Full and Fractional Counting in Bibliometric Networks

TL;DR

This paper compares full and fractional counting methods in bibliometric networks, highlighting how fractional counting at the network level can normalize link weights and clarify network structures, with a focus on different schemes and their applications.

Contribution

It introduces a new fractional counting scheme for bibliometric networks and clarifies distinctions among existing approaches, providing analytical methods and routines for application.

Findings

Fractional counting normalizes link weights in bibliometric networks.

Different fractional counting schemes can significantly affect network structure interpretation.

The paper offers practical routines for applying these schemes to bibliometric data.

Abstract

In their study entitled "Constructing bibliometric networks: A comparison between full and fractional counting," Perianes-Rodriguez, Waltman, & van Eck (2016; henceforth abbreviated as PWvE) provide arguments for the use of fractional counting at the network level as different from the level of publications. Whereas fractional counting in the latter case divides the credit among co-authors (countries, institutions, etc.), fractional counting at the network level can normalize the relative weights of links and thereby clarify the structures in the network. PWvE, however, propose a counting scheme for fractional counting that is one among other possible ones. Alternative schemes proposed by Batagelj and Cerin\v{s}ek (2013) and Park, Yoon, & Leydesdorff (2016; henceforth abbreviated as PYL) are discussed in an appendix. However, our approach is not correctly identified as identical to their Equation A3. Here below, we distinguish three approaches analytically; routines for applying these approaches to bibliometric data are also provided.