
    Jun 26, 2021 at 10:05
  • 10
    Would your answer add value to the question that is the duplicate target over and above the existing answers? If so why not delete your answer from where it is now and then write an answer to the original question? Jun 26, 2021 at 10:08
  • 6
    @RobertLongson a better alternative is raising a costum mod flag and ask for the answer to be merged with the duplicate target. This does have a disadvantage that the original answer may need editing to fit under the new question.
    – bad_coder
    Jun 26, 2021 at 10:23
  • 2
    Worth noting that views in SEDE aren't tracked for answers specifically, just for questions, which might make this kind of analysis more difficult.
    – zcoop98
    Jun 28, 2021 at 16:14
  • If others feel your answer is a better answer then the answer at the duplicate target. People could ask a mod, or gold badge holder, to reverse the duplicate closure. Duplicate closure is intended to guide people to the best Q/A pair available. Asking to do things on a question you have an answer on is a bit tricky, as people could perceive you as biased in such case.
    – Luuklag
    Jul 12, 2021 at 7:07
  • 2
    @bad_coder I don’t think mods have the ability to just move an answer; they have to merge the entire question which defeats part of the purpose of closing questions as duplicates (making the main question more discoverable).
    – ColleenV
    Jul 12, 2021 at 10:01
  • 1) what "other factors"? 2) I second the comment suggesting to flag for merging. of course, please follow any site procedure, such as meta.stackoverflow.com/a/400200/11107541 on SO. 3) as others have stated, SEDE does not track views on answers, nor does the system, and it doesn't track those views' history- just the current all-time total.
    – starball
    Apr 2 at 0:07
  • The correct solutions for the situation of the answer and question you've used as an example have all been mentioned here: 1) You, as the answer author, move your answer to the dup-target question by copy & pasting the Markdown, then edit it to fit the dup-target question and delete your original answer from the duplicate question (you'll need to flag to ask for deletion, as you won't be able to delete your answer, due to it being accepted). 2) Flag requesting that the questions be merged (first, edit your answer so it fits the dup-target). 3) Flag to have the duplicate direction reversed.
    – Makyen
    Apr 2 at 1:18
  • Your request here is, at best, quite difficult. You're looking for "reduced views" of "well-received answers" as a result of duplicate closure. In order to find that, one must first find a baseline for the number of views of the question prior to duplicate closure, but duplicate closure is, usually, done very soon after the question is posted (within 7 hours for your example question). The number of views a question gets varies dramatically over time, so such a short period prior to closure makes comparison invalid.
    – Makyen
    Apr 2 at 1:38
  • Comparisons of views between questions is also largely invalid, as that also vary dramatically, based on the question's topic and largely on the searchability of the different questions, as the vast majority of traffic is driven from search engines (also hot network question, or not). Keep in mind that both the dup and dup-target are edited over time, often specifically with the intent to increase searchability, so you'd need to track when those edits took place and the variance in views after each edit (of both the question and all answers, not to mention new answers on the duplicate-target).
    – Makyen
    Apr 2 at 1:38
  • Overall, I doubt that what you're asking for here will result in reasonable or valid data, due to the nature of what you're wanting to look at. While the duplicate-target typically gets more views, there are many examples where the duplicate question gets more views. So, any analysis would need to account for such cases too.
    – Makyen
    Apr 2 at 1:39
  • @Makyen I have not asked someone to do such a study, I'm asking if anyone has already done one. That said, one might isolate questions that were closed after say a month or two, and look for precipitous drops in views or answer upvotes. If a precipitous drop after closure is indeed clear, then maybe we can start to think about what that means to a site which is all about good answers.
    – uhoh
    Apr 2 at 1:55
  • 1
    Unless search engines take into account the dup-closure, then, for answered questions, there's likely to be no difference in views of the duplicate. Based on the fact that a number of duplicates have more views than their duplicate target, I'd guess that search engines don't. The closure might increase the views of the duplicate-target, given that there's then a prominent link at the top of the duplicate question. As far as I'm aware, there's nothing that records how many users navigate away to the duplicate-target without looking at the answers on the duplicate when they land there first.
    – Makyen
    Apr 2 at 2:03
  • @Makyen I see you've given this some thought. By coincidence I've just written an answer to a has anyone...? question which says something similar; "if they had, it wouldn't have worked, because..." I'm amenable to such an answer here as well; I love awarding bounties!
    – uhoh
    Apr 2 at 2:41

0

You must log in to answer this question.

Browse other questions tagged .