Is PLOS ONE losing favour with palaeontologists?
Is PLOS ONE falling out of favour with Palaeontologists?
By looking at a simple plot of publications through time, it would seem that the zenith of PLOS ONE has been reached after a surge in increase as it dominated the market, but is now in sharp decline. There have been a couple of such blips, with subsequent recovery phases, but the overall trend would seem to be one of a decline since 2014.
Here’s the simple R code I used to do create this plot:
# Using R to visualise Paleontology publications in PLOS ONE through time
# Install package from within R
install.packages("rplos")
# Load package from library
library("rplos")
# Search for the term 'paleontology'
searchplos('paleontology', 'id,publication_date', limit = 10000)
# So there's a lot of data..
# Let's try plotting it!
# Plot results through time, no limit on data
plot_throughtime(terms='paleontology',limit=10000)
# You can try this with any field really
plot_throughtime(terms='political science',limit=3000)
Well, I’m going to guess that this trend it’s because of the rise of Nature Communications and it’s little cousin, Scientific Reports, as well as PeerJ. Or an expansion of the ‘megajournl’ market.
Now, the research published in any of these journals is objectively no different than if it were published in PLOS ONE, or any other venue. But it’s a marketing thing, isn’t it. PLOS ONE has a bit of a reputation of a sort of ‘dustbin’ for research, publishing anything and everything that’s scientifically sound. For some reason, many equate this as the peer review being less rigorous (“peer review lite”..), whereas its proponents will simply state that it is more objective, as it doesn’t do weird things like trying to assess research impact before it’s had a chance to even make an impact.
PLOS ONE also doesn’t exactly have the greatest reputation after a couple of Editorial ‘SNAFUs‘, and issues with the publication process (e.g., it still doesn’t offer page proofs). I will say that while many journals suffer from these sorts of issues, PLOS ONE definitely gets a skewed amount of researcher angst when a mistake is revealed, for some reason. I don’t know why this is, but all you have to do is look at the relative amount of social media ranting when a PLOS ONE paper gets retracted compared to when it happens in Nature.
If this seeming trend of decreasing PLOS ONE papers were a cost thing, then all of the research being lost from PLOS ONE ($1495 per article) would be going to PeerJ ($1095), depending on membership fees ($399/author for 1 paper/year for life), or another cheaper venue (many Palaeo journals such as APP or Palaeo Electronica are free to publish in and still Open Access). Both Nature Communications ($5200) and Scientific Reports ($1675) largely do the same thing editorially as PLOS ONE too, although the former is more selective, but have the advantage of having the Nature brand attached to it. And many researchers are suckers for brands, forgetting it is the value of their work which builds the brand of a journal, not the journal conferring some sort of quality mark on their research.
But I don’t think we’ll see that PeerJ or these other cheaper/free venues are exclusively ‘snaffling’ PLOS ONE’s authors in the data, as we’re still seeing many Palaeo papers appearing in a whole range of venues. And this is because the cost of publishing doesn’t matter to researchers – most of the time it doesn’t make a difference if it costs $100 or $5000 – they either can or cannot afford it, and for those who can, they are rarely accountable for the funds which they draw upon to do so.
I’ve even got two papers in Nature Comms, one in PeerJ, and one in PLOS ONE. In my experience, the entire publication process was no different between either journal, except that the former was around 4 times the cost per article than PLOS ONE (we got a fee waiver for PeerJ because they’re awesome).
So, I don’t really know what to make of this so far. It seems like the ‘megajournal’ market is well and truly open (no pun intended), and I would hazard a guess that researchers care more about journal reputation than the cost of submission, because generally our incentive, financial [in]equality, and accountability systems in research are all pretty messed up.
I’ll keep tracking the data and update this blog whenever something interesting happens. Incidentally, is it possible to get the same data here for Nature titles and PeerJ? Would loooove to do that comparison!
You say that Nature Communications “largely [does] the same thing editorially as PLOS ONE too” but this isn’t the case. It’s website says “Papers published by the journal represent important advances of significance to specialists within each field.” It is selective, not a megajournal like PLOS One, Scientific Reports and PeerJ, do shouldn’t be lumped together with them.
Oops, good spot Anna! Will modify that.
PLOS ONE are losing out for two main reasons:
– They are expensive relative to the quality of service. The lack-of-proofs issue and the *very high* frequency with which significant errors appear in published papers as a result has left many people very disillusioned with them. Why publish in PLOS ONE with no proofs and a load of mistakes that aren’t your fault in the final manuscript for $1495 when you can publish in PeerJ, Scientific Reports or Royal Society Open Science for similar or less money (or nothing in the case of RSOS), and get page proofs?
