13
$\begingroup$

The issue of unanswered question fraction is always in discussion (and it being a factor for site graduation).

In my opinion a lot of mechanisms are unknown in biology (even for some commonly observed processes). This is unlike the problems addressed in stackoverflow or any other discipline for which there will always exist a precise answer (or perhaps less number of unanswerable questions).

Have a look at these two questions:

  1. What are the side-effects of long-term liquid breathing?
  2. How does Sodium Valproate cause neural plasticity

There may be more like these. They are good questions, logically posed and clear. The issue is that we don't know the answers for these; the research community doesn't know the answers (I assume that because for these two questions I could not find any research papers that answer them).

What is/should be our policy on such currently unanswerable questions? They do increase our unanswered question count. It is justifiable to keep those questions until the answers are found out and someone posts them but I guess the purpose is to also give answers that are useful at present. One can always add new answers later.

$\endgroup$
3
  • 4
    $\begingroup$ A problem with such questions is that once they fall off the start page, most people will not find them and write an answer. $\endgroup$
    – Chris Mod
    Nov 18 '14 at 12:35
  • $\begingroup$ In the same vein, what about a question like this: biology.stackexchange.com/questions/21412/… It also seems unanswerable due to a lack of information, which was well discussed in the comments. Should it be closed or should someone summarise the reason it cannot be answered in an answer? $\endgroup$
    – canadianer
    Dec 9 '14 at 11:02
  • $\begingroup$ @canadianer Well. In that case the OP has to provide more information such as the plasmid map. The kind of unanswerables that I am talking about is those that have not been researched upon, yet. In my opinion it is best to close that question. $\endgroup$
    – WYSIWYG
    Dec 9 '14 at 16:22
10
$\begingroup$

I have answered a number of questions on Stack Overflow by essentially saying "This is not possible" or "The software you're using doesn't have that feature" or something similar, and more often than not these get accepted as the correct answer. They aren't one-line answers, as I try to back up my assertion with proof, of course, citing any references I can.

Here, I would think it's perfectly acceptable to write up an answer saying that no research has been done on this subject, or the pathway has not yet been elucidated, or whatever is appropriate for the question. Back this up by explaining what you've done to determine this, and perhaps link to groups that are working on this field, if you can find any.

We've discussed that we tend to answer questions in the comments a lot, so even if there's a comment saying "This hasn't been studied" or whatever, these questions, as long as they're on-topic, deserve an answer, even if it's just expanding on the comment.

$\endgroup$
1
  • 1
    $\begingroup$ Re the last paragraph - That's what I saw in the comments on my question (the first listed in the question here). I got the feeling that I essentially got a null result - there's nothing that we know about liquid breathing - and I felt that the comments given were helpful. $\endgroup$
    – HDE 226868
    Nov 19 '14 at 0:22
6
$\begingroup$

I can see two problems with closing questions which are good but have what we deem to be unanswerable due to a lack of suitable research or scientific material:

1) It is based on an assumption that we have enough breadth of expertise among the very regular visitors, some questions may be answerable just not by people that are visiting the site (at all or often enough). I simply don't believe that we have enough highly expert prolific contributors to cover every known facet of biology. I think we would have to increase the number of close votes required because in many cases (where research activity is low) I don't think 5 users could have the joint breadth and depth of knowledge to say that an answer doesn't exist.

2) There may well be an answer soon enough, I think it is perfectly valid to answer one of these questions by saying it is currently unknown while discussing related things are known and trying to infer some kind of expectation - it could even prompt someone in to doing the research!

$\endgroup$
1
  • $\begingroup$ I agree with you. To say that an answer doesn't exist we need to say where all we have looked. $\endgroup$
    – WYSIWYG
    Dec 3 '14 at 17:34
2
$\begingroup$

I recently looked at old unanswered questions, and many of these should be closed in my mind (and I voted to close on a number of them). There are certainly many good unanswered Qs (which should remain open), but I think we could in part adress the fraction of answered Qs by going through and closing Qs that are unclear/off-topic etc. This is an easy way to increase the answered%, but will also increase the quality of the site by either removing bad content or pushing askers to improve on their questions.

$\endgroup$
2
  • $\begingroup$ Good idea. I will have a look at these questions, too. $\endgroup$
    – Chris Mod
    Nov 24 '14 at 14:02
  • $\begingroup$ @Chris Nice to see that this little effort has already pushed our answered% up one point. $\endgroup$ Nov 24 '14 at 23:19
1
$\begingroup$

2 questions? No big deal - leave them open. I'd say if a large percentage of questions asked are unanswerable atm because of limited research, that would be a case for closing these questions as unclear or whatever, but as there are at most a handful of these right now, a better way to use your 'unanswered' time would be to leave those and answer some of the old forgotten questions no one paid attention to earlier. At least, if you think you can answer well.

