Editing Impact factor (section)

===Distribution skewness===
[[File:Journal impact factor Nature Plos One.png |thumb |440x440px |Journal impact factors are influenced heavily by a small number of highly cited papers. Most papers published in 2013–14 received many fewer citations than indicated by the impact factor. Two journals ([[Nature (journal)|''Nature'']] [blue] and [[PLOS One|''PLOS ONE'']] [orange]) are shown to represent a highly cited and less cited journal, respectively. The high citation impact of ''Nature'' is derived from relatively few highly cited papers. Modified after Callaway 2016.<ref name="Callaway_2016">{{cite journal |vauthors=Callaway E |title=Beat it, impact factor! Publishing elite turns against controversial metric |journal=Nature |volume=535 |issue=7611 |pages=210–1 |date=July 2016 |pmid=27411614 |doi=10.1038/nature.2016.20224 |bibcode=2016Natur.535..210C |doi-access=free}}</ref>]]

Because citation counts have highly [[skewed distribution]]s,<ref name="Beat it, impact factor! Publishing"/> the mean number of citations is potentially misleading if used to gauge the typical impact of articles in the journal rather than the overall impact of the journal itself.<ref>{{cite news |title=Citation Statistics |author=Joint Committee on Quantitative Assessment of Research |publisher=International Mathematical Union |date=12 June 2008 |url=http://www.mathunion.org/fileadmin/IMU/Report/CitationStatistics.pdf}}</ref> For example, about 90% of ''[[Nature (journal)|Nature]]''{{'}}s 2004 impact factor was based on only a quarter of its publications. Thus the actual number of citations for a single article in the journal is in most cases much lower than the mean number of citations across articles.<ref>{{cite journal |vauthors= |title=Not-so-deep impact |journal=Nature |volume=435 |issue=7045 |pages=1003–4 |date=June 2005 |pmid=15973362 |doi=10.1038/4351003b |doi-access=free}}</ref> Furthermore, the strength of the relationship between impact factors of journals and the citation rates of the papers therein has been steadily decreasing since articles began to be available digitally.<ref>{{cite journal |doi=10.1002/asi.22731 |title=The weakening relationship between the impact factor and papers' citations in the digital age |year=2012 |vauthors=Lozano GA, Larivière V, Gingras Y |journal=Journal of the American Society for Information Science and Technology |volume=63 |issue=11 |pages=2140–2145 |arxiv=1205.4328 |bibcode=2012arXiv1205.4328L |s2cid=492922}}</ref>

The effect of outliers can be seen in the case of the article "A short history of SHELX", which included this sentence: "This paper could serve as a general literature citation when one or more of the open-source SHELX programs (and the Bruker AXS version SHELXTL) are employed in the course of a crystal-structure determination". This article received more than 6,600 citations. As a consequence, the impact factor of the journal ''[[Acta Crystallographica]] Section&nbsp;A'' rose from 2.051 in 2008 to 49.926 in 2009, more than ''[[Nature (journal)|Nature]]'' (at 31.434) and ''[[Science (journal)|Science]]'' (at 28.103).<ref>{{cite news |date=21 June 2010 |title=New impact factors yield surprises |url=https://www.the-scientist.com/new-impact-factors-yield-surprises-43246 |archive-url= |archive-date= |access-date=23 October 2024 |work=The Scientist |vauthors=Grant B}}</ref> The second-most cited article in ''Acta Crystallographica Section&nbsp;A'' in 2008 had only 28 citations.<ref>{{cite web |date=17 June 2010 |title=What does it mean to be #2 in Impact? |url=http://community.thomsonreuters.com/t5/Citation-Impact-Center/What-does-it-mean-to-be-2-in-Impact/ba-p/11386 |archive-url=https://web.archive.org/web/20171012192240/http://community.thomsonreuters.com/t5/Citation-Impact-Center/What-does-it-mean-to-be-2-in-Impact/ba-p/11386 |archive-date=12 October 2017 |access-date=2018-07-16 |vauthors=McVeigh M}}</ref>

Critics of the JIF state that use of the [[arithmetic mean]] in its calculation is problematic because the pattern of citation distribution is skewed<ref name="Lariviere 2016">{{cite bioRxiv |title=A Simple Proposal for the Publication of Journal Citation Distributions |biorxiv=10.1101/062109 |year=2016 |vauthors=Larivière V, Kiermer V, MacCallum CJ, McNutt M, Patterson M, Pulverer B, Swaminathan S, Taylor S, Curry S |display-authors=6}}</ref> and citation distributions metrics have been proposed as an alternative to impact factors.<ref>{{cite web |vauthors=Kiermer V |title=Measuring Up: Impact Factors Do Not Reflect Article Citation Rates |publisher=[[PLOS]] |year=2016 |url=http://blogs.plos.org/plos/2016/07/impact-factors-do-not-reflect-citation-rates/}}</ref><ref>{{cite web |url=http://www.the-scientist.com/?articles.view/articleNo/46495/title/Ditching-Impact-Factors-for-Deeper-Data/ |title=Ditching Impact Factors for Deeper Data |work=The Scientist |access-date=2016-07-29}}</ref><ref>{{cite journal |vauthors=Corneliussen S |title=Bad summer for the journal impact factor |journal=[[Physics Today]] |date=2016 |issue=8 |page=22472 |doi=10.1063/PT.5.8183|bibcode=2016PhT..2016h2472C }}</ref>

However, there have also been pleas to take a more nuanced approach to judging the distribution skewness of the impact factor. Ludo Waltman and Vincent Antonio Traag, in their 2021 paper, ran numerous simulations and concluded that "statistical objections against the use of the IF at the level of individual articles are not convincing", and that "the IF may be a more accurate indicator of the value of an article than the number of citations of the article".<ref name="Waltman 366"/>