Bayesian Hierarchical Models for Longitudinal Meta-Analysis in Sports Science: A Practical Tutorial Using brms
Abstract
Meta-analysis in sports science has predominantly focused on cross-sectional comparisons, leaving the temporal structure of longitudinal evidence largely unexploited. This tutorial introduces Bayesian hierarchical longitudinal meta-analysis as a principled framework for synthesizing repeated-measures data and illustrates its implementation using the R package `brms`. We demonstrate a complete analytical workflow---from model specification and prior selection to convergence checks, posterior predictive checks, and reporting—using a simulated dataset based on pubertal stature growth in male youth athletes. The worked example highlights how polynomial growth models recover biologically meaningful parameters, such as age at peak height velocity, with full uncertainty quantification. Additionally, we compare three prior specifications using leave-one-out cross-validation to test robustness. Fully annotated code scripts are provided, providing researchers with the practical tools and methodological reasoning necessary to apply longitudinal Bayesian meta-analysis to their own work in sports science and related fields.
Related
- Ensemble clustering: A practical tutorial
- Choosing Stimuli Wisely by Using Individual Differences: A Step-by-Step Tutorial for the One-Stimulus-Per-Factor Principle in Vision Science and Optometric Assessment
- A practical risk framework for large language model use in life science research
- A hierarchical Bayesian model for dynamic phenotypes in multi-environment trials with genotype by environment interaction
- Behavioral Mapping of Hierarchical Dynamical Models Accelerates Bayesian Parameter Inference
BibTeX
@misc{junior2026bayesian,
title = {{Bayesian Hierarchical Models for Longitudinal Meta-Analysis in Sports Science: A Practical Tutorial Using brms}},
author = {Sandro Balbino Júnior and Ahlan Benezar Lima and Humberto M. Carvalho},
howpublished = {PsyArXiv},
year = {2026},
month = sep,
doi = {10.31234/osf.io/mkfsv_v1},
url = {https://osf.io/preprints/psyarxiv/mkfsv_v1/},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Bayesian Hierarchical Models for Longitudinal Meta-Analysis in Sports Science: A Practical Tutorial Using brms
AU - Sandro Balbino Júnior
AU - Ahlan Benezar Lima
AU - Humberto M. Carvalho
PY - 2026
DA - 2026/09/02
PB - PsyArXiv
DO - 10.31234/osf.io/mkfsv_v1
UR - https://osf.io/preprints/psyarxiv/mkfsv_v1/
AB - Meta-analysis in sports science has predominantly focused on cross-sectional comparisons, leaving the temporal structure of longitudinal evidence largely unexploited. This tutorial introduces Bayesian hierarchical longitudinal meta-analysis as a principled framework for synthesizing repeated-measures data and illustrates its implementation using the R package `brms`. We demonstrate a complete analytical workflow---from model specification and prior selection to convergence checks, posterior predictive checks, and reporting—using a simulated dataset based on pubertal stature growth in male youth athletes. The worked example highlights how polynomial growth models recover biologically meaningful parameters, such as age at peak height velocity, with full uncertainty quantification. Additionally, we compare three prior specifications using leave-one-out cross-validation to test robustness. Fully annotated code scripts are provided, providing researchers with the practical tools and methodological reasoning necessary to apply longitudinal Bayesian meta-analysis to their own work in sports science and related fields.
ER -