Bayesian Structural Equation Modeling

摘要: Abstract Structural equation models (SEMs) with latent variables are routinely used in social science research, and of increasing importance biomedical applications. Standard practice implementing SEMs relies on frequentist methods. A simple concise description an alternative Bayesian approach is developed. Furthermore, a brief overview the literature, specification SEMs, outline Gibbs sampling strategy for model fitting provided. inferences illustrated through industrialization democratization case study from literature. The has some distinct advantages, due to availability samples joint posterior distribution parameters variables, that highlighted. These provide important information not contained measurement structural parameters. As using study, this can often valuable insight into relationships.