Interpretation of regression coefficients under a latent variable regression model

Abstract In standard linear regression where the predictor matrix X is of full rank, the regression coefficients are clearly defined as the parameters B appearing in the linear regression model. In latent variable models there is no direct relationship between the predictor variables and response variables. Rather they are both related to an underlying reduced‐rank set of latent variables. Recent papers have proposed different methods for …