What drives Open Access Advantages, and what are the implications for scientometrics?

The tendency OA research to receive more citations and online attention than paywalled work - has been widely discussed but only partially reported. Existing studies often focus on narrow samples, single metrics, or fixed time periods, leaving unanswered questions about how Open Access Advantages (OAA) vary across disciplines, OA type, journal status, and article age. Furthermore, much research neglects to understand the causes of the OAAs. This talk presents results from large scale analyses of ~45 million journal articles published between 2010-23, examining OAAs across citations and 6 altmetric indicators. Results show that OAAs are real, tend to persist over time but are highly heterogeneous. OA citation advantages range from modest (1.25–2x in medical and life sciences) to negligible (technical) or declining (social sciences). Altmetric based OAAs are often stronger, although are more mixed: OA articles are ~3x more likely to be mentioned on Twitter, and >2x as likely to appear in news sources. OAAs vary by OA type and journal status, with Green OA frequently outperforming Gold OA, and high status journals showing stronger advantages. Temporal analyses reveal early year boosts for OA articles that persist in the following years. These results suggest that OAAs are not a single phenomenon but result from a number of mechanisms, including visibility, discoverability and early access – the effects of which are often driven by extraneous forces, relating to subject norms and the growth of individual disciplines. The talk will discuss implications for scientometrics, including how OA reshapes impact measurement, and how evaluators should interpret OA related advantages.