Title: bgms: Bayesian Analysis of Graphical Models of Binary and/or Ordinal
Variables

Author: Maarten Marsman
 
Abstract:
Multivariate analysis using graphical models is rapidly gaining ground in
psychology. In particular, Markov Random Field (MRF) graphical models have
become popular because their graph structure reflects the conditional
associations between psychological variables. Although most psychological
variables are assessed on an ordinal scale, the analysis of MRFs for ordinal
variables has received little attention in the psychometric literature. In
this talk, we propose an MRF for ordinal data to fill this gap and present a
new methodology to test the structure of the proposed MRF. Testing the
structure of an MRF requires us to determine the plausibility of the opposing
hypotheses of conditional dependence and independence. To this end, we propose
a Bayesian approach using the inclusion Bayes factor to quantify the (lack of)
evidence for a given edge. The proposed methodology is implemented in the R
package bgms. We will demonstrate the functionality of the bgms package for
network analysis of binary and/or ordinal data and discuss current and future
developments of the proposed model and methodology.