Take for example the following answer.

"Having the misfortune of being near a rapist is what increases the chances of being raped, regardless of attire. Opportunities for rapists are not so abundant that they can afford to be very choosy. It's situational or compulsion based." - Based on opinion and not backed up by the given reference

"The suggested correlation is so ridiculous that you'd be hard-pressed to find many in-depth studies on it." - the correlation, that rapists may choose to target women with more sexually suggestive dress, is not ridiculously a priori

"Most convicted rapists do not remember what their victims were wearing" - That is certainly relevant information, but not conclusive. Convicting someone takes rather a long time, so maybe they forget. Maybe they had too much to drink that night?

"Victims range in age from days old to those in their nineties, hardly provocative dressers." - that doesn't prove anything

"A Federal Commission on Crime of Violence Study found that only 4.4% of all reported rapes involved provocative behavior on the part of the victim. In murder cases 22% involved such behavior (as simple as a glance)" - Well, of course you don't provoke the random guy you see as you walk down the dark street.

The sole reference given is a safety bulletin, not a study.

Yet, it still has a score of 46.

What should we do when we see answers like this? Is it appropriate to post about especially tenuous, but upvoted answers on Meta?

  • 1
    This whole question is a problem. The accepted answer also cites at least one study which is totally irrelevant and the other two studies are contradicted by one cited in another answer. A “real” accepted answer should collate those research results. Oct 18, 2012 at 8:05
  • @KonradRudolph it's also a relatively old question. Maybe we could simply lock it and add "do not use this as a reference" notice?
    – Sklivvz
    Oct 18, 2012 at 8:15
  • @Sklivvz Ah, if we had such a notice … I’m against locking though since I think this is a valid question and, taken together, the two properly referenced answers are quite good. I’m in a mind of just deleting the unreferenced answer since it’s actually pretty much “not an answer” at all. Oct 18, 2012 at 8:25
  • Personally think this (not this question, the one being referenced!) question should be deleted. Its not a good fit for the site IMO.
    – Jamiec Mod
    Oct 18, 2012 at 10:39
  • @Jamiec: If we were to declare all questions like this off-topic, then we'd have to declare almost all sociology ones off topic too
    – Casebash
    Oct 18, 2012 at 11:51

1 Answer 1


I think we should use meta only if there is any doubt about the answers being broken or not.

If the problem is that they are objectively unreferenced and salvageable (as your example), I see no reason to deviate from our policy:

  1. Challenge the OP to provide references, or directly fix the answer

  2. Flag for mod attention to add a notice if needed

  3. If the OP explicitly refuses to fix (or if the answer is effectively abandoned for more than a month or two), flag for deletion (or delete if you have the powers)

Furthermore, if the answer is deemed unfixable, e. g. If it is an anecdote, skip points 1. and 2. and just delete, with an appropriate, kind and helpful message.

All deletions are undoable, either by high rep users or mods (regular users can't undo deletions by mods, though they can still flag for undeletion).

Finally, it is also perfectly appropriate to point out in meta if one honestly believes a policy has been misapplied, and if one disagrees with any policy, but we should avoid posts of the "my answer has been deleted in accordance to a policy I disagree with" kind. Those should be closed as not constructive.

  • Regular users can't undo deletions by mods, though they can still flag for undeletion
    – Mad Scientist Mod
    Oct 18, 2012 at 8:41
  • Incorporated in the answer
    – Sklivvz
    Oct 18, 2012 at 14:40
  • How can I as a regular user flag for undeletion if I can't see there was a deleted answer in the first place?
    – gerrit
    Oct 28, 2012 at 22:19
  • 1
    As a regular user you can see your own deleted answers, @gerrit
    – Sklivvz
    Oct 28, 2012 at 22:49

You must log in to answer this question.

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