In my answer to Is Genetically Modified food safe to consume?, I cite a study that I call controversial, and link to a Nature article summarising debate on the contents. When I'm no expert, I rely on judgement by others to tell the difference between crackpot theories and the scientific method. In this case, however, one comment motivated a downvote because I gave undue weight to a particular scientific paper, that he considers debunked (again; as an outsider, I cannot tell the difference between debunked and an ongoing debate, short of the journal retracting a paper for fraud or similar).

Lacking expertise in the topic, what are the criteria for determining what undue weight entails?


1 Answer 1


That was just an individual's opinion that informed their downvote.

I don't think as a non-expert in a topic, that you should be expected to have the ability to always determine if something has been given undue weight.

Answer as best you can, let comments and votes come in, and if you agree, edit your answer given the new information you may have gained.

As far as I can tell, there is no established policy against "undue weight" on this site. It is a concept well established at Wikipedia, and flows from their policies on Neutral point of view and Reliability. I would guess that reading Wikipedia's policy on the matter would be the best way to figure out what that paritcular user means when they caution against undue weight.

Skeptics.SE doesn't have an explicit "Neutral point of view" policy. Any guidance towards a neutral point of view and away from undue weight is all case-by-case as informed by the people that happen to be interested in a paritcular question or answer.

  • The individual is a diamond moderator and ranks in the top 20 users by reputation, perhaps that makes a difference.
    – gerrit
    Apr 6, 2013 at 19:15
  • 1
    That's okay. Moderators don't have any special control over the content of answers or community consensus. In the capacity that they voted on and commented on your question, they were acting as a normal user.
    – user5582
    Apr 6, 2013 at 19:17
  • 2
    Good answer. The fact that Skeptics doesn’t have an equivalent of WP:UNDUE is IMHO an oversight, however. We should have it, for two reasons: (a) it’s rational. You need to define some kind of prior (even if implicitly) to reach conclusions. (b) the alternative would mislead readers of the site. Case in point: gerrit’s answer. Apr 6, 2013 at 19:21
  • 1
    @Konrad: I don't disagree with your points, but every time we lift the standard, we have to ask "Are we shooting ourselves in the foot by making it too hard for one mortal to answer a question?" Many answers here are based on a single source, especially esoteric nonsense claims. Must we wait for a second study before we can answer? I'm really asking "How much is 'undue'?" and "What are the consequences? Downvoting?"
    – Oddthinking Mod
    Apr 7, 2013 at 5:13
  • 1
    @Oddthinking I don’t think that’s related. On the contrary, a single source can absolutely be enough if it represents the best available evidence. What I’m objecting to would be to juxtapose that one study with a second study of highly inferior quality, and concluding that the evidence is mixed. Apr 7, 2013 at 9:12
  • 1
    @Konrad: ah, the issue of false balance. Understood.
    – Oddthinking Mod
    Apr 7, 2013 at 10:57
  • 2
    Shouldn't voting take care of these qualitative issues?
    – Sklivvz
    Apr 8, 2013 at 9:39
  • @Sklivvz I believe it would, if comments for improvement are also left. As in this case, the answerer may not realize they have given undue weight. At Wikipedia, the policy is workable because articles are edited continuously by many people. Here, a single person is responsible for their answer's content, and a policy requiring those answers to have neutral point of view or to avoid undue weight would be unworkable (due to good faith ignorance). Something lighter may be helpful though: a guideline, or a meta answer that downvoters could point to, etc.
    – user5582
    Apr 8, 2013 at 16:58

You must log in to answer this question.

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