Project
Statistical methods for capturing emotional connectedness
Interpersonal relationships impact our health and well-being, whereby interpersonal similarity in for example experienced emotions is an important facilitator. Until now, this interpersonal similarity has mostly been examined using variable-centered (i.e., one variable at a time) and cross-sectional (i.e., one measurement at a time) approaches. However, interpersonal similarity often concerns multiple variables simultaneously (e.g., multiple discrete emotions) and can be expected to change over time and contexts. In this PhD-project, we therefore propose a new statistical framework that combines existing methods to optimally capture dyadic similarity (i.e., between two persons) in a multivariate, timepoint-specific way. We propose to compute similarity in person-specific profiles for each timepoint of each dyad separately. This will allow to examine the similarity of profiles in moment-to-moment experiences, how this similarity changes over time and how it can be associated to time-varying and person-level covariates.
However, building a new framework, albeit based on existing techniques, brings some challenges. This PhD project will handle these challenges in two parts: the first part will tackle the statistical challenges that come with the new framework and the second part will go deeper into challenges related to the application of the framework.