Relational Turbulence: A Latent Model Test of Theoretical Propositions

ABSTRACT In this study we simultaneously tested theoretically specified predictions from relational turbulence theory (RTT) using a fully latent structural regression model. A total of 807 college students in dating relationships responded to survey items measuring variables embedded in RTT, including relationship parameters (relational uncertainty and interdependence), biased cognitive appraisals, emotions, relational communication, and relational turbulence. Results from structural equation modeling revealed that RTT predictions were consistent with our data, but a respecified model with less restrictive constraints was a better fit. This less restrictive model allowed relational uncertainty to predict negative affect, and partner interference and facilitation to predict biased cognitive appraisals. Propositions of RTT were sustained.