Applications of Categorical Recurrence Quantification Analysis to Communications in Teams: Lessons Learned and Best Practices

As technology enables increasingly large, complex, and distributed team structures, new methods are needed to assess coordination in action teams. Recurrence Quantification Analysis (RQA) offers a powerful tool for capturing team coordination dynamics, yet its application to these modern teams remains methodologically underdeveloped. This paper identified and examined four characteristics of large action teams that contribute to analytical challenges when using RQA. We presented a systematic framework of Pre-Analytical Treatments, which are data preprocessing methods that can make RQA more computationally feasible while preserving meaningful team-level patterns. We then empirically evaluated three specific Pre-Analytical Treatments that either preserved temporal structure (Time Series) or controlled for periods of silence (Sequential, Dead-air Randomized). Using communication data from a 14-person expert team, we applied each treatment and examined their effects on RQA metrics and correlations with team trust, cohesion, and workload measures. Results demonstrated that these preprocessing decisions significantly affect RQA output relationships with team-level outcomes. Rather than advocating for a single approach, this study provides researchers with an empirically-grounded framework of methodological trade-offs to consider when applying RQA to large teams, offering both theoretical clarity about the impact of preprocessing decisions and practical guidance for selecting appropriate treatments based on research objectives.