Xu, Man posted on Wednesday, March 30, 2011 - 4:55 am
