Disciplinary differences in altmetrics for social sciences

Purpose The purpose of this paper is to contribute to the understanding of altmetrics in different disciplines of social science: first, by investigating the current richness and future potential of altmetrics in the selected social science disciplines and then by evaluating the validity of altmetrics as indicators of research impact in each discipline through correlation analysis. Design/methodology/approach This study uses three approaches to understand the current richness and future potential of ten altmetric measures in nine selected disciplines: first, investigate the distribution and trend of altmetric data; second, verify the relationship between citation rate and altmetric presence of the discipline using Pearson correlation; and third, perform word frequency analysis on tweets to understand different altmetric presence in different disciplines. In addition, this study uses Spearman and sign test to find the correlation between altmetrics and citation counts for the articles that receive altmetric mention(s) to test the validity of altmetrics as indicators of research impact. Findings There is a steady increase in the number of articles that receive altmetric mentions in all disciplines studied. In general, disciplines with higher citation rates have higher altmetric presence. At the same time, altmetrics are also an effective complement to citation in disciplines with low citation rates. There is a moderate correlation with Mendeley and significant but weak correlations with Tweets and CiteULike in seven disciplines. Altmetrics appear effective as a predictor of citation counts in seven out of nine disciplines studied. However, there is low presence and lack of correlation with citation count in business-finance and law disciplines. Originality/value This paper furthers the understanding of altmetrics in social science disciplines. It reveals the disciplines where altmetrics are most effective, potentially useful, and fairly applicable. In addition, it presents evidence that altmetrics are an effective complement to citation in disciplines with low citation rates.