References are not automatically required in any other StackExchange site, not even the ones that are much more based on rigorous scientific research. One adds them when necessary and appropriate. A blanket rule to always include references is counterproductive, especially in a soft-science topic like health which suffers from a lack of rigorous scientific research, where you sometimes have not well motivated research being done done and bad conclusions drawn as a result.One should not confuse "health" with "medicine" where one does try to uphold more rigorous standards but where things are still far from perfect compared to e.g. physics.

Simple example: can exposure to microwave radiation from mobile phones affect health? One can look at the peer reviewed literature published in medical journals, the conclusions are all over the place and you then have a WHO report that out of an abundance of caution argues caution. But simple physics will tell you that negative health effects are pretty much ruled out to a degree that you can't extract from the medical literature. That fact can, of course, also be extracted from the scientific literature, but the point I'm making here is that there is hardly any value of including references in an otherwise well written answer. You can pretty much get a citation for pretty much any claim you write down on this topic within a huge margin, much larger than what is plausible given all the scientific knowledge.

When it comes to general health issues that has more to do with how the human body functions or is supposed to function when it is not ill as opposed to problems caused by illness, then problems are much larger. If you have developed a drug X to treat illness Y, then the burden of proof is on you to prove that X works to cure Y. That's a straightforward problem we know what the end point is so the research you're going to do is well defined.

But if the problem is whether vitamin D is good for health, then you don't know a priori what to look for which severely complicates things. In case of drug X which is not made by the human body nor found in a normal diet, the null hypothesis can justifiably be taken to be that it does nothing, so that the burden of proof is fully on you to show that it works. In case of vitamin D we know that it does have a biological role also beyond bone health. So, taking the null hypothesis to be that it does nothing like in case of drug X would likely lead to skewed result. It may still be useful in the sense of showing what some RCT have proven but I would never stop taking my high dose vitamin D pills just because a RCT did not find a link between vitamin D and heart disease.

Nevertheless, the official advice in the medical literature is based on treating vitamin D just like that drug X , which is why I don't take it serious. I can of course still motivate what I'm doing from citing the literature but the reason why I would then cite these sources is then based on my own reasoning, and that's why I think it would not add much to include those refs. I could still include them to refer to refer readers to more information, but that should not per se be construed as additional support for the validity of my reasoning. Whether I then decide to add the refs or not is then purely an editorial decision up to me, it should not be a requirement to validate the answer.


I would refer you to this meta question, which has already been asked and answered.

Should we require references to back up all answers?

Strongly encourage and required both got a large number of upvotes. Also note, that neither of these answers were provided by moderators, they were supplied by users like yourself.

Your answer on that suggesting that "in the lack of evidence, we rely on our body design" was downvoted, and that you have been asked and warned by the moderators to back up your claims with the strongly encouraged/required references. Currently you have 22 answers (Some deleted) and in many of them are requests to back up your assertions with references. Your response to the request for references is to highlight a few words and say "e.g. here" for the reader to go to another place to read the article. As pointed out by a CM moderator in this answer to another meta question, this is bad practice as well.

You may also note my answer on that question, stating that fitness (A much softer science than health) challenges assertions as well, and tries to steer away from "bro-science" and "common knowledge" answers without backup. (And that was written well before I became a moderator).

Currently the community has spoken that they want their information backed up, and by all appearances this is another attempt to circumvent that desire. If the community wants to come out and collectively say that references are superfluous, then we can revisit the stricture on references. Until that time, we would expect you to adhere to the structure set in place by the users of the community.

