DCovR: A modeling framework for capturing the concurrent and delayed effects of human interaction

Human behavior, feelings and thoughts continuously fluctuate across time, especially when interacting with other people. Studies on this topic mostly focus on interactions in dyads (such as romantic couples), and try to disentangle partner-induced fluctuations of the behavior from the within-person fluctuations. For example, if a person feels sad, it will partially depend on his own nature to which extent he will still feel sad at the next moment in time, but his partner could also help ending a sad episode, by saying comforting things, etc. Because the currently available methods fall short when multiple variables are being measured, I will develop a new framework for unraveling the interpersonal dynamics of dyads, named DCovR.

DCovR will handle multiple variables, by reducing the variables of each person separately to a limited number of summarizing variables, called components. Simultaneously, we regress the components of both persons on one another to capture concurrent mutual influences. However, some partner-induced effects occur only after a while. For example, sometimes comforting words of a partner have to sink in before there is an emotional response. Therefore, the second objective is to capture the delayed effects as well. The third objective is to reveal the differences and similarities between the interpersonal dynamics of multiple dyads.