A theory construction methodology for network theories in psychology

In recent years, there has been a growing call to advance psychological theorizing through formal modeling. We answer this by introducing a methodology for developing psychological theories using probabilistic network models (PNMs). Originating in statistical mechanics, PNMs describe networks of interacting elements and have already shaped prominent theories in attitude, emotion, and decision research. We present a systematic guide on how to develop, analyze, and validate PNMs. Central to our framework is a survey of six base models which researchers can start from, extend, and adapt to their context. For each of these models we discuss existing applications and analyze them using two newly developed tools: a NetLogo model for simulations and an R package for visualizing mean-field dynamics. As a case study, we demonstrate the application of PNMs in theory development, before discussing the assumptions and limitations of the framework.