– There have been some very high-profile problems with the editorial process, particularly with their process for assigning appropriate editors to manuscripts, and with fee waivers (including the notorious period of time in which people were asked to provide personal bank information as evidence of inability to pay). These issues have tarnished their brand seriously in a way that hasn’t happened yet to PeerJ, Scientific Reports or Royal Society Open Science.
I don’t think that it is any more helpful to your career to have a paper in Scientific Reports, PeerJ or RSOS than PLOS ONE (Nature Communications with its selective editorial policy and high impact factor is a different kettle of fish). I think it is simply that people are hacked off with the way PLOS ONE operates and have decided to go elsewhere. I haven’t published there since 2014, having previously published seven papers there, and doubt that I will again. My research group doesn’t publish there either.
Hey Richard,
Thanks for chipping in on this one, appreciated as always!
My main thing here was to have some numbers attached to this and measure the scope of any potential problem. Anecdote is fine, but every journal out there has issues of some sort. See Gabi’s recent fiasco with RSOS, for example. Every day on Twitter I see some journal messing up somehow, like authors being invited to asked to review their own papers etc. We could discuss these issues across the board until the cows came home to little effect. The scale of this at PLOS ONE is perhaps higher, or perhaps just discussed more in public, as part of a snowballing reputation decline, as you mention. Either way, I’m interested in seeing what the effect this is having on venue choice for authors. Also I’m hoping that Joerg Heber, who used to be the EiC for Nature Comms but is now at PLOS ONE, will whip the journal back into shape!
Hope all is well in Brum!
Hi Jon,
It’s true that my statements above are based on anecdote. But I have worked as a volunteer editor for both PLOS ONE and Scientific Reports, as well as specialist journals such as JVP, Zootaxa and Pal Zeitschrift. As you know, I’m particularly interested in PLOS ONE. I worked as a volunteer editor for them for two years, handling around 60 manuscripts, and was very invested in them – I believed in what they were doing, and wanted to see them succeed. I published seven of my own papers there, including some of my best anatomical work. I invested significant amounts of time and energy in PLOS ONE, but resigned in early 2014 because I was extremely disillusioned with the direction of change and the way the journal was being run. I had numerous interactions with authors as an editor, and have had many conversations about this since.
It is true that pretty much all other journals have problems. Some of PLOS ONE’s stem from the unique challenges of being a megajournal, and I saw similar problems at Scientific Reports. For example, it is very easy for a specialist journal to assign appropriate editors to manuscripts. When as a megajournal you handle vast numbers of manuscripts spanning a huge diversity of subjects it is difficult to identify appropriate editors for all of them. When I was on the editorial board PLOS ONE were not very good at this, and I frequently got asked to handle very inappropriate manuscripts (e.g. a manuscript on homicide rates and temperature in Chicago), but I saw similar issues at Scientific Reports. This is one of the reasons that I decided I was no longer going to edit for megajournals, and now only volunteer my time to JVP.
The proofing issue is a big one for PLOS ONE. It is true that Gabi had very serious issues recently at RSOS (I was a co-author on the paper), BUT she could resolve nearly all of them by going through several rounds of proofs. As editor at PLOS ONE I saw similar issues happen during the typesetting process, but the authors had no chance to correct them, and it often took PLOS ONE months after publication to fix issues, if they were willing to do so at all.
I agree that editorial lapses at PLOS ONE tend to get a lot of attention on social media, probably more than deserved in some cases. But the flip side is that I see open access advocates are much more willing to make excuses for PLOS ONE when such editorial mistakes happen than they are for other journals. It cuts both ways.
I find it interesting in your data that the decline in submissions to PLOS ONE appears to begin in 2014. Coincidentally, I resigned from the editorial board in February 2014, as mentioned above, because I disagreed with the direction of travel of the journal and believed that it was damaging its reputation in the community. For interest, below are some of the points that I made at the time in my resignation letter. They are essentially the same points as I have made above:
“I am resigning because of my ongoing frustration with what I perceive to be a gradual and ongoing decline in the services offered to authors and editors by PLOS ONE. I will summarise here some of the problems that have led me to decide to focus my editorial contributions elsewhere…
(1) Author proofing and manuscript production. For the “no proofs” model to work the production of final manuscripts needs to be highly efficient and quick to fix any problems appearing in the final manuscript. Unfortunately, I have repeatedly encountered serious problems with PLOS ONE inserting major formatting errors after manuscripts were accepted, or reproducing figures at incorrect sizes. These errors are often quite serious, make authors (justifiably) angry, and are alienating substantial segments of the palaeontological community. As both editor and as an author I have found PLOS ONE unhelpful, unwilling and slow in correcting such errors, and this further compounds the resentment felt by authors. Given the relatively high APC (higher than that of some other journals that do offer proofing), I do not understand why an author proofing service is not available.
