Testing Group Differences in Network Models using Bayesian Model Selection: A Tutorial
Abstract
Network models are increasingly used to model multivariate data and many research questions are about whether networks differ across groups. This tutorial provides an accessible introduction to Bayesian model-selection for testing group differences in networks, as implemented in the R package bgms. This approach provides posterior inclusion probabilities and Bayes factors for each group difference, allowing researchers to distinguish evidence for a difference from evidence for equality. We focus the tutorial on group differences in ordinal network models, since the vast majority of data in cross-sectional research are measured on ordinal scales. We explain the model parameters, prior choices, missing-data handling, convergence checks, checks of model fit, and the interpretation of posterior summaries of group differences. The workflow is illustrated with a reproducible comparison of problematic pornography use symptom networks in women and men.
Related
- Analysis and Reporting Guidelines for Bayesian Graphical Modeling of Cross-Sectional Data: A Tutorial in JASP
- Bayesian Hierarchical Models for Longitudinal Meta-Analysis in Sports Science: A Practical Tutorial Using brms
- A tutorial on selecting, translating, and reporting variance priors in Bayesian multilevel models.
- Deep Bayesian Filtering for Nonlinear Structural System Identification: A Tutorial on Structural and Physical Inductive Biases
- 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
BibTeX
@misc{haslbeck2026testing,
title = {{Testing Group Differences in Network Models using Bayesian Model Selection: A Tutorial}},
author = {Jonas Haslbeck and Mona Klau and Tessa F. Blanken and Maarten Marsman},
howpublished = {PsyArXiv},
year = {2026},
month = nov,
doi = {10.31234/osf.io/rx745_v1},
url = {https://osf.io/preprints/psyarxiv/rx745_v1/},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Testing Group Differences in Network Models using Bayesian Model Selection: A Tutorial
AU - Jonas Haslbeck
AU - Mona Klau
AU - Tessa F. Blanken
AU - Maarten Marsman
PY - 2026
DA - 2026/11/02
PB - PsyArXiv
DO - 10.31234/osf.io/rx745_v1
UR - https://osf.io/preprints/psyarxiv/rx745_v1/
AB - Network models are increasingly used to model multivariate data and many research questions are about whether networks differ across groups. This tutorial provides an accessible introduction to Bayesian model-selection for testing group differences in networks, as implemented in the R package bgms. This approach provides posterior inclusion probabilities and Bayes factors for each group difference, allowing researchers to distinguish evidence for a difference from evidence for equality. We focus the tutorial on group differences in ordinal network models, since the vast majority of data in cross-sectional research are measured on ordinal scales. We explain the model parameters, prior choices, missing-data handling, convergence checks, checks of model fit, and the interpretation of posterior summaries of group differences. The workflow is illustrated with a reproducible comparison of problematic pornography use symptom networks in women and men.
ER -