Methodological problems with the empirical testability of sociocybernetic theories

Abstract

Theories in sociocybernetics usually have a high degree of complexity, which may lead to methodological problems when these theories are tested in confrontation with “real world data”. From a review of the literature it appears that only a small percentage of the sociocyberneticians make an attempt to test their theory empirically. Two alternative approaches are used: the deduction from the theory of a series of univariate and bivariate hypotheses, subsequently tested with the usual statistical methods; and the reformulation of the theory into a simulation model and comparing the trajectories generated by the computer with time series of “real world data”;. Both approaches have their advantages and risks. It is shown in a simulation experiment that the inclusion of various sociocybernetical ideas (self‐referencing, goal‐referencing, morphogenesis) leads to a model with many degrees of freedom, and consequently of a large “no man’s land” in between the set of “verifiers” and the set of “falsifiers” of the theory.

Citation

van der Zouwen, J. (1996), "Methodological problems with the empirical testability of sociocybernetic theories", Kybernetes, Vol. 25 No. 7/8, pp. 100-108. https://doi.org/10.1108/03684929610149684