A Flexible Python-Based Toolkit for Analyzing Team Communication

Team communication isn’t just empty talk — it encapsulates rich, dynamic processes of navigating challenges, coordinating taskwork, and resolving conflict. However, analyzing communication data is often both computationally and labor-intensive. Researchers must make a number of decisions regarding which features to focus on, how to operationalize them, and the level of analysis with which to aggregate them. This costly process makes it difficult to explore different methods of measuring team communication, as well as to test the robustness of findings across various decision pathways. In this work, we introduce a Python-based toolkit that streamlines the analysis of team communication. Our fully modular design extracts 164 communication features (and counting), and it gives researchers flexibility to examine versions of these features across three levels of analysis (utterance, speaker, and conversation). Drawing on empirical data for teams completing a moral reasoning task, we illustrate key applications of the toolkit: bootstrapping exploratory data analysis, generating measures for constructs of interest, and conducting sensitivity analysis of findings.