Comparison of self-citation patterns in WoS and Scopus databases based on national scientific production in Slovenia (1996–2020)

No Thumbnail Available
Issue Date
2021-02-06
Embargo End Date
Authors
Budimir, Gordana
Rahimeh, Sophia
Tamimi, Sameh
Južnič, Primož
Advisor
Citation

Budimir, G., Rahimeh, S., Tamimi, S., & Južnič, P. (2021). Comparison of self-citation patterns in WoS and scopus databases based on national scientific production in slovenia (1996–2020). Scientometrics, doi:10.1007/s11192-021-03862-w

Abstract

Comparative studies of Web of Science and Scopus databases relate mainly to journal coverage and citation indicators. The aim of this study is to compare self-citation patterns in these databases. 25 years of scientific production (1996–2020) in Slovenia was analyzed. To offset limitations and errors of deriving data directly from databases, we used the COBISS/SciMet portal, which systematically collects records of citations with various authors identifiers for the total national scientific production. Both databases were harmonized according to the Frascati/OECD classification scheme of research fields. Self-citations were determined by comparing author identifiers rather than their names. Scopus shows better results in self-citation counts. This is mainly due to its higher coverage of local academic journals published in Slovenia and other regional journals, mostly in Humanities, where a fifth of all documents received about 3% more self-citations. In Engineering & Technology and Social Sciences, about 4% and 3% of all documents received approximately 2% more self-citations. However, these differences cause less than 3% more self-citations per researchers and two databases did not substantially differ in the relative citation ranking of researchers. Also, similar patterns of faster ageing of self-citations, rather than citations, were found in both databases, indicating a similar diminishing impact of self-citations on citations over time for all fields.

Table of Content
Description
Click on the DOI link to access the article (may not be free).
publication.page.dc.relation.uri
DOI