Using network centrality measures to improve national journal classification lists

Abstract

Research productivity for the scholar is often evaluated on the basis of his/her journal articles; however, specific journals are said to possess a higher measure of impact than others (Archambault & Larivière, 2009; Garfield, 2006; Glänzel & Moed, 2002). When a scholar decides where to publish, (s)he might consider a journal’s impact factor. Although Garfield (1973) claimed that citation counts to individual articles will determine the impact factor of a journal, newer evidence points to the contrary: a journal with a high impact factor can also influence an article’s readership and subsequent citation rates (Larivière & Gingras, 2010). This “chicken-and-egg” dispute (i.e., citations first or impact first?) can be tested, but can still have negative consequences for how journals are selected, rated, listed, and used by policymakers for developing measures of scholarly performance. For instance, in countries like Denmark and Spain classified journal lists are now being produced and used in the calculation of nationwide performance indicators. As a result, Danish and Spanish scholars are advised to contribute to journals of high “authority” (as in the former) or those within a high class (as in the latter). This can create a few problems.