2021/03/04 by Donald R. Williams · 111 citations
Biochemistry, Genetics and Molecular Biology · Mathematics · Neuroscience · Psychology · #Artificial intelligence #Bayesian inference #Bayesian network #Bayesian probability #Bioinformatics and Genomic Networks #Computer science #Conditional dependence #Data mining #Econometrics #Frequentist inference #Functional Brain Connectivity Studies #Graphical model #Inference #Machine learning #Mathematics #Mental Health Research Topics #Predictability #Statistics #Variable-order Bayesian network
paper · doi:10.1080/00273171.2021.1894412
published in Multivariate Behavioral Research 56(2), 336-352 (Taylor & Francis)
openalex publication_date 2021/03/04 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/31
Gaussian graphical models (GGM; “networks”) allow for estimating conditional dependence structures that are encoded by partial correlations. This is accomplished by identifying non-zero relations in the inverse of the covariance matrix. In psychology the default estimation method uses ℓ1-regularization, where the accompanying inferences are restricted to frequentist objectives. Bayesian methods remain relatively uncommon in practice and methodological literatures. To date, they have not yet been used for estimation and inference in the psychological network literature. In this work, I introduce Bayesian methodology that is specifically designed for the most common psychological applications. The graphical structure is determined with posterior probabilities that can be used to assess conditional dependent and independent relations. Additional methods are provided for extending inference to specific aspects within- and between-networks, including partial correlation differences and Bayesian methodology to quantify network predictability. I first demonstrate that the decision rule based on posterior probabilities can be calibrated to the desired level of specificity. The proposed techniques are then demonstrated in several illustrative examples. The methods have been implemented in the R package BGGM.