PLOS ONE is poised to lose its claim as the largest journal in the world soon, with rival journal, Scientific Reports, taking its place.
Scientific Reports, published by Springer Nature was a relative latecomer to the scene, and, at the time of its launch, many skeptics did not believe that there would be room for yet another multidisciplinary open access journal. Both Scientific Reports and PLOS ONE are run by an immense, diffuse group of section editors, accept papers without novel findings, and charge the same article processing charge (APC) — $1,495. A trader on Wall Street would claim that these journals were providing fungible commodity services.
But, are these journals really interchangeable? If so, what explains Scientific Reports growth at the apparent expense of PLOS ONE?
- Journal Impact Factor.
- Data Availability Policies
- Publication Delay
Journal Impact Factor (JIF). In the last (2015) release, Scientific Reports received a score of 5.228 compared to 3.057 for PLOS ONE. In addition, Scientific Reports has maintained a score above 5 for the last three consecutive years whereas PLOS ONE‘s JIF has declined. While many editors and publishers eschew the indicator, scientists consider the JIF as an important factor when deciding where to submit their work, according to Nature Publishing Group’s 2015 Author Insights survey as well as the recent University of California’s Pay It Forward report.
Data Availability Policies. In 2014, stronger data availability policies were enacted across all PLOS journals, requiring authors to make “all data underlying the findings described in their manuscript fully available without restriction, with rare exception.” All PLOS manuscripts must include a data availability statement and authors are strongly encouraged to make their data available in a public archive before publication. In contrast, Scientific Reports’ policy merely states that authors should share upon request. Research by Tim Vines has shown that the willingness and ability for authors to share data declines significantly with time, and having a weak data availability policy is only marginally better than having no policy at all. While advocates of open science argue for full data transparency, there is considerable resistance from the biomedical community, and freedom from a strong editorial policy may provide Scientific Reports with a competitive advantage.
Publication Delay. Anecdotally, colleagues told me that it was faster to get a paper published in Scientific Reports than PLOS ONE. To confirm this, I took the last 100 papers published in each journal and calculated the time from submission to acceptance, acceptance to publication, as well as the full publication process. Time to acceptance was about a month shorter for Scientific Reports, (median days = 99 [IRQ: 78–146]) than it was for PLOS ONE (132 days [IRQ: 100–173]). Although once accepted, PLOS ONE was able to publish the papers about a week sooner (19 days [IRQ: 14–23]) than Scientific Reports (27 days [IRQ: 25–28]). In all, the median time from submission to publication was 126 days for Scientific Reports and 151 days for PLOS ONE, a difference of about three weeks.
According to the 2015 Author Insights Survey, Chinese authors rate the JIF and Time (both time to acceptance and time to publication) as more important than other respondents. This is not surprising, as many institutions reward Chinese authors with monetary rewards for publishing in high impact journals, and publication numbers confirm these institutional rewards. For papers published since 2015, 39% of Scientific Reports authors were from China, compared to 18% for PLOS ONE (source: Web of Science).
In sum, given two large multidisciplinary open access journals with similar editorial structure and publication cost, authors appear to be favoring the one with the higher Impact Factor, faster publication time, and more lenient data availability policies.
As I’ve argued in the past, drawing vast numbers of submissions is not always the wisest business model for longterm success. PLOS ONE drew record submissions immediately after they received their first JIF. If the kind of author seeking publication in a megajournal is primarily motivated by the JIF, then the same downward turn observed with PLOS ONE is likely to happen with Scientific Reports as well. Given that the JIF reflects the publication preferences of authors publishing two and three years ago, we may see waves of authors oscillating between these two journals every few years.
An unpredictable publication flow and revenue stream through APCs will have very different effects on the two publishers. Springer Nature has an enormous, diversified stable of journals and revenue streams, which allows them to play a long-term strategy game with Scientific Reports. Annual revenue fluctuations with one journal are not going to put Springer Nature in financial trouble. In contrast, PLOS’ income is almost exclusively based on APC revenue, with 97% of their 2014 revenue coming from publication fees. More importantly, 91% of all 2015 papers published in PLOS journals were published in PLOS ONE, the remaining 9% split among six other journals. As revenue from PLOS ONE functions to subsidize the publication costs of these six other titles, downward pressure on PLOS ONE puts the entire organization at risk.
Given that the leading edge of the above publication graph reflects papers submitted in the winter and spring of 2016, the influence of the last Journal Impact Factor scores should start revealing themselves later this fall.