On polychoric and polyserial partial correlation coefficients: a Bayesian approach

摘要

This article provides the estimation method for multivariate polychoric and polyserial correlation coefficients by using the simulation-based Bayesian method. It also shows that the partial version of the polychoric and polyserial correlation coefficients can be estimated using the corresponding estimates of the simple version. A simulation study illustrates the proposed method. Further, an application of the method to subjective well-being data is provided.