By Joshua A Hirsch
Once a year, in almost ritualised fashion, the editors of peer-reviewed journals anxiously await publication of the Thomson Reuters Impact Factor. Journals in the “neuro” space are no exception to this rule.
In 1955, Eugene Garfield proposed the idea of an impact factor to identify the most significant journals in a given field; originally, the measure was intended as an aid to librarians for building and maintaining their journal collections. Impressively, he founded the Institute of Scientific Information (ISI) in Philadelphia, USA prior to receiving his doctorate in Structural Linguistics from the University of Pennsylvania in 1961. ISI’s flagship product, and another Garfield innovation, was the Science Citation Index (SCI), which, at the time, collected publication and citation data on roughly 4,000 journals. Today, the SCI’s online descendant, the Web of Science, covers more than 12,000 journals. In addition to providing current awareness on topics in the journal literature to ISI’s customers, the SCI generated the citation data from which the Impact Factor was (and still is) derived. Since 1975, using these citation figures, Impact Factors have been calculated for the academic journals covered by the former ISI, now officially known as the Intellectual Property & Science business of Thomson Reuters.
In selecting journals for indexing (and, ultimately, for an Impact Factor calculation), specialists in journal coverage at Thomson Reuters follow a methodology of evaluation that has been described at http://sciencewatch.com/articles/connecting-dots-it-all-starts-data. Once a journal is selected, Impact Factors are calculated once per annum for those journals, and the figures are reported in the Thomson Reuters Journal Citation reports.
How is the Impact Factor calculated? A journal’s impact factor represents the average number of citations for each paper that had been published in the prior two years. For example, if each article published in calendar years 2011 and 2012 was cited twice in calendar year 2013, the 2013 Impact Factor of our newly created journal would be two. In reality, the calculation is more complicated than this. For example, the 2013 Impact Factor is published in 2014. Moreover, not all materials printed in a journal are included in the analysis that determines Impact Factor, just articles that are termed “citable items.” (See M E McVeigh, S J Mann, JAMA, 302 (10): 1107-9, 2009.)
Review articles tend to be the most often cited type while case reports are least cited. This can have a number of distinct influences on publications found in journals.
The importance attached to the Impact Factor likely varies by environment. It is by its nature not designed to evaluate the importance of an individual article or author, and Thomson Reuters expressly disavows and actively discourages this practice; however, it is reasonably commonly used that way. Indeed, the Impact Factor is not without controversy. In December of 2012 the San Francisco Declaration on Research Assessment (DORA) took form. In essence, DORA takes issue with the use of the Impact Factor as a method of evaluating the merit of an individual scientist; a role, one could argue, that the Impact Factor was never intended to assume. The Declaration, which was published in May of 2013, has achieved fairly a reasonable measure of support from a broad constituency.
One thing is for certain, 2015 will likely not be different than 2014 or to prior years as editors and publishers of journals “ritually” await publication of the prior year’s Impact Factor.
Joshua A Hirsch is the director of Interventional Neuroradiology/Endovascular Neurosurgery at Massachusetts General Hospital, Boston, USA