(2) Fee waivers. There appears to be an ongoing attempt to make it ever more difficult for authors to request and obtain fee waivers, even when they clearly state that they have no access to funds. One of my PhD students recently had an unfortunate interaction with PLOS ONE with regard to this (which I have already discussed with the Editorial Office and PLOS senior management and have now resolved), but anecdotally it has been suggested to me by colleagues that he is not alone in this experience. Obviously it is important that enough fees are collected for the journal to run, but the fee waiver is critical to supporting junior researchers and researchers in institutions and countries without access to APC funds.
(3) Assignments of manuscripts to editors. I have registered my frustration elsewhere with the extremely inappropriate invitations that I have received as an editor, and the sheer volume of invitations sent out. The former is a very serious issue that is very damaging to the scientific credibility of PLOS ONE. Back in August last year, I noticed a large increase not only in the number of editorial invitations but in the number of utterly inappropriate editorial invitations… I would note that I have very specific and quite narrow keywords linked to my editorial account… I was shocked by the manuscripts that I was being asked to edit, but I also did not feel that my feedback of what to me appears to be a very major problem was treated sufficiently seriously.”
Might be interesting to normalize this against growth in the field or use of a term in general. You should be able to use rcrossref get counts for for a much broader swath of publications, though it will only search metadata fields.
That’s not a bad idea. Any more advice on how to do this for a noob?
> # You can try this with any field really
Interestingly, the general pattern and timing of decline seems to hold with almost any search term:
plot_throughtime(terms=’ecology’,limit=10000)
plot_throughtime(terms=’sex’,limit=10000)
plot_throughtime(terms=’manufacturing’,limit=10000)
While I haven’t published yet with PloS, seeing that now-retracted creationist hand paper a couple of years back turned me off considering Plos.
The papers with the highest retraction rates include Nature, Cell, and Science. Will you concede from publishing with them (and their associated publishers) in the future too? Retraction is so common to journals, that you’d basically have to boycott all publishers rather than single any specific one out.
And that whole issue was sour. It was more like a retraction due to Twitter outrage, that was probably due to language/cultural differences and could have been resolved with a simple corrigendum, rather than any publication of ‘legitimate’ Creationist research. The authors attempted to explain this in the comments of the article, and PLOS handled the entire situation was handled terribly. That being said, isn’t the ‘Arsenic Life’ paper published at Science still not retracted after like 10 years? 😉
I just don’t see singling out any individual publishers that otherwise do/try to do a good job at publishing useful, especially in a climate when there are so many substantially worse issues with other scholarly publishers.
This comment got a bit long, sorry!
Very interesting thread. I’m not enamored with electronic megajournals and haven’t published in one yet, mainly because they all seem to do single column pages. That might seem trivial, but it’s much easier to read narrow (to a point) columns than wide columns.
PLOS ONE, at least, used to format two columns but stopped around 2015. Why? Because it’s cheaper to format the pages if they’re single column. In fact, it’s darn cheap because there IS no formatting! It’s just a text stream of a given width, with periodic figures. The problem is, it’s harder to read than narrower columns. Plus, if you print the PDF, which I do sometimes, it’s even harder to read and a big waste of space due to wiiiiiide margins on each side. In fact, the wide margins give away the problem with single-column: if the text is full-page width, it gets *really” unpleasant to read.
There are other forums with people claiming that one column is better than two (well, I saw one such forum at least). Maybe those people never print a PDF, or maybe they have better monitors than I do, or maybe they never actually read a paper–maybe they just look at the pretty pictures. But the one and only reason for switching to single column instead of two columns is: it’s much easier and thus cheaper to format.
Btw, Royal Society Open Science jumped from free to $1500 per article at the start of 2018. That ain’t free.
Thanks for this comment, Paul! Yes, I guess reading style and format is a tricky one. Wouldn’t it be great if there was a way you could switch between them for articles? This post was written in Febrary 2017, and I am not capable of seeing the future when it comes to journal APCs.