Misuse and abuse of journal "impact factors"

The scientific journal is the technical indicator most widely used for evaluating scientific journals. The impact factor was set up by the Institute of Scientific Information (ISI) in the 1960s, and it measures the frequency with which individual articles - the average article - in a journal are cited over a period, usually a year. It is calculated by dividing the number of all citations of articles published in a particular journal during the current year by the articles published in that same journal during the previous two years. Since 1975, ISI has published Journal Citation Reports annually as part of the Science Citation Index. At first, ISI used the impact factor to collect statistical data to evaluate and compare the journals indexed, so that librarians had a useful tool for the management of library journal collections [1]. The use of the impact factor, however, has gone far beyond that aim. Besides playing a major role in market research for both publishers and advertisers, the impact factor has been more and more used by science managers and decision-makers as an indicator of the individual scientific value of researchers in science policy. Research funding is frequently influenced by the impact factors of the journals in which applicants for grants have published their articles. Another publishing-based way to accredit knowledge is by measuring the frequency with which articles are cited. The number of citations of a scientist's work has even been recommended as the best single indicator of scholarly recognition [2]. A reference cited in a paper makes an explicit linkage between the research described in the paper and previous work that is a source of information for the author [3]. The citation impact of a given paper implies that the authors citing it have found the paper useful for their work. These conceptual associations have been called intellectual transactions, and they are indexed in ISI's databases. The impact factor and citation impact are closely related. In fact, a paper's citation impact may depend on the impact factor of the journal where it has been published [2]. Scientific journals and science policy