Relationship between citation counts and Mendeley readership metrics: A case of top 100 cited papers in Physics

摘要

Purpose - Social media has given way for the development of various new altmetric indicators. Mendeley readership count is one such indicator. The purpose of this paper is twofold. First, the paper aims to investigate the relationship between citation counts and Mendeley readership counts. The paper also evaluates the relationship between Mendeley readership metrics for two different time periods, thereby investigating its nature as an altmetric indicator. Design/methodology/approach - Data were collected using the Scopus database. Top 100 papers in Physics published during 2005 as well as in 2010 that received the largest number of citations were selected. Mendeley readership data were collected using Mendeley readership statistics for documents indexed in Scopus. For establishing a relationship between citation counts and Mendeley readership, correlation was calculated between the citations in Scopus database and Mendeley readership. The difference in Mendeley readership for different time periods was also investigated. Findings - The paper showed that for both the years, Mendeley readership counts were in positive correlation with citation counts. For the year 2010, it was found that Mendeley readership counts were in strong positive correlation with citation counts, whereas for 2005, they were in moderate positive correlation. Research limitations/implications - One of the limitations of this paper is that with time more scientists and researchers may join Mendeley causing various changes in data and giving different results. Also, the paper has focused on the highly cited papers in Physics. Originality/value - Very few studies have been conducted in the area of altmetrics, as it is a comparatively new and emerging field of research. The findings of this paper offer insights to the question whether Mendeley readership counts can be used as an alternative to traditional sources of bibliometric indicators like citations, h-index, etc. The paper also evaluates the difference in the nature of traditional bibliometric indicators and Mendeley readership counts.