Engineering systems design is a dynamic socio-technical process where the social factors, such as interdisciplinary interactions, and technical factors, such as design interdependence and the design state, co-evolve. Understanding this co-evolution can lead to behavioral insights, resulting in efficient communication pathways and better designs. In that context, we investigate how to generate behavioral insights to inform effective structuring of interdisciplinary interactions in engineering systems design teams. We present an approach that combines the predictive capabilities of computational modeling with contextual information from empirical data. A stochastic network-behavior dynamics model quantifies the co-evolution of design interdependence, discipline-specific interaction decisions, and the changes in system performance. We employ two datasets, one of the student subjects designing an automotive engine and NASA engineers designing a spacecraft. Then, we apply Bayesian statistical inference to estimate model parameters and compare insights across the two datasets. The results indicate that design interdependence and social network factors such as reciprocity and popularity have strong positive effects on interdisciplinary interactions for the expert and student subjects alike. An additional modulating impact of system performance on the number of interactions is observed for the student subjects. Inversely, the total number of interactions, irrespective of their discipline-wise distribution, has a weak but statistically significant positive effect on system performance in both cases. However, we observe that excessive interactions mirrored with design interdependence and inflexibility in design exploration reduced the system performance. These insights support the case for open boundaries in systems design teams to improve system performance.