Interactions

Interaction is a key ingredient of explanations developed by analytical sociologists, as interaction is a central reason why individual behavior aggregates to unexpected collective phenomena. Here, I formulate recommendations for modelers of interaction in theories explaining macro-phenomena based on micro-behavior. To this end, I reflect on various processes underlying interaction and show that they can shape interaction in very different ways. Next, I echo a proposal by Lopez-Pintado and Watts who argued that modelers should develop so-called “influence-response functions”. Rather than describing the actual underlying process of social interaction, these abstract functions describe only the outcome of interaction: changes in individuals’ attributes. Using the literature on social-influence in networks as an example, I demonstrate that influence-response functions are powerful tools to understand why interaction generates unexpected macro-outcomes, and to compare competing models. Influence-response functions guide empirical research and facilitate communication between researchers studying fundamentally different phenomena and seemingly contradictory micro-theories.