Opening the Black Box: A Multilevel Framework for Studying Group Processes

Abstract

Over time, groups can change in at least two important ways. First, they can display different trajectories (e.g., increases or decreases) on constructs of interest. Second, the configuration of group members’ responses within a group can change, such that the members become more or less similar to each other. Psychologists have historically been interested in understanding changes in groups over time; however, there is currently no comprehensive quantitative framework for studying and modeling group processes over time. We present a multilevel framework for such research—the multilevel group-process framework (MGPF). The MGPF builds on a statistical approach developed to capture whether individual members of a group develop a shared climate over time, but we extend the core ideas in two important ways. First, we describe how researchers can gain insights into group phenomena such as group leniency, group learning, groupthink, group extremity, group forming, group freezing, and group adjourning through modeling change in latent mean levels and consensus. Second, we present a sequence of model-testing steps that enable researchers to systematically study and contrast different group processes. We describe how the MGPF can lead to novel research questions and illustrate its use in two example data sets.