Models for Dyadic Data. Chapter in Cambridge Handbook of Research Methods in Clinical Psychology

Abstract

Chapter (accepted and in press) prepared for the forthcoming Cambridge Handbook of Research Methods in Clinical Psychology. Edited by A. Wright & M. N. Hallquist. We describe models for analyzing data from dyadic systems such as therapist-client, mother-children, or romantic partners, among others. We define key characteristics of dyadic systems and then identify clinical research questions related to dyadic systems and processes that unfold over time. We use these questions to select a set of statistical models and data analytic techniques for answering clinical research questions related to dyadic research.

Dyadic pattern analysis using longitudinal Actor-Partner Interdependence Model with Markov chains for unique case analysis

“…The aim of capturing a time dynamic in dyadic interactions is achieved by using dynamic systems analysis, which considers the temporal evolution of behavioral sequences over discrete time points, in contrast to static models that do not study the temporal dynamics of variables (Estrada et al, 2020;van Rijn, 2008;Kendall & Comer, 2014). The use of sequences is justified to achieve greater ecological validity, which can be defined as a measure of how well a model can predict real-life behavior (Walls et al, 2006).…”

Methodology for Identification, Visualization, and Clustering of Similar Behaviors in Dyadic Sequences Analyzed Through the Longitudinal Actor-Partner Interdependence Model With Markov Chains

“…When data is collected over several short periods, such as several successive days, the data reveals changes in the behavior of each individual over time while considering the mutual influence of each member of the dyad. An adaptation of the APIM, called the longitudinal actorpartner interdependence model (L-APIM), is used to analyze this data (Estrada et al, 2020). When the data are categorical, these dyadic sequences can be modeled using homogeneous Markov chains.…”

Momentary profile similarity measures to capture similarity in multivariate dyadic time series

“…At the same time, such research is on the rise. This surge is driven by the understanding that interpersonal relationships are dynamical in nature, with their characteristics continuously changing in function of internal and external contexts (8,24). For example, two colleagues might showcase a low similarity in personality related behaviors in a working context where one of them acts rather introverted and conscientious, while the other one acts more extraverted and carefree.…”