Impact factor – benefits and limitations

Abstract

Every summer Thomson Scientific's Journal Citation Reports (JCR) announces their impact factor lists covering more than 8000 (in 2013–8474) of the world's most highly cited, peer-reviewed journals in about 200 fields of science. Although the first scientific journal, the Philosophical Transactions of the Royal Society of London, was founded in 1665, the significant increase in number of scientific journals first became an issue in the 20th century. The concept of counting references to rank scientific journals was proposed in 1927 (McGarty 2000). Then, Garfield suggested that reference counting could measure 'impact', and the term 'impact factor' was introduced in 1963. In the next year, Garfield introduced the first Science Citation Index as a five-volume print edition indexing 613 journals and 1.4 million citations. (Williams 2007) At that time, impact factor analysis was to help in the management of library journal collections in order to buy the most cost-effective journals. Another application was the correlation data on citation frequency and impact with subscription costs (Grzybowski 2009). Until today the number of scientific journals aroused enormously, estimated to 126,000 in 1996 (Ha et al. 2006). It is quite natural that in such plenty of journals, there appeared the need for an indicator of journal scientific performance. Journal Citation Reports (JCR) have grown from 600 in 1964 to 2400 journals in 1972, and every year the number of journals indexed in the database is enriched by 200 new. Impact factor is calculated from the formula IF = A/B, where the numerator A is the total number of citations of the journal articles in the previous two years for the year in consideration; the denominator B is the total number of items published in the same two years as in A. It is quite obvious that impact factor is directly related to citations. The more citations the journal have, the higher is the impact factor value (Grzybowski 2009). The possible range of journal's impact factor values is closely related to the field of research. Journals oriented into basic science subjects, like molecular biology (Nature Reviews Molecular Cell Biology, IF 36.458), immunology (Annual Review of Immunology, IF 41.392), physics (Reviews of Modern Physics, IF 42.860), or biochemistry (Annual Review of Biochemistry, IF 26.534) have much more citations than clinical subjects, like ophthalmology and dermatology (Table 1.). In the group of top ten 'impact factor' journals for 2013, there are only three journals oriented to clinical medicine, CA-Cancer Journal for Clinicians (IF 162.500), New England Journal of Medicine (IF 54.420), and Lancet (IF 39.207). Moreover, the impact factor is directly related to the scope of the journal – the larger it is the higher is impact factor value. This means that even the best journals of more specialized fields, like ophthalmology (the highest impact factor in this group for 2013 has Progress in Retinal and Eye Research with 9.897) never achieve the impact factor of general medicine journals, like New England Journal of Medicine (IF 54.420) or Lancet (IF 39.207). The more specialized reviews of impact factor calculations and analysis might be found elsewhere. (Garfield 1972, 1996; Seglen 1997; Grzybowski 2009). It should be noted that calculations of impact factor might be biased (Seglen 1997; Brown 2007; Grzybowski 2009). Although it is mostly related to articles citations, other factors include number of scholarly records published in the journal. Some journals increased the impact factor value by reducing the number of these items. It is interesting to note that the highest impact factor journal in 2013 CA-Cancer Journal for Clinicians (IF 162.500), published in 2011 only 19 articles with the citations number of 4 228. On the other hand, the most cited journal in the same period was Nature with 37 642 citations, and only 5th position in impact factor ranking (IF 42.351). This was due to the large number of 841 published articles in the same time. It is also known that citations of journal articles differ significantly. An analysis of 7,528 journals has shown that as few as 300 journals account for more than 50% of the total number of citations (Testa 2012). It means that the impact factor of the journal is driven by a small number of highly cited studies (Seglen 1997). It was also shown that for about 38 million source items about half were not cited at all. On the other hand, it is well known that most cited are methodology articles with the Lowry's article on protein measurement from 1951 with nearly 300,000 hits. (Brown 2007) Thus, the journal impact factor should not be used as an indicator of individual article or researcher value. Due to its simplicity and convenience, the journal impact factor is and probably will be in the nearest future the most popular parameter of journal quality. However, as it was shown above, it might be greatly influenced by the type of the article (review vs original article), subject specialty and the sort of research field. It should be also remembered that the impact factor can be used only to evaluate journals, not individual articles.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 10 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.