Page 125 - Becoming Metric Wise

116 Becoming Metric-Wise

reference lists. Copying mistakes from earlier reference lists occurs all
too often (Liang et al., 2014).
As it is in practice impossible to include all influences (through publica-
tions) in a reference list, citations always undercount scientific influence.

5.6.2 Self-Citations

The term self-citation has been used with different meanings. If the citing
paper has one or more authors in common with the cited paper one usu-
ally describes this as a self-citation. Yet, one could, more precisely make a
distinction between different authors. Indeed, if an article written by A
and B is cited in an article written by B, then this certainly is
a self-citation for author B, but what about author A? One could make
a distinction here between a self-citation on article level (then this is a
self-citation) and a self-citation on author level (then this is a self-citation
for author B and not for author A).
References to articles published in the same journal as the one in
which the citing article appears are called journal self-citations. This type
of self-citation is discussed further on in Chapter 6, Journal Citation
Analysis. When citations are used for evaluation purposes then citations
of articles authored by people working in the same research group or
department as the citing author are also called “self-citations,” see e.g.,
Moed et al. (1985a).
This leads to a hierarchy of self-citations:
• Author self-citations.
• Article self-citations.
• Research group self-citations.
• Country self-citations (in the context of own-country preference).
Besides these different forms of self-citations there are, of course, cita-
tions by outsiders (the remaining ones). Clearly in terms of diffusion of
scientific knowledge this group is the most important one, while author
self-citations may reflect no diffusion effect (possibly only from one
author to a coauthor).
Tagliacozzo (1977) conducted a systematic study of author self-
citations in the areas of physiology and neurobiology. The main conclu-
sions were that only few articles do not include self-citations, and the
numbers of author self-citations per article are widely distributed.
Moreover, authors are inclined to cite their own work more abundantly
than the work of any other single author. This observation coincides with