The article shows the possibilities of using the blockmodeling technique as a method of clustering network data in sociological research by conducting the secondary analysis of data related to structure of the community of St. Petersburg sociologists. The methodology of blockmodeling, data and results of the original research are briefly described. Using the blockmodeling algorithm of the CONCOR program, colleagues initially identified three clusters – “West End”, “East End” and “North End”, which differed in affiliation to organizations, publication strategies and orientation of scientists to the Western and domestic scientific community. The article describes the procedure of the blockmodeling algorithm used for secondary analysis, based on an indirect approach and hierarchical clustering. Using this method, we discovered a community structure similar to that found in the original study, but also, we had the possibility to take a deeper look at the selected groups, referring their structures to the “core-periphery” type in a complex form. The intersection of clusters obtained by different methods makes it possible to cross-validate the results of the analysis carried out by two independent research teams. The work can serve as a guide for researchers from other fields dealing with the problems of identifying related subgroups, since the described blockmodeling algorithm is universal and does not depend on the specifics of the subject.
Read full abstract