Structural Similarities between Research Impact Indicators: Factor Analysis of Citation and Altmetric Indicators

By investigating the structural similarities between altmetric and citation indicators through factor analysis, the present study attempts to identify similar indicators and experimentally group them based on their impact dimensions. Applying a citation analysis method, it concentrates on a purposive sample consisted of papers published in PLoS journals during 2010-2012. The results of the factor analysis led to identification of a model consisted of 3 factors explaining 53 percent of the variance of the latent variable, i.e. “the impact of research outputs”. The experimental model seems to adhere to the theoretical model proposed by Junping & Houqiang (2015). According to the yielded model, the factors differ in terms of the kind, extent and depth of the impacts they explain. The first factor, called “impact at perception level”, implies the widest, though the most superficial, impact. The “perception level” can be considered as the “consumption” level that does not necessarily lead to “usage”. At another point of the impact continuum, there exists “impact at social media level” that is relatively higher in terms of the interaction between users and texts. The level is relatively more limited in terms of the quantity of users, but deeper in terms of the impact the related indicators may have. At the other extreme, one may find “the impact at usage level”, that is achieved after reading the texts, deeply interacting with and selecting them purposely. The finding indicates the power of altmetrics in measuring different types of impact with different extents and depths. It helps understanding the nature of the impacts the indicators may represent and thereby help recognizing the association between the indicators and overcoming their multiplicity and chaos.