Motivated by a medical diagnosis problem that generates a wide range of problem-solving behaviors, three social scientists with different domains of expertise and different perspectives collaborated using a system dynamics modeling approach to do theory building in the social sciences. The collaboration led to a model of dynamic problem solving grounded in a unique dataset of 39 anesthesia residents confronting a simulated patient with a breathing problem, drawing upon and linking together theories of sensemaking and decision making. In this paper, we describe the collaborative journey, show glimpses of several models that we produced along the way, and reflect on some lessons that we learned about linking models, theory and data, about model-based collaboration, and about communication. Copyright © 2013 System Dynamics Society