Collaboration analytics often focuses on assessing and monitoring individuals during collaborative problem-solving (CPS). A defining feature of CPS is the interdependence that exists between individuals when they work together -- that is, how they respond to and influence one another over time. While models that account for the impact of interdependence at the individual level of analysis (interdependent models) exist, they are often highly complex. This complexity makes them potentially difficult to use in assessments and systems that need to be explainable for educators, learners, and other researchers. Measures of the impact of interdependence at the individual level of analysis could inform decisions as to whether interdependent models should be used, or whether simpler models will suffice. Such measures could also be used to investigate specific questions about interdependence in collaborative settings. In this paper, I present a novel method of measuring the impact of interdependence on individuals using epistemic network analysis. To provide evidence of the validity of the measure, I compare it to qualitative findings that describe the impact of interdependence on individuals participating in team training scenarios. To demonstrate the value of the measure, I use it to assess the impact of interdependence in these data overall and to test hypotheses regarding the collaborative task design. My results suggest that the measure can distinguish between individuals who have been impacted by interdependence differently, that interdependence is impactful in these data overall, and that aspects of the task design may have affected how some individuals were impacted by interdependence.