Qiao Hu posted on Saturday, July 28, 2018 - 9:10 pm
I run a two-level Bayesian structural equation modelling (BSEM). I can read the significance of indirect effects in MPLUS output directly, do I need to add a more bootstrap procedure to check the indirect effects?
With Bayes you don't need bootstrap; Bayes gives you all you need directly. Bootstrapping is used to allow for non-symmetric confidence intervals. Bayes provides non-symmetric CIs as part of its estimation.