AUTHOR=Johal Simran K. , Ferrer Emilio 

TITLE=Variation in emotion dynamics over time is associated with future relationship outcomes

JOURNAL=Frontiers in Human Neuroscience

VOLUME=Volume 18 - 2024

YEAR=2024

URL=https://www.frontiersin.org/journals/human-neuroscience/articles/10.3389/fnhum.2024.1331859

DOI=10.3389/fnhum.2024.1331859

ISSN=1662-5161

ABSTRACT=The application of time series models – such as vector autoregressive (VAR) models – to intensive longitudinal data has allowed researchers to explore the dynamics of multiple processes over time. Previous research has shown that incorporating information about these dynamic processes can improve the prediction of distal outcomes of the system – for example, predicting future relationship quality or break-up amongst romantic couples (Castro-Schilo & Ferrer, 2013). Yet these VAR models typically assume that the processes under study are weakly stationary, which is often an unrealistic assumption in psychological data. One way to account for potential non-stationarity in the process is to allow the parameters in the VAR model that represent the dynamics to change continuously over time. 

The current project expands on previous work by investigating whether variation of the dynamic parameters in VAR models over time was associated with future outcomes. In a sample of 148 romantic dyads, we examined whether the variation in parameters representing the couple’s daily emotion dynamics of affect was associated with relationship dissolution and relationship quality one to two years later. Our results demonstrate that certain autoregressive and cross-lagged parameters show significant variation over time, indicating fluctuations in different aspects of emotion dynamics. Furthermore, this variation is associated with future relationship outcomes, above and beyond the mean values of those parameters.