Lying via citations

posted by Carl V Phillips

When students are taught to write their first toy research paper, in grade 4 or so, they learn to find and cite sources for their claims.  The mere act of finding some source for a claim is a major intellectual development, moving for the first time beyond just writing what they believe.  But by the time they finish high school or at least university, those students learn that just because someone said something, even if the source seems to have credentials or the document appeared in some impressive forum, does not make it a legitimate source to cite (or, if they do not learn that, they should ask for their tuition to be refunded).  Moreover, the cited material needs to support the core claim that it is attached to not some non-core aside in the statement, or only some aspect of the claim, and it needs to be a genuine source of that information, not just someone making an unsubstantiated claim and writing it down.

Unfortunately, the corollary to that — that readers need to be wary of authors violating that rule — is not learned nearly as often.  Liars who make “sciencey” claims realize this and take advantage of it.  (Of course, I suspect some of them may not even know that blind use of citations is D-minus work in itself, but that just means they are both liars and know nothing about scientific inquiry.)  Indeed, some of them go even further and just tack on source citations that do not even seem to relate to the claim being made, figuring no one will notice.

Consider the ERS anti-THR position paper that I started discussing yesterday. The particular passages I quoted and responded to then:

The European Respiratory Society, ERS is opposed to the use of all tobacco and unapproved nicotine delivery products such as cigarettes, chewable tobacco, and emerging products that include electronic cigarettes (e-cigs), snus, dissolvable tobacco and waterpipes.

In response to the successful increase in tobacco-free policies, the tobacco industry has developed these new products, allowing consumers to obtain nicotine without the use of a cigarette.[1] In many cases, these new products are claimed and/or perceived to be ‘harm reducing’ or safe alternatives to conventional cigarettes; however, there is no reliable science to substantiate this claim.[2] Rather, available research suggests that these products pose a significant health risk to citizens, placing them at continued high risk for disease and negative health outcomes. [3,4]

Reference [1]  is to a paper in one of the anti-tobacco advocacy journals from 2007.  Even setting aside the lie about how the “tobacco industry developed” these “new” products (see previous post), how exactly would a paper from 2007 been able to report on the new dissolvables (some existed then, but did not even qualify as a niche product; they are overwhelmingly a product of the last few years) or even e-cigarettes (had started to become recognized by researchers then, but were barely on the scene).  But worse, the paper is a review of literature on perceptions, beliefs, and marketing, but the claim in the sentence is about the motive of the industry, something that the review could not have addressed.  Grade: F for utter lack of support for the claim in the source, and for citing a puff piece from 2007 (ancient history in THR terms) in a current position statement.

Reference [2] is an unsigned position paper of unknown provenance put out by a political organization (FCTC).  It should not be cited as scientific evidence for anything, in much the same way Wikipedia should not.  Either one can be used for a resource to find a real source of information, but if content cannot be traced back to a real source, then it has to be considered as being at the “I saw it on the internet somewhere” level.  Moreover, as far as I can tell, the document — as political as it is — never goes quite so far as to blatantly lie that there is no evidence that smoke-free alternatives are lower risk than smoking.  Grade: D for citing an unsigned political document, lowered to F for citing it inaccurately.

Reference [3] is one of the blurbs from FDA making absurd claims about e-cigarettes (look for the upcoming series on lies about THR by the US government).  Even if FDA were a reliable source on this topic, this is just a little statement directed at consumers, and so is similar to quoting a “public service” television advertisement or a pamphlet found in a medical office.  And guess what?  It still does not actually include any claims that “these products” cause substantial health risks — not even about e-cigarettes, let alone the other three smoke-free products among the list of “these” that are not even mentioned.  Thus, it most certainly does not contain claims that “research suggests” that this is true (and, of course, the blurb is certainly not research in itself, though perhaps someone like Christina Gratziou thinks it is).

The adjacent reference [4] is little better.  It is a lie-filled “advisory” from the Canadian government that is part of their rationalization for banning e-cigarettes.  And yet, again, even as laden with anti-THR lies as it is, it does not actually claim that there are substantial health risks or identify any research.  It is worth reiterating that neither of the cited sources addresses any product other than e-cigarettes at all, even though the sentence clearly refers to the entire list of products.  Grade:  I really wish there was something below F.  Oh, wait, there is: academic misconduct.

Next comes:

There is a growing concern about the potential health risks associated with e-cig usage and exposure. A recent sample of the product was found to contain carcinogens and toxic chemicals such as diethylene glycol, an ingredient used in antifreeze.[5]

That, of course, refers to the biggest anti-THR lie of recent times, in which the US FDA lied about unimportant contaminants they found in e-cigarettes.  Citing that thoroughly debunked propaganda is C-minus work in itself (only that high because there was at least something that might be called a study behind it).  But, wait.  The citation is to “Utah Tobacco Prevention And Control Program- position on E-Cigs”.  Yes, you should be laughing out loud by now — a supposedly learned European scientific society could not even manage to cite the original junk science, but instead cited a position paper from a little activist group in Utah.  Definitely lowers it to an F.

ERS fully endorses the Article 14 guidelines of the WHO Framework Convention on Tobacco Control[7] and therefore maintains that the best health advice to smokers is to quit smoking.

At least that bit is honest (the citation is to an FCTC document, of course).  The only honest statement in the entire broadside.  Translation:

We blindly follow what the FCTC says, and have neither opinion nor understanding of our own to offer on this topic.

I skipped [6] — that one gets a whole post of its own.

Last point (which you can skip if you are not particularly interested in the theory of citations):  It has long been clear to me that it is a very bad thing that public health adopted the endnote method of citation (in the original, those numerals are the standard superscript, by the way) used by natural sciences, rather than the embedded name and date format (e.g., “(Phillips et al., 2007)”) used in other social sciences.  First, it makes it very difficult to use the content footnotes that are needed in the more complicated social sciences, but not so much in simple natural science reporting.  More important, it tends to hide the information away from all but very attentive readers who jump to the end of the paper for every citation.  In decent natural science reports, this is probably not so necessary, but in social sciences it is quite often very crucial who said a particular thing and when.  Often any expert reading the paper would recognize that the particular source is biased or useless from the embedded citation, but might miss it in the endnote.  This, of course, does not fully explain why modern “public health” people lie so much and get away with it, but it does contribute.  I have a feeling that if the original editors in this field could see what they wrought, they would have chosen the social science style of citations.