Data and latent position modelling results for a coauthorship network on the topic latent position modelling.
The raw data was downloaded from Scopus on 22nd January 2022. It consists of all the Scopus entries that cite Hoff, P. D., Raftery, A. E., & Handcock, M. S. (2002). Latent space approaches to social network analysis. Journal of the American Statistical Association, 97(460), 1090-1098.
We cleaned these data to correct for duplicate names appearing due to middle names.
Then, we constructed a network of coauthorships between all those individuals with at least two publications.
Thus, any two individuals are connected with an edge if and only if:
-
they appear in at least one bibliographic entry together;
-
they appear independently in at least two bibliographic entries.
The resulting network is binary and undirected, and it is stored in the file data_coauthorships_lpm.RData.
After this, we used the R package latentnet to fit a basic latent position model, a latent position clustering model, and a latent position clustering model with random effects. Model choice for the number of dimensions and for the number of clusters was performed using the BIC criterion implemented in latentnet. The fitting algorithm was run with default parameter values.
The results for each model fitting are available from the folder '/results'. The results are provided as an RData file, in pdf, and in html format.