Read a recent update to this piece: PLOS ONE Output Drops Again In 2016.
PLOS ONE’s 2015 Impact Factor is expected to rise, the result of its shrinking size. As reported earlier this year, the open access mega-journal has experienced two successive declines in article output, from a peak of 31,509 research papers in 2013 to 28,107 in 2015–a reduction of 3,402 papers or 11%.
Published each summer by Thomson Reuters, the Journal Citation Reports (JCR) includes various annual journal-level performance measures. The indicator most eagerly awaited is the Journal Impact Factor — a simple arithmetic average of the citation performance of a journal’s articles. Editors and publishers anxiously await this measure as it affects their relative ranking within their discipline. Authors, in turn, are heavily influenced by journal Impact Factors (often used to describe a journal’s prestige) as a guide for where to submit their manuscripts. While PLOS does not support the use of the Impact Factor, their authors seem just as interested in this metric as everyone else.
The Journal Impact Factor is constructed by summing all citations made to papers published in the previous two years and dividing by the number of papers published in those two years. There is a little more nuance to this approach, as the number of papers in the denominator is limited to research articles and reviews — what Thomson Reuters calls “citable items.” Whereas some journals publish many papers that are not considered citable (e.g. editorials, perspectives, commentary, letters, etc.), this is not the case for PLOS ONE. This will help simplify our example of how size affects Impact Factor calculations. In a future blog post, I will revisit and provide more detail on the classification and counting of citable items.
Last year, PLOS ONE received an Impact Factor of 3.234, which was the result of dividing 177,706 citations made in 2014 to 54,945 papers published in 2012 and 2013. We can further dissect the journal’s IF score into two separate parts, measuring independently how papers published in 2012 compare with papers published in 2013. For 2012 papers, there were 95,631 citations made to 23,447 papers for an effective Impact Factor of 4.079; for 2013, there were 82,075 citations made to 31,498 papers for an effective Impact Factor of 2.606. The reason why there is such a large performance difference between these two annual cohorts is that 2012 papers were measured in their third year of publication while 2013 papers were measured in their second.
The Impact Factor of a shrinking journal is therefore based upon the performances of more older papers, which tends to create an upward influence on its IF calculation. Conversely, a growing journal is evaluated on the performances of more younger papers, which tends to create a downward influence on its IF calculation.
What can we expect from PLOS ONE‘s next Impact Factor?
If we assume that PLOS ONE papers published in 2014 have a similar citation profile to papers evaluated last year, then we can expect its 2015 Impact Factor to rise marginally, from 3.234 to 3.360 or by about 4%. It is very difficult to get an actual performance reading, as the Web of Science (the index from which the JCR citation metrics are derived) limits citation reports to 10,000 records–just one-third of the papers published in PLOS ONE each year.
I should note that the above calculation ignores changes to the journal over the past year (data availability policy, publication fee increase, competition from other open access journals, platform upgrades and publication delays) that may change the kinds of papers submitted to PLOS ONE. Taken together, these changes may countervail the rise we expect from journal shrinkage.