Special Section: Model Identification and Estimation for Longitudinal Data in Practice

This special section focuses on model identifiability and estimability in the context of analyzing realistic educational, psychological, and more generally, social and behavioral sciences longitudinal data. Submissions should focus on statistical models commonly used for analyzing longitudinal data—including but not limited to—linear or nonlinear mixed-effects models, generalized mixed-effects models, and latent growth curve models. Longitudinal data has a clustered structure where measures are repeatedly collected on the same individuals over time and are nested within individuals or other higher-order groups. Empirical research in social and behavioral sciences aims to identify and estimate the different group-level effects. These specific group-level effects can be hard to detect if data on groups is sparse, lending to the issue of model identifiability. Moreover, datasets in applied research are often accompanied by a significant amount of noise, small sample size, missing data, nonlinearity, systematic heterogeneity, and other issues that are often compounded.

Special methodological consideration is needed to analyze these “messy” datasets to ensure researchers and practitioners are, indeed, modeling the underlying phenomena. The overarching objective of this special topic is to involve quantitative researchers to exchange novel perspectives on statistical considerations for modeling longitudinal data in challenging applied settings. Examples of such contributions may include but are not limited to (a) tools that can help researchers and practitioners assess whether models under consideration are identifiable, (b) guidance on how to obtain reproducible, stable, and precise parameter estimation for a given model, and (c) guidance on how to assess or compare the appropriateness of different models for a given data application.

Manuscripts published in this special section will be methodologically rigorous and illustrate the application of innovative statistical methodology with one or more real data examples of general interest to educational, psychological, social, or behavioral scientists.