Can <scp>M</scp>endeley bookmarks reflect readership? A survey of user motivations

Abstract

Although M endeley bookmarking counts appear to correlate moderately with conventional citation metrics, it is not known whether academic publications are bookmarked in M endeley in order to be read or not. Without this information, it is not possible to give a confident interpretation of altmetrics derived from Mendeley. In response, a survey of 860 M endeley users shows that it is reasonable to use M endeley bookmarking counts as an indication of readership because most (55%) users with a M endeley library had read or intended to read at least half of their bookmarked publications. This was true across all broad areas of scholarship except for the arts and humanities (42%). About 85% of the respondents also declared that they bookmarked articles in M endeley to cite them in their publications, but some also bookmark articles for use in professional (50%), teaching (25%), and educational activities (13%). Of course, it is likely that most readers do not record articles in M endeley and so these data do not represent all readers. In conclusion, M endeley bookmark counts seem to be indicators of readership leading to a combination of scholarly impact and wider professional impact.