$\endgroup$
4
  • $\begingroup$ I guess there are more like these. I tried answering some old forgotten/missed questions and while doing so, found questions like these. $\endgroup$
    – WYSIWYG
    Nov 18 '14 at 12:22
  • $\begingroup$ Yup, answer the others, if there are a huge amount of unanswerable questions left, then is the time to take care of those. $\endgroup$
    – J. Musser
    Nov 18 '14 at 12:25
  • $\begingroup$ Many of the unanswered questions are indeed interesting, but pretty much unanswerable. By the way: does the 87% answer rate concerns accepted answers only, all answered questions regardless acceptance, or all answered questions including those with only comments? $\endgroup$
    – AliceD Mod
    Nov 18 '14 at 14:08
  • $\begingroup$ @ChrisStronks the unanswered questions list is subtitled "questions with no upvoted or accepted answers" but I couldn't find evidence that this is the same definition used in the targets page. I would guess it is the same, but I can't confirm that. $\endgroup$
    – trichoplax
    Nov 19 '14 at 20:10
1
$\begingroup$

I have an idea about this.

Some of these questions have informative comments. It would be appropriate to compile those comments (retaining the names of the people who posted them) and post them as as community-wiki answers.

$\endgroup$
1
$\begingroup$

I can tell you math has a lot of unanswered questions. As of yesterday at 18:06, by SE time, our bot told us we have 81963 unanswered questions. In fact, we have so many unanswered that we have a room dedicated to solving them, meet The Crusade of Answers. Additionally, we use a sql query on SEDE called crossbow which finds questions by user over 180days (I think) that are considered unanswered; that is, the question has an answer with no upvote. We can then read the answer and determine if it should be voted on, meet Crossbow.

If you really feel the unanswered questions are getting outline, you can try what we do and start a room dedicated to finding, answering, and voting on worthy answers of unanswered questions. Presently, I don't think your unanswered is too crazy. I am not sure what SO's number of unanswered questions are but I would think higher than 100k.

The user currently known as Woodface noted that we can use the search feature to find questions with negative and zero score multiple answers with

answers:1 isanswered:no hasaccepted:no closed:no

You can add [your favorite tag] if you want to search in tags to the above query.

$\endgroup$
0
$\begingroup$

I have now marked a number of questions from the end of the unanswered questions list (which received no votings and only little attention for closure. Most of them as "unclear what you are asking" since there is no real tag for inactive.

Additionally I stumbled upon a number of closed question which did not met the criteria to be deleted automatically. I nominated a number of the for deletion, how do we proceed here? Go through the questions (sort them by votes and then go to the end of the list) and nominate them for deletion to increase our quality?

Any suggestions?

$\endgroup$
7
  • 2
    $\begingroup$ I don't think closed Qs count towards the site statistics thought, so deletion is probably not that important (even so, cleaning up might still be a good thing). For your first point; I don't think Qs should be closed just for being inactive - they should be closed for a normal, valid close reason. But maybe this is what you meant? The Qs you voted to close were certainly problematic (and many are now closed). $\endgroup$ Nov 24 '14 at 23:24
  • $\begingroup$ @fileunderwater I think closed questions without answers do count. I think we can remove these anyway (except for the rare case of finding hidden gems here what I find relatively unlikely), especially when they already received multiple downvotes. I have targeted these yesterday and will go on with it. These questions do not add any value to the site. $\endgroup$
    – Chris Mod
    Nov 25 '14 at 6:41
  • $\begingroup$ You might be correct. Since e.g. upvoted but not accepted answers still count towards the answered% I just assumed that closed Qs were not included (since they cannot even be answered at the moment). I can only find clear info on this though, only some discussions in SOMeta $\endgroup$ Nov 25 '14 at 10:38
  • $\begingroup$ @fileunderwater Even if they do not count they can go. You can find them when you click on questions, then on vote (which sorts the questions by number of votes) and then go to the end of the list. There are downvoted, closed and mostly unanswered questions. $\endgroup$
    – Chris Mod
    Nov 25 '14 at 10:46
  • $\begingroup$ For the record, I just calculated the ratio of answered open Qs over open Qs ($#(closed:no answers:1) / #(closed:no)$) and this replicates our current answered% (0.868), so I do not think closed Qs are included. I agree that many of the closed Qs can go though - currently 194 of closed Qs lacks any answers while 252 have at least one answer. $\endgroup$ Nov 25 '14 at 10:54
  • $\begingroup$ yeah I agree with fileunderwater that questions should not be closed for being inactive. $\endgroup$
    – WYSIWYG
    Nov 25 '14 at 11:34
  • $\begingroup$ I agree here, too. I am not blindly voting to close old unanswered questions. $\endgroup$
    – Chris Mod
    Nov 25 '14 at 11:36
-1
$\begingroup$

I have a few ideas as well:

  1. I think we should have a message under the question stating that:

As of ******* date, there is not enough experimental/research data to give a satisfactory answer to this question

  1. The ability to search for questions marked as such should be added. We need not close them if this can be done.

When in future such questions are viewed and it is realised that it can be answered, the community (the user who knows the answer) can answer it.

$\endgroup$

You must log in to answer this question.

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