Shop Forum More Submit  Join Login
By Emily Willoughby and Jonathan Kane

In nearly every university course I've taken that required any amount of independent research, the professor has always jokingly told the class: "No citing Wikipedia!" Such a request would tend to be followed by some amount of laughter from the students, because it seemed so ridiculous: who in their right mind would seriously cite Wikipedia, even for a silly undergraduate research paper?

Imagine my shock and surprise to learn, years later, that it's apparently quite common—and not just among students, but professional academics, researchers and medical doctors as well!

One recent and particularly appalling example comes from an article by medical anthropologist Alondra Oubré in the high-profile blog Scientia Salon, maintained by philosopher Massimo Pigliucci. Oubré is a fairly well-known author, responsible for books such as Race, Genes and Ability: Rethinking Ethnic Differences among others. In the aftermath of Nicholas Wade's new book A Troublesome Inheritance, Oubré published a scathing diatribe criticizing Wade's treatment of a version of the human gene that codes for monoamine oxidase-A, known scientifically as MAOA and colloquially as "the warrior gene". As described in this paper, this gene is responsible for an enzyme that acts as a neurotransmitter degrader and which plays a role in arousal, including emotion and impulse control. One particular variant of this gene is shorter than the most common variant, and produces less of the relevant enzyme. This variant appears to correlate with less impulse control and greater risk of aggressive behavior in the people who carry it. And it also seems to be the case that this variant is unequally distributed among ethnic groups, with the highest incidence among sub-Saharan Africans and Maori (from whom the term "warrior gene" originates).

The problem arose when Oubré reported the numbers for this unequal distribution: for example, she wrote "In the Add Health database, 5.5% of African American men, 0.9% of Caucasian men, and 0.00067% of Asian men" had the gene variant associated with increased aggression. But the Add Health database said no such thing, and, in fact, these specific numbers can be traced directly to a piece of Wikipedia vandalism that was eventually undone. Her article misleadingly cited these statistics to the Psychiatric Quarterly paper linked above, but the numbers she gives are not the same as the numbers given in that paper, so there's no explanation for where she got the fake numbers other than that she copied them from the Wikipedia article. Note that the number given for Caucasian men was off by nearly a factor of 10!

These are far from the only problems with Oubré's article on the MAOA allele, the rest of which you can read in this excellent article at the Unsilenced Science blog. This blog post points out one thing that's especially significant: these dishonest statistics continue to exist in her article even after corrections were submitted to her blog, as well as to Pigliucci, the blog's editor. Taken together, these events clearly demonstrate how reliance on Wikipedia can inform the academic work even of people who should absolutely know better. This isn't the first time such an egregious violation has been committed, nor will it be the last.

The distortion of scientific research seems to especially be a problem in medical articles, both because of their prevalence of vandalism and misinformation and because of the fact that both doctors and laymen actually rely on them. In fact, a May study published in The Journal of the American Osteopathic Association found that 9 out of 10 Wikipedia articles on medical conditions contained some kind of obvious factual error when compared to peer-reviewed studies. This probably wouldn't seem like a big deal if doctors didn't rely on Wikipedia as a source of factual information, but unfortunately they do, with 50% of physicians admitting that they've consulted Wikipedia for health-related information. One of the best examples of an inaccurate medical article on Wikipedia having far-reaching consequences is the case of "glucojasinogen". In 2007, an anonymous IP address trolled the Wikipedia article on diabetic neuropathy by adding a reference to the fake condition it called glucojasinogen, which it wrote resulted from excess glucose in the blood and was associated with erectile dysfunction and epigastric tenderness. This was, of course, a made-up term and was complete bullshit, but it stayed in the article long enough to get repeated by three academic journals: Research in Pharmaceutical Sciences, the International Journal of Health Research, and the Indian Journal of Pharmaceutical Sciences. Today, the fake term "glucojasinogen" has become one of the first Wikipedia neologisms, with a significant presence on Google Scholar.

Scientists' reliance on Wikipedia has increased by an order of magnitude in the past four years, with peer-reviewed medical papers experiencing a more than a tenfold increase in citations to Wikipedia between 2010 and 2013. This is one reason why Wikipedia's hoaxes so often find themselves appearing in books and articles elsewhere. Some of these hoaxes are just good for a laugh, such as the case of "Azid", a prank that resulted in a friend's name becoming synonymous with an Indian chicken dish. Depressingly often, though, they have the potential to do a lot of harm, both to one's health and to public understanding of controversial topics. This is especially true in cases like Oubré's article, when the authors of articles are unwilling to correct the errors even after they've been pointed out.