  • 1
    The community doesn't get it, they ignore the fact that a lot of bad science has made it into the peer reviewed journals and that there is a lack of research on some very relevant issues. The requirement of references leads to a false promotion of certain ideas. It's better to be honest, if no references are given and you give your own reasoning for a position, as then it's very transparent that what you say follows only from what you've written, you don't invoke another authority. Sep 8 '15 at 22:48
  • 1
    @CountIblis - The community doesn't get it..." When at odds with a number of knowledgeable people, it is not always the most prudent course to assume everyone else is wrong and you're right. Sep 8 '15 at 23:11
  • 1
    @anongoodnurse argumentum ad populum isn't the most prudent course either.
    – Dr. Duncan
    Sep 9 '15 at 16:37
  • 1
    @Dr.Duncan - Argumentum ad populum is the model that all stack exchange sites work on. The community upvotes good answers, downvotes bad ones. On meta, downvotes indicate disagreement with the proposal. Currently this proposal has 3 downvotes, against 35 cumulative upvotes on whether or not to require references. Currently the Count has 22 answers with a collective -14 score, which indicates the community does not believe the majority of his answers have merit.
    – JohnP Mod
    Sep 9 '15 at 16:50
  • 2
    @CountIblis - So, because a few bad studies have made into peer reviewed journals, that invalidates every published study, and we should just blindly trust what people say? If you feel an answer uses a study that is flawed or erroneous, you are welcomed, even encouraged, to post an answer that refutes it and shows why it is bad. To me, that is much more valuable than a bunch of random stuff with no corroboration, as it would expose and educate.
    – JohnP Mod
    Sep 9 '15 at 16:52
  • @JohnP the main paradigm of medical science is flawed when it comes to general health issues as it relates to healthy people with no significant health complaints. There is not just a "few flawed studies" there are big holes in the entire field. The fundamental problem is that biological plausibility arguments are going to be disfavored. While there is some discussion about that in the literature, the main paradigm is still the same. That doesn't mean that references are alway worthless, just that the requirement of backing up everything one says in every instance is over the top. Sep 9 '15 at 17:10
  • Note that may of my answers were downvoted mainly because of a lack of refs. If I had written down something completely stupid but backed up by refs than such answers would have gotten many upvotes. Sep 9 '15 at 17:11
  • Wow, That answers a lot of questions, @JohnP. The entire model of this site is based upon a logical fallacy and mob rule. So, all I have to do is write things that people want to hear, then I'll be popular enough to exert power over other users. Thank you.
    – Dr. Duncan
    Sep 9 '15 at 17:21
  • 2
    @Dr.Duncan - Reductio ad adsurbum. Well played.
    – JohnP Mod
    Sep 9 '15 at 17:40
  • 1
    "If I had written down something completely stupid but backed up by refs than such answers would have gotten many upvotes." Two observations: 1. if something were completely stupid, you'd have difficulty in finding good reference, and 2. Critical readers can usually spot bad answers, even those with (usually bad or, as mentioned before, misinterpreted) references. The lack of references is only one of a number of possible reasons for DVs, though you'd have fewer DVs if you did cite references. Sep 11 '15 at 8:05
  • The issue is that for general health issues not related to pathological issues, good references in the medical literature will tend to approach the problem from a wrong, pathological, angle, leading to possibly stupid conclusions. Medical science is notorious for getting some issues systematically wrong over many decades due to this approach. Sep 11 '15 at 18:58

...especially in a soft-science topic like health which suffers from a lack of rigorous scientific research...

The first thing that came to my mind when I read this sentence was the amazing science done by Louis Pasteur. There was nothing "soft" about his science, and countless millions of people are alive today only because of his tremendous scientific contributions. Your faulty characterization of medical science reveals a deep lack of understanding not only of medicine but of science in general.

One should not confuse "health" with "medicine"...

Medicine is all about health. There are plenty of healthy people who know nothing of science or medicine, but that hardly means they are experts on the subject, or that a site such as health should allow any person who may have some health to express their erroneous opinions.

...where you sometimes have not well motivated research being done done and bad conclusions drawn as a result.

What has this to do with our site? Every field has suffered from badly designed studies and/or misinterpretation of results.

Similarly, your decision not to take the medical literature seriously is not an adequate reason to allow you or others to post answers without citations. Probably the great majority of people who write answers consider them well written, even if they are completely wrong. That's because of a failure to distinguish beliefs from facts. No one is immune to this common error.

That is why beliefs are not enough on Health.SE.

  • Comments are not for extended discussion; this conversation has been moved to chat.
    – JohnP Mod
    Sep 9 '15 at 14:24

You must log in to answer this question.

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