XI Conference – European Congress of Methodology

A multilevel Ornstein–Uhlenbeck process with crossed random effects for multivariate time series

24 Jul 2025, 11:30

30m

Faculty of Social Sciences and Communication. (The Pyramid)/. - Foyer (Faculty of Social Sciences and Communication. (The Pyramid))

Poster Statistical analyses Poster Session 3

Speakers

José Ángel Martínez-Huertas (UNED)Dr Emilio Ferrer (University of California, Davis)

Abstract

Stochastic differential equation (SDE) models are very useful for analyzing time series. In Psychology, some SDE models have been used for the study of affect dynamics. For example, the Ornstein–Uhlenbeck (OU) process is a simple but very interesting model for this purpose. In this context, some authors have extended that SDE model to a multilevel framework using Bayesian estimations to study individual differences in their parameters. Here, we propose the OU process, a SDE model, to analyze multivariate time series by means of crossed random effects for individuals and variables. Our extension allows to estimate the variability of different parameters of the process, such as the mean (µ) or the drift parameter (φ), across individuals and variables of the multivariate system using a Bayesian framework. In this presentation, we illustrate the estimations and the interpretability of the parameters of this multilevel OU process in an empirical study of affect dynamics, and also conduct a simulation study to evaluate whether the model can recover the population parameters generating the OU process. Our results support the use of this model. Thus, we conclude that it can be a useful approach to analyze multivariate time series of affect dynamics.