Bivariate dyadic patterns analysis using longitudinal actor-partner interdependence model and Markov chains for single-case [Author Accepted Manuscript]

The Longitudinal Actor-Partner Interdependence Model (L-APIM) is an interesting tool to analyze longitudinal variables in a dyad context. When these variables are categorical, L-APIM can be modeled using Markov chains, which captures both dyadic and time dependences by identifying interaction patterns through the classic actor and partner effects. In dyadic analysis, understanding the dependence between two variables may be important. To address this issue, it is necessary to extend univariate dyadic Markov chains modeling to simultaneously consider two variables. This article introduces a method that first detects the need for bivariate analysis and then analyzes interaction patterns in bivariate dyadic sequences. It details how to adapt the univariate Markov chains approach for bivariate analysis. Additionally, we provide a tutorial based on simulated data for applying this method to a simulated dataset using the R language for a single-case analysis. By focusing on a single-case analysis, the study validates the proposed method ensuring clarity.