When research on topics such as MAOA is obfuscated by scientific authors, whether due to laziness or for ideological reasons, it can result in specific repercussions. As is pointed out in this widely-used textbook, it is fairly well-established that violent crime disproportionately affects African-American communities, although it is an unanswered question whether the reason for this is entirely due to cultural and socio-economic factors or whether it also involves genes such as MAOA. Because African-Americans are over-represented as both perpetrators and victims of violent crime, black community leaders have long regarded crime in black communities as an important social issue which needs to be addressed, including Barack Obama in his comments following the death of Trayvon Martin. Although this may seem counter-intuitive, if further research were to conclusively show that MAOA is involved in race differences in crime rates, this actually would be a step towards solving the problem. There already is some research suggesting that this gene's relation to aggression can be changed by environmental factors in early childhood, and further research in this area might clarify what sort of environmental intervention is necessary to prevent MAOA from increasing the risk of violence—but this will only be possible with an understanding of exactly how this gene plays a role, if it does.

By reporting statistics that came from Wikipedia vandalism, even after being made aware that this is what they were doing, Alondra Oubré and Massimo Pigliucci have done a massive disservice to progress in this area. Most people assume that Wikipedia vandalism typically just involves edits like this, without realizing that the problem runs far deeper. Your undergraduate professor was right: relying on Wikipedia for factual information is a terrible thing to do for a variety of reasons, and scientists need to learn this lesson more than anyone.
Add a Comment:
TheCapeWildMan Featured By Owner Sep 8, 2014
Instead of complaining about how inaccurate Wikipedia is, I think you should join the community as an editor and help improve the accuracy! Wikipedia needs as many hands as possible who are qualified and able to fact-check for accuracy.
Agahnim Featured By Owner Sep 26, 2014
What makes you think we haven't tried that? Our cynicism with respect to Wikipedia is mostly the result of our own experiences there, as well as observing other people having the same experience in the same situation.

This wasn't just the result of inexperience on our parts, either. It was a situation that involved about twenty different people, and virtually all of them agreed there was a problem, but it took three years to get anything done about it. Part of why Wikipedia's articles about psychology and genetics tend to be in poor condition is because while this issue was going on, several of the editors who used to maintain these articles quit out of protest.

I can tell you more detail about this if you want to know, but it's not very interesting unless you've got an interest in corrupt online politics.
doctormo Featured By Owner Edited Aug 15, 2014  Professional Digital Artist
Wikipedia is a very important leading open access platform which is not controlled by a government nor a corporation. I think science professionals are starting to get a better handle on what wikipedia is and is not and cut down on the aggressiveness of attacking it as a website. In the Free Software world that gave wikipedia both ideas and it his software, we have always used a much stronger reputational association. Anyone can edit the wikipedia source code, but those changes don't get committed into anything (no least wikipedia.orgs running service) without serious review. There's no factual checking required for source code so perhaps it's an easier job, but at this point of wikipedia's fortunes they should have switched to better controls.

Having said that. I would rather see a thousand wikipedias in it's current state than the pox of corporate services that will only lead to bad things as they take away from a computer user's freedom.
DarkSaxeBleu Featured By Owner Aug 15, 2014  Hobbyist General Artist
It is not all necessarily bad information - if one bothers and checks it out through other sources and compares it with further information. But it's good if one stumbles across it on a free site such as Wikipedia. Citing of which may be a problem because the entries change/get deleted frequently ;)

Yes, website is being abused and Wiki people do not always hit it 100% - but, as you surely are aware of,

neither do the published academical papers and books which claim to be the Authority on certain subjects:

much of the material can be a subjective opinion which gets taken for granted and just cited from book to book -

without anybody checking the information at the source or without exposing so served ''facts'' as problematic.

You'll find loads of these ''private idealised invasions'' especially in studies/biographies - besides, pick up a 50 years old Encyclopaedia - and compare it with a current one. Changes are made constantly and are not necessarily more or less true - all of it is simply a record of how a certain thesis is evolving (as well as maybe the motives of its presenters. A pretty interesting study on its own!)
PeteriDish Featured By Owner Aug 15, 2014  Hobbyist General Artist
this is so true! and not only at a university, same stuff when I was at highschool as well XD
GhostInThePines Featured By Owner Aug 14, 2014  Hobbyist Photographer
It's sad that professionals in the scientific, medical, and academic communities are now citing Wikipedia. It's just compounding an already disturbing problem in the research communities of distributing bad information and misinformation.

Unfortunately, Wikipedia isn't the only problem site among information sources... there's also a growing issue among open-access publishing sites too!…

I wrote a brief journal about this back in March when The Wildlife Society's newsletter ran the article from the link above.
ZoPteryx Featured By Owner Aug 14, 2014  Hobbyist General Artist
And just when I thought it was safe to go back to the doctor!  I can understand using Wikipedia as a quick way of finding references, but I'd always be sure to quote from the reference/paper itself, not the Wikipedia summary someone else wrote.
JohnRaptor Featured By Owner Aug 14, 2014   General Artist
Wow. I had no idea doctors and serious academics were citing Wikipedia. That's, disheartening.
Add a Comment:

:icondomain-of-darwin: More from Domain-of-Darwin

More from DeviantArt


Submitted on
August 14, 2014


8,540 (1 today)
1 (who?)