CiteULike bookmarks are correlated to citations at journal and author levels in library and information science

Abstract

Aiming to explore the applicability of bookmarking data in measuring the scientific impact, the present study investigates the correlation between conventional impact indicators (i.e. impact factors and mean citations) and bookmarking metrics (mean bookmarks and percentage of bookmarked articles) at author and journal aggregation levels in library and information science (LIS) field. Applying the citation analysis method, it studies a purposeful sample of LIS articles indexed in SSCI during 2004---2012 and bookmarked in CiteULike. Data are collected via WoS, Journal Citation Report, and CiteULike. There is a positive, though weak, correlation between LIS authors' mean citations and their mean bookmarks, as well as a moderate to large correlation between LIS journals' impact factors on the one hand and on the other, their mean bookmarks, and the percentage of their bookmarked articles. Given the correlation between the citation- and bookmark-based indicators at author and journal levels, bookmarking data can be used as a complement to, but not a substitute for, the traditional indicators to get to a more inclusive evaluation of journals and authors.