Usage, content and citation in open access publication: any interaction effects?

• Published in: Scientometrics 2021-12, Vol.126 (12), p.9457-9476
• Main Authors: Vílchez-Román, Carlos, Vara-Horna, Arístides
• Format: Article
• Language: English
• Subjects: Application programming interface; Cognitive ability; Computer Science; Documents; Empirical analysis; Information Storage and Retrieval; Library Science; Open access publishing; Social factors
• ISSN: 0138-9130; 1588-2861
• DOI: 10.1007/s11192-021-04178-5

This study aims to validate an empirical model, at document level, that explains the interaction among content, usage, and citation within open access publications. The PLoS site was the information source for this study. Using an R API (Application Programming Interface) for PLoS ONE, 776,465 records were downloaded on August 24, 2018. Those records (from 2006 to 2018) were organized according to the PLoS’ thematic areas. The empirical framework was validated using mediation analysis. For computing the parameters’ significance, bootstrapping with 500 replications for the general model and each thematic area was used. When usage was included as the mediating variable within the model, the total effects of cognitive and social variables got better predictive capability, as expressed by the explained variance of citation ( R 2  = 0.282) and usage ( R 2  = 0.333). The same trend was observed for the indirect effects after carrying out the mediation analysis by categories. Promotion campaigns of scientific publication should reinforce the widespread adoption of easy-to-use social media because, besides the velocity and variety of diffusion channels, the extended use guarantees that journal’s papers will reach increasing audiences. This is one of the first studies that analyze the interaction effects of variables at the article-level within open access publications.