Measuring network-driven citations: An adjusted citation count metric

Abstract

Citation count is extensively used within research systems to measure research performance and relevance. However, high numbers of citations could result from many different factors. This paper proposes an adjusted citation count metric to help identify citations from researchers’ networks. Using data from the Web of Science core collection, this research suggests and illustrates an approach to compute the proportion of citation counts from within and out of author networks, defined as the set of their co-authors. The results reveal the trends and effects of author networks on citation counts. The paper explains how these findings could be used to assess research performance. Its utilisation will depend mainly on the objectives of the assessor and the stage of the researcher’s career. An algorithm is also proposed that could be used to automate the computation of the proposed metric. It could serve as a step towards refining and integrating this metric into existing bibliometric analysis packages built on the R software platform.