Preregistration Guidelines for Longitudinal Network Analyses

Over the past decade, longitudinal network analyses have gained significant traction in psychological science. These approaches, applied to intensive time-series or panel data, require a high level of flexibility and involve numerous modeling decisions, which can introduce considerable degrees of freedom into the process. Despite their growing use for confirmatory research, there remains a notable lack of preregistration guidelines. The present paper introduces a comprehensive checklist with guidelines for preregistering of longitudinal network analyses. The checklist is designed to be a resource for both authors and reviewers, ensuring that all critical aspects of the preregistration process are adequately addressed. We specifically focus on four widely used models: Panel Graphical Vector Autoregression (VAR) models, cross-lagged panel network analysis (CLPN), multilevel VAR (mlVAR), and Group Iterative Multiple Model Estimation (GIMME). We offer detailed guidance on preregistering studies using these approaches to mitigate biases and enhance transparency. Finally, we describe the importance of preregistration and the need to adapt and refine these guidelines as the field continues to evolve.