The aim of a blockmodeling is to reduce a large potentially incoherent network to smaller and less complex network (Doreian et al. 2005). Compared to other community detection methods, blockmodeling allows not only detection of highly connected groups of units, but also the relations between the obtained groups of units (de Nooy et al. 2011).
The most common blockmodeling procedures are indirect and direct blockmodeling. In the case of a direct blockmodeling, the structure of the network has to be pre-specified. The most known structures of a blockmodels are: center-periphery, hierarchy, cohesion and others (Doreian et al. 2005).
The research done on empirical networks shows that the networks' structure can change in time. One of many examples could be find in the context of development of new scientific fields. As Bettencourt et al. (2009) explain, new scientific fields emerge with establishing of ties between small unconnected groups of researchers until one big component is formed. During the further development of a scientific field, its' standards and methodologies, with the increasing level of specialization and also with the influx of a new researchers, the revers process is starting - one big component begins splitting to a smaller and less connected groups of researchers.
Although the reasons and factors forcing the change of a type of blockmodels are usually well explained in the specific context of a studied phenomenon, the general factors that affect the transition from one into another type of a blockmodels are less known and have been so far studied to a limited extend.