Scientometric and network analyses

Authorship and citation data were analyzed using descriptive statistics, scientometric indices, and network analysis methods \citep{Nita_2019a,Barab_si_2002,Aria_2020}. For scientometric data, we considered the following metrics: the number of articles published by an author, the number of citations received by hierarchical modeling papers, the local h-index of an author (indicating the number of publications for which an author has been cited by articles in our database at least that same number of times), and the number of citations from our database received by the papers referenced in the Literature Cited section (i.e., papers included in the reference sections of analyzed hierarchical modeling papers)\citep{Aria_2017}.
To understand co-authorship and co-citation patterns, we used network analysis \citep{borgatti2018analyzing}. We generated two undirected, unweighted networks: (1) a co-authorship network, where authors (nodes) are linked to other authors (edges) if they share at least one coauthor (including themselves), and (2) a co-citation (co-references) network, where paper references (nodes) relate to other paper references (edges) if they share at least one reference (including the references that serve as nodes). The co-authorship network was used to identify network leaders, while the co-citation network highlighted the most important papers referenced in the field \citep{RN770}. For each of the two networks, we also calculated two node-level centrality metrics: degree and normalized betweenness \citep{borgatti2018analyzing}. Degree centrality of an author represents the number of direct connections that the author has with other authors in the network and helps to identify the most collaborative authors (i.e., the authors with the highest number of connections). Betweenness centrality measures the extent to which an author lies on paths between other authors in the network \citep{Nita_2019a}. Such authors may be considered ”bridge” authors (i.e., they have the potential to influence the research topics in a co-authorship network) \citep{borgatti2018analyzing,Nita_2019a}. Co-authorship and co-citation networks were calculated using VosViewer \citep{RN770} and graphically represented using the NodeXL app \citep{RN768}. Node-level metrics were calculated using the R igraph package \citep{RN769}. Scientometric indices were extracted via bibliometrix R package \citep{Aria_2017}.