Jointly published by Akad��miai Kiad��, Budapest Scientometrics, and Kluwer Academic Publishers, Dordrecht Vol. 56, No. 2 (2003) 235���246 Received August 12, 2002. Address for correspondence: DAG W. AKSNES Norwegian Institute for Studies in Research and Higher Education (NIFU), Hegdehaugsveien 31, NO-0352 Oslo, Norway E-mail: Dag.W.Aksnes@nifu.no 0138���9130/2003/US $ 20.00 Copyright �� 2003 Akad��miai Kiad��, Budapest All rights reserved A macro study of self-citation DAG W. AKSNES Norwegian Institute for Studies in Research and Higher Education (NIFU), Oslo (Norway) This study investigates the role of self-citation in the scientific production of Norway (1981- 1996). More than 45,000 publications have been analysed. Using a three-year citation window we find that 36% of all citations represent author self-citations. However, this percentage is decreasing when citations are traced for longer periods. We find the highest share of self-citation among the least cited papers. There is a strong positive correlation between the number of self-citations and the number of authors of the publications. Still, only a minor part of the overall increase in citation rates that can be found for multi-authored papers is due to self-citations. Also, the share of self- citation shows significant variations among different scientific disciplines. The results are relevant for the discussion concerning use of citation indicators in research assessments. Introduction Self-citations account for a relatively large share of all citations. Given the cumulative nature of individual research, citing oneself may be considered as a natural and acceptable procedure.1 On the other hand, scientists also tend to cite themselves as a result of egotism, for establishing their own scientific authority or to make their former works visible.2 Particularly when citations are used as indicators for assessing scientific impact, these citations are often treated as problematic.3,4 The reason is that self-citations do not reveal much about the impact of a work in the scientific community. Against this background it is important to know how prevalent self-citation is and how it influence on the citation indicators. A self-citation is usually defined as a citation in which the citing and the cited paper have at least one author in common. However, the term is sometimes used for other kinds of citation linkages (e.g., journal self-citations and institutional self-citations5). Also, in a more restricted version only publications having identical first authors are included as author self-citations.6 Methodologically, self-citations can be counted in two different ways, diachronously or synchronously.2 An author���s synchronous self- citations are those contained in the citations the author gives, that is in the reference lists of his publications. The diachronous self-citations are those included in the citations an

D. W. AKSNES: A macro study of self-citation 236 Scientometrics 56 (2003) author receives. While an author���s synchronous self-citation rate can be calculated by considering the papers he has authored or co-authored, a citation database must be adopted in order to estimate diachronous self-citation rates. Several studies have analysed self-citations synchronously. According to Tagliacozzo7 the share of self-citations amounted to approx. 17% in the two fields plant physiology and neurobiology. In another synchronous case study, Lawani2 found a self- citation rate of 15% in agronomic literature, while the corresponding share in cancer literature was 10%. Bonzi and Snyder,8 in a case study including the social sciences and the humanities, found an average self-citation rate of 11%, varying from 16% in the physical sciences (chemistry and geology) to 3% in the social sciences (economics and sociology). Rather similar results were presented by Snyder and Bonzi.9 These studies have shown that self-citation rates tend to vary between disciplines. A major drawback with all the studies, however, is their small sample size. For example, Tagliacozzo���s study7 involved approx. 180 articles, while Bonzi and Snyder8 analysed only 120 publications. Still, it appears that a synchronous self-citation rate of 10 to 20% is typical within the natural sciences and medicine. Diachronous self-citation rates, on the other hand, may differ from those calculated synchronously. In order to estimate such self-citations rates one needs bibliometric information on the citing papers. In some bibliometric studies of research performance the percentage of diachronous self-citations has been included as an indicator. For example, a case study of physics in the Netherlands (1985-1994) found a self-citation rate of 29%.10 Similarly, 29% of the citations in a study of Dutch chemistry (1980- 1991) represented self-citations.11 The percentage of self-citations has also been shown to be highest the first years after publication.12 Still there is need for more systematic information on self-citation rates. This is particularly important for assessing the representativity of citation indicators as performance measures. The aim of this study is to contribute to the knowledge on self-citation, and self- citation rates in particular. In the study we analyse the scientific production of Norway (1981-1996), in total more than 45,000 publications. We assess the element of self- citation in this scientific production and how the share of self-citation varies according to different parameters such as scientific discipline, overall citedness and number of authors. Finally, we discuss the implications of the findings with respect to the use of citations as indicators.