Openness trends in Brazilian citation data: factors related to the use of DOIs

Research output: Contribution to journalJournal articlepeer-review

  • Rogério Mugnaini
  • Grischa Fraumann
  • Esteban F. Tuesta
  • Abel L. Packer
Digital object identifiers (DOIs) are important metadata elements for indexing and interoperability, as well as for bibliometric studies in times of openness. This study analyses the use of DOIs in the cited references of articles by authors from Brazilian institutions, their possible influencing factors and differences among areas of knowledge. It measures the extent to which the citation datasets are open for reuse by others in terms of the availability of DOIs. 226,491 articles were retrieved from Web of Science (2012–2016), making a total of 8,707,120 cited references, 68% of which include DOIs. The results showed that the hard sciences have higher percentages of DOIs in their cited references. The factor type of collaboration showed higher percentages when there is international collaboration, being significantly different from the other categories. However, when the analysis was conducted inside the areas, the international collaboration was found to be different particularly in the soft sciences and a couple of other areas. The articles with DOI attributed, as well as those with mention of research funding, had a significantly higher percentage, even in the interaction with the areas of knowledge. Among the open access routes the green routes showed the highest percentages, followed by golden (DOAJ and other) and Bronze, but green routes articles proved to be not significantly different from those not openly accessible. Finally, the principal collaborating countries also showed the greatest influence on the DOI attribution, with the exception of Peru and South Africa. Our findings provide evidence that studies on the availability and usability of DOIs can assist researchers, by underlining the importance of making greater use of this persistent identifier, as well as to provide consistency to citation analysis.
Original languageEnglish
JournalScientometrics
Volume126
Pages (from-to)2523-2556
Number of pages33
ISSN0138-9130
DOIs
Publication statusPublished - 2020

ID: 248567955