Coverage of Data Sources and Correlations Between Altmetrics and Citation Indicators: The Case of a Brazilian Portal of Open Access Journals

Abstract Analyzes the coverage of bibliometric and altmetric data sources, and the correlations between bibliometric citation indicators and altmetric indicators. The analyses were performed by area of knowledge, making it possible to identify differences between them. A total of 11,955 articles published in the journals of the Journal Portal of the Federal University of Goiás were analyzed. The main results were the high coverage of 95.7 and 96.7% found for the altmetric indicators of article downloads and accesses to the abstract, followed by 53.6% for Mendeley readers. The Pearson correlations between altmetric indicators and citation indicators vary from weak to moderate, the strongest correlations being those between the altmetric indicator of accesses to abstracts and the Google Scholar citations indicator (r = 0.475), and between Mendeley readers and Google Scholar citations (r = 0.467). Correlations between tweets and other altmetric indicators are always negative and very weak. The coverage of altmetric indicators varies among the areas, the highest values being found among Life Sciences. We concluded that the use of altmetric indicators, especially of Mendeley readers, can be considered complementary to bibliometric indicators.