Modeling longitudinal dyadic processes in family research.

In this article, several dyadic analyses are applied to illustrate how they can be used to answer distinct research questions regarding associations between dyad members over time (longitudinal interdependence). This article focuses on how to conceptualize and empirically assess distinct dyadic processes, including time-sequential processes involving change in rank-order, parallel change processes involving intra-individual changes, dynamic dyadic processes involving both intra-individual changes and time-specific deviations (from intra-individual change), and accelerated dyadic processes involving acceleration of intra-individual change. These dyadic processes are depicted by four different dyadic models; a cross-lagged autoregressive model, a dyadic latent growth model (with and without structured residuals), and a dyadic latent change score model, respectively. These four longitudinal dyadic models are illustrated using a sample of 251 husbands and wives in enduring marriages. Each model focuses on a different dyadic process demonstrating distinct ways to empirically assess longitudinal interdependence; thus, when analyzing data, dyadic researchers must weigh the advantages and disadvantages of each and select the modeling approach that is most appropriate for the research question. (PsycInfo Database Record (c) 2021 APA, all rights reserved).