Modeli društvenih mreža kao posrednici između teorije i empirije: primer statističkih modela mrežne dinamike
Social Network Models as Mediators between Theory and Empirics: the Example of Statistical Models of Network Dynamics

Summary/Abstract: In this paper we present the argument for application, investigation and interpretation of a specific class of social network models as mediators between the theoretical and empirical perspectives on various problems of sociological interest. Our argument starts from the epistemological position which claims that (mathematical) models in science can mediate between theoretical constructs and data due to their epistemological autonomy. We claim that statistical models of network dynamics satisfy the criteria for this type of autonomy on the basis of: the way they are constructed, their research function, their application as an instrument, and the possibility of learning from the models. After we briefly describe this class of network models, we present the case for the first two criteria of autonomy and after reviewing recent applications in various social sciences we explain how these models function as an instrument and how they enable the researcher to learn from the results and use that knowledge for improvement of future research. These attributes are consequences of two statistical and mathematical properties of these models: the stochastic nature of data fitting and the modelling procedure focused on the individual actor rather than the aggregate data. In the discussion of the paper we claim that statistical models of network dynamics can offer a solution of bridging a gap between theory and data in several fields of social sciences. By emphasizing the need for autonomous models, we explain how they can improve the fruitfulness of the mathematical in social sciences with regard to existing theoretical pluralism and pre-paradigmatic state of social sciences.

• Details
• Contents