We have established that this site “strongly encourages” references in answers. Although it does not appear to reflect full consensus, I even upvoted the answer that states that we should require references. However, as a moderator,* I have to make a decision: does “strongly encouraging references” mean that I should immediately delete all answers that do not include references?

*Soon and very soon, this will also be a decision relevant to 4k+ users.

  • 6
    I don't think a good answer should be deleted just because it lacks references. Can downvote if they are not provided but no need to delete.
    – Joe W
    May 4, 2015 at 0:11
  • 1
    @JoeW - It can't go both ways. An answer can always be undeleted if they fix it and add references. I would think that asking for references and deleting if they are not provided after a suitable time is appropriate.
    – JohnP
    May 4, 2015 at 0:40
  • 1
    It is a lot harder to get something un-deleted then it is to get it deleted in the first place especially if a moderator is the one deleting it. If an answer has no problems other then the lack of references it should not be deleted and let voting deal with the issue rather then deletion. Just because an answer has references doesn't make it a good one and just because it doesn't have them doesn't make it a bad one.
    – Joe W
    May 4, 2015 at 1:04
  • @JoeW I think I expressed a similar sentiment much less concisely in the answer posted with this question.... ;-)
    – Susan
    May 4, 2015 at 1:06
  • No. The rest of this is for character requirements.
    – Fomite
    May 4, 2015 at 5:37

1 Answer 1


N.B.! Although the question was asked on behalf of the moderation team, this answer represents only my personal opinion and should not be construed to reflect the view of the moderation team I work with, the community team who appointed me, or anyone else.

I do not want to do this. I can and will down-vote, comment, and add post-notices:

Some of the information contained in this post requires additional references. Please edit to add citations to reliable sources that support the assertions made here. Unsourced material may be disputed or deleted.

And I will participate is such disputation and deletion (after disputation). However, I think the notion that moderators should reflexively delete anything without a reference neglects the nuance of the situation. Posts can not be (sanely) divided into: has a reference | doesn’t have a reference.

Has a reference (but still ‘aint no good)

  • Includes a reference (or twenty) that is tangentially related but doesn’t actually support the claims in the answer.
  • Includes reference(s) that back up the obvious claims in the answer but don’t speak to the more substantial aspects that actually need a source.
  • Includes reference(s) that, taken out of context, seem to back up an important claim, but, when the data are viewed with a more critical eye, they do not. (See also, How to Lie with Statistics.)
  • Includes reference(s) to the University of X website that contains someone else’s opinion with no author cited, let alone peer review or references to primary literature. You may as well have made it up yourself.

Doesn’t have a reference (but still has value)

This is harder. Most of these indeed deserve DV and post notices given the stance this site has already taken. However:

  • Many are actually better than the ones in the group above. In academic literature, there are certain assumptions made about background knowledge. Some questions here could be answered by invoking basic concepts that could be found in any introductory textbook. An expert who drops by and sees such a question is likely to simply answer it. We’re different! We need references! I know.

  • Having good, unreferenced information sitting around, getting down voted, graying out, is good for the community.

    • Reminder: include references when you answer.
    • Opportunity: participate in cleaning up this site by editing to include references and improve answers.
  • Having bad, unreferenced information sitting around, getting down voted, graying out, is good for the community:

    • Reminder: include references when you answer.
    • Opportunity: participate in cleaning up this site by voting to delete crap.*

I’m not arguing for lower standards; I’m arguing for higher ones. I want to see critical comments, DVs, and eventual VTDs in all four categories in the first group in addition to the second group. We need to foster a discerning community that appreciates subtlety and approaches science with rigor.

The presence of references doesn't correlate consistently with quality or even acceptability. Simply hitting the delete button because there’s no link may be an easy out that confers the appearance of scrupulousness, but it’s not a wise approach.

*Pending a substantial body of 4k users, flagging for mod deletion (after a reasonable period of time has elapsed with consistent DVing and no attempts at a rescue by the OP or good Samaritans with Pubmedding skills) is reasonable. I would be happy to work with the community to quantify «reasonable period of time» — 30 days?

  • 4
    One note on the "people don't want to DV and lose their precious 1 point" front: if the above system were implemented perfectly (which it never will be, of course, but we could approach it), a user would be guaranteed (the opportunity for) a refund on DVs for unreferenced answers. If it never gets fixed, it gets deleted (-->refund); if it gets fixed, it pops back to the top and you can change your vote (-->refund).
    – Susan
    May 4, 2015 at 6:39
  • 2
    I know some people don't like post notices (seeing them as sort of a badge of shame), but I think adding them "serves notice", so to speak, that something needs to be fixed. Comments should add whatever further detail is needed. Then, later, you can search for posts with notices and delete ones that didn't get fixed. I think a two-stage approach will serve you better than immediate deletion. May 7, 2015 at 14:31
  • With respect to getting rep back: even if votes don't get undone, it'll only be a really tiny amount of rep, unless the user kept on posting iffy answers without references after being informed, in which case the downvotes are probably well-deserved anyway.
    – Cascabel
    Jul 24, 2015 at 5:01
  • what if an answer in comment does have good relevant references, but slightly not really answer the question? For example, if the question ask about if X happens, with condition Y1 goes along, then the comment says X will happen with condition Y2, then will it a good comment?
    – Ooker
    Sep 29, 2015 at 19:31

You must log in to answer this question.

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