The goal of this paper is to present several methods of visualizing student activity and collaboration in problem-based learning (PBL) in ways that can augment understanding of the complexity within PBL classrooms as well as to provide insights into the use of specific visual representations to address research questions in PBL. Grounded in sociocultural theory, we consider these representations to be an aid to mediate researchers' interpretation of the multiple data streams generated in PBL. For quantitative analysis, we introduce social network system (SNA), structural equation model (SEM) and path modeling. In terms of qualitative analysis, we exhibit chronologically-ordered representation of discourse and tool-related activity (CORDTRA), event maps, and spatial representations of physical activity. By reviewing articles that utilized these representations in PBL, we present several examples of how these visualizations were helpful in interpreting complex data and illuminating how students learn in PBL and other forms of collaborative inquiry. Although such visual representation methods enable us to visualize and trace complex dynamics and communicate findings with readers, they are not self-explanatory. We further discuss potential pitfalls of using visualizations and how to fully take advantage of those tools in PBL research studies.