On the causes of subject-specific citation rates in Web of Science

It is well known in bibliometrics that the average number of citations per paper differs greatly between the various disciplines. The differing citation culture (in particular the different average number of references per paper and thereby the different probability of being cited) is widely seen as the cause of this variation. Based on all Web of Science (WoS) records published in 1990, 1995, 2000, 2005, and 2010 we demonstrate that almost all disciplines show similar numbers of references in the appendices of their papers. Our results suggest that the average citation rate is far more influenced by the extent to which the papers (cited as references) are included in WoS as linked database records. For example, the comparatively low citation rates in the humanities are not at all the result of a lower average number of references per paper but are caused by the low fraction of linked references which refer to papers published in the core journals covered by WoS.