Assortative mixing, preferential attachment, and triadic closure: A longitudinal study of tie-generative mechanisms in journal citation networks

This study reveals the roles of three tie-generative mechanisms, namely, assortative mixing, preferential attachment, and triadic closure, in forming citation links in journals through the exponential random graph modeling approach. The study also adopts a longitudinal design to examine how the roles of the three mechanisms evolve over time. The data involve citation exchanges in Internet research among 680 journals in 12 subject areas from 2000 to 2013. Assortative mixing by discipline and publication foci is a significant tie-generative mechanism in journal citation networks. The magnitude of assortative mixing by discipline remains stable over time, whereas that by publication foci declines over time. Journals in Internet research demonstrate an increasing preference for influential journals to form citation links over time. The indegree of journals does not exert a significant impact on citation link formation among journals, whereas the outdegree of journals imposes a significantly negative impact on citation link formation among journals. Triadic closure is an important force that facilitates the formation of citation links among journals. The findings of this study improve our knowledge of the organizing principles that underlie journal citation networks and advance our understanding of the production process of scientific knowledge in Internet research.