The only thing missing is... well... um... the paper.
The Nature News article reporting the results of the (as yet nonexistent) paper even links to the paper's DOI, which looks promising until you click it and find out it doesn't actually lead anywhere. So, like any good Web 2.0 denizen, I left a comment there: "The DOI for the cited article does not appear to exist, nor can I find anything about this paper on the PNAS website. Any suggestions?" The response I got was, "Please note that PNAS releases the news of its upcoming papers on Mondays at 5pm Eastern time, but the papers themselves may go live at any time during that subsequent week. So this paper will appear eventually... check back on Friday and you should find it."
So, I thought, what better to do while I'm waiting for the paper to be published than to pick apart some bad science writing? Oh, there's the usual assortment of inaccuracy and sensationalism, but my complaints here are going to be more focused on the specific content, that is, DNA barcoding.
Scalpel, please.
Here is my dart-board collage of the offending titles and sentences from a selection of three news pieces reporting the key findings of the mystery DNA barcoding paper. I gave up after three when it became clear they were all just copying and pasting from the same (lame) press release. In bold are the bits that need trouncing, followed immediately by the actual trounce.
"'Barcode' of plants mapped" (title, inthenews.co.uk). Mapped? What, with like a road atlas? I don't need to go into detail on why this is just so wrong, simply because Miss Prism has already done it with verve.Needless to say I am looking forward to the paper actually getting published so I can blog it using words a little more informative than "usually", "majority", "a few" and "closely related".
"Scientists have identified a 'barcode' gene that can be used to distinguish between the majority of plant species on Earth." (inthenews.co.uk). Well, actually, according to the news release they tested 1600 species of orchids in the study that is (supposedly) going to be published and yet there are nearly 300,000 recognised plant species. Now, unless you're the US Electoral College, 1600/300,000 does not a majority make.
"As such scientists can use the gene to distinguish between different plants, even closely related species." (inthenews.co.uk). How closely would that be? ...seems like an important question.
"matK Gene Is A "Barcode" DNA For Plants (title, ScientificBlogging)." These guys can't even get the word order right. It's a DNA barcode not a barcode DNA.
"They found that when one plant species was closely related to another, differences were usually detected in the matK DNA. (ScientificBlogging)". Is ScienceBlogging having a secret internal competition on who can be the most vague? "Usually" in this context could mean one of two things for a putative plant DNA barcode: 1) spectacular success, 2) cringe-worthy failure.
The matK gene may not, however, be able to be used to identify every plant species on Earth. In a few groups of species, additional genetic information may be required for species-level identification because hybridization - where species cross-breed and genetic material is rearranged - may confuse the information provided by matK. (ScientificBlogging). Okay, where to begin. First of all, hybridization is not just a minor problem in the plant kingdom, it's more the norm. Second, genetic material is rearranged as a normal result of the normal process of normal sexual reproduction, so hybridization doesn't have anything special going for it there. Third, hybridization does not "confuse the information" but rather the information is inadequate to tell when hybridization has occurred.
The team behind the discovery found that DNA sequences of the gene 'matK' differ among plant species, but are nearly identical in plants of the same species. (Bio-Medicine). Not so much a discovery as a trial of something that had already been proposed and partially tested.
Posts
2 comments:
LOL, amazing. I never paid attention to scientific journalism and never read press releases prior to blogging. My only news was from the news sections of Science and BioScience, 2 I've subscribed to for many years (and national geographic too). Now I read it all the time and the science journalists are consistently getting it wrong. I don't know why more scientists get into journalism. I know sensationalist headliness and talking points bring in the readers, but you still beed to be somewhat accurate!
Seems I'm a tad late to this post, but I want to point out one glaring inaccuracy in your premise: Science journalists and perhaps some bloggers do have access to the full journal papers before they appear online, via embargoed press packets.
As a working editor, I can log in to special sites for Nature, PNAS, Science, etc. and download pdfs of any paper that will be public as of the following week. I just can't report on it until the embargo lifts, which in Nature's case means the Wednesday of the week after I get the paper.
In the age of teh Internets, it seems more journals are releasing papers waaaaay before they plan to run them in the print editions, and therefore they also ease up on how quickly they post their papers to the web. Sorry, but journalists have no control over that.
As for sensationalist headlines, anything from a maximum character count to optimizing your headline for Google searches can affect the exact words chosen. Can you write a catchy but more accurate headline that's 55 characters or less, with spaces?
I do, however, agree that there's nothing worse than reporters parroting a press release, and wildly lament the lapse in critical thinking a 24/7 publishing world has wrought. Especially when we do have advance access to the paper, there's no excuse for not getting the facts right and presenting a fully balanced view.
Anyway, sorry to rant, but I hope this sheds some light on what's actually going on in science journalists' worlds...
Post a Comment