How untrue “facts” propagate in the scientific literature
August 5, 2014
Short post today. Go and read this paper: Academic urban legends (Rekdal 2014). It’s open access, and an easy and fascinating read. It unfolds a tale of good intentions gone wrong, a chain of failure, illustrating an important single crucial point of academic behaviour: read what you cite.
References
Posted by Mike Taylor
Filed in credit where it's due, paleontologists behaving badly, stinkin' authors
13 Comments » Filed in credit where it's due, paleontologists behaving badly, stinkin' authors
13 Responses to “How untrue “facts” propagate in the scientific literature”
Leave a Reply
This site uses Akismet to reduce spam. Learn how your comment data is processed.
Sauropod Vertebra Picture of the Week 
August 5, 2014 at 8:50 am
I recently posted on RTFS (Read The Fu.king Source) a short while ago at dinosaurpalaeo. It is a real problem that neither authors nor reviewers relly check them.
August 5, 2014 at 8:59 am
Here’s the post Heinrich is referring to.
August 5, 2014 at 9:36 am
thanks, Mike! I haven’t mastered c&p on my cellphone yet.
August 5, 2014 at 9:37 am
I can’t help but feel like there’s a hidden test of character in that paper. Some unsubstantiated claim or inaccurate citation somewhere we’re supposed to notice…
August 5, 2014 at 9:41 am
I also had wondered whether someone would swiftly take Rekdal to task for an unsubstantiated citation; but it hadn’t occurred to me that he might have planted such a thing as a deliberate test of readers! Sneaky!
August 5, 2014 at 10:32 pm
Good thing I don’t study spinach! :-D
August 6, 2014 at 10:41 am
Thanks Mike, an entertaining read.
August 6, 2014 at 5:41 pm
Great article! I especially like this biting sentence: “If we are concerned about the consequences of ‘low-quality research’, ‘junk science’, or the prevalence of academic urban legends, there are good reasons to shift the focus away from uncited and ignored publications, and worry instead about those that are cited, but which should not have been published.”
August 14, 2014 at 1:03 am
On a related note, I was reading over a graduate student’s manuscript pre-submission (This happens a lot – I check out other people’s stats, and have a reputation for a mean first paragraph.) and remarked that they had not cited the foundational literature where appropriate, instead citing a later (in some cases, century-later!) paper that itself cited the original source. The student told me that they _had_ cited the original sources, but their advisor told them to take them out and replace with something with a more recent date. Tricky situation, since the advisor is co-author.
I sent this student the spinach article.
The student was horrified: is anemic, and the doctor had recommended spinach!
So, between the physician and the advisor, this student is totally f***ed, and learns this all in one paper!
All I can say is, with the younger authors, please remember it isn’t always about the author’s *character.* Some of them seem to be poorly advised, and/or powerless to argue. I recommend simply correcting these types of error, and leaving out assumptions and moralizing.
(This is not a paleo lab I am telling you about. Paleo labs are SO egalitarian compared to other disciplines!)
August 14, 2014 at 7:46 am
“The student told me that they had cited the original sources, but their advisor told them to take them out and replace with something with a more recent date.”
Ouch! Thanks for this sobering anecdote. And for describing palaeontologists as egalitarian :-)
August 14, 2014 at 9:51 am
I would, and have, cited the original work, and then added “see (blah) for a recent account”. The long original paper in question is from the 60s published in a book series, and someone extracted the basic definitions and explained them concisely with elementary examples and put it on the arXiv in 1998. Since the details are fiddly, it’s always worth giving not just due credit (and the original author is very sensitive about this), but a reference that is available to all.
August 14, 2014 at 9:52 am
Thanks, David, that seems like precisely the right solution. I will adopt it henceforth.
October 2, 2014 at 3:49 pm
for “assuming-dinosaur” and “Mike Taylor”:
There is a trap in there. If you read the abstract, and skip the words “appears to have” you will get a message that is diametrically opposite to what I say in the article itself: That the decimal point error probably never occurred.
There is a large number of tweets and comments about the “Academic Urban Legends” article out there, and quite a few of them, despite my sincere effort to contribute to debunk the urban legend about the decimal point error, actually do the opposite: They retell the “fact” of the decimal point error that made the world eat more spinach – with a link to my SSS article.
These tweets are written by people who have quickly and superficially read the abstract, and not the article itself, and they can serve as good examples of yet another academic shortcut that can help spread academic urban legends: basing ourselves on superficial reading of abstracts/titles and not the content of the article itself.
Perhaps I should have written the abstract somewhat differently, but to be honest, I am glad I didn’t. There are so many layers of irony about iron in spinach. Why not add another, and equally revealing one?