The SEM Reliability Paradox in a Bayesian Framework

Within the frequentist structural equation modeling (SEM) framework, adjudicating model quality through measures of fit has been an active area methodological research. Complicating this conversation is research revealing that a higher measurement portion SEM can result in poorer estimates overall than lower models, given same misspecifications. Through population analysis and Monte Carlo simulation, we extend earlier to recently developed Bayesian evaluate whether these indices are susceptible reliability paradox, context using both uninformative informative priors. Our results show paradox occurs for RMSEA, some extent, gamma-hat PPP (measures absolute fit); but not CFI or TLI relative fit), across (MCMC) (maximum likelihood) frameworks alike. Taken together, findings indicate behavior newly adapted map closely their analogs. Implications utility identifying incorrectly specified models discussed.