Are there differences in terms of citations between journal articles, book
chapters and books?
There are 138 books, 134 articles and 22 book chapters in the list of
publications with the highest number of citations. A comparison using
Mann-Whitney U tests shows that books are cited more frequently than
articles (Mdnbooks = 510,Mdnarticles = 341,U= 6550,p< 0.001). No
differences were found between chapter and books or chapters and articles,
but that may also be caused by the small number of chapters mentioned.
Are there differences in terms of citations between overview studies
versus data based studies?
Of the 291 publications that could be classified, 70 are overview articles and
221 data-driven articles. Overview articles are cited somewhat more frequently
(Mdn= 466) than data-driven articles (Mdn = 407), but the difference is
not significant (U= 7240,p= 0.419). This may be partly due to the large
difference in sample size.
Are there differences in terms of citation between single-authored publica-
tions and multi-authored publications?
One of the factors that make it hard to compare impact between dis-
ciplines is the number of authors. In science and medicinefive or more
authors is quite normal. The list of most cited papers in the science-oriented
open access journalPLOS-ONEcontains no single-authored articles, and
most of them havefive or more authors. In AL there are few articles with
more than three authors. In the analysis a distinction was made between
single- and multi-authored, but multi is typically two or three. Of the 243
publications mentioned, 139 are single-authored and 104 multi-authored.
Single-authored publications attract more citations (median = 513) than multi-
authored publications (median = 384). This difference is significant (U=6163,
p= 0.049).
9.7 Leaders and citations
In order tofind out to what extent there is a relationship between number of
citations and leadership, the data on the number of times someone is men-
tioned as a leader from Table 4.1 in Chapter 4 have been combined with the
citation scores. The correlation is significant (Spearman’srs= 0 .44;p< 0.01),
but not very high. Figure 9.4 shows the scattergram of the data.
The data show a clustering of informants, with relatively low scores on
both indications and a few with rather low citation totals but higher listings
as leaders. So leadership partly depends on academic status and impact, but
other factors clearly play a role, too.
114 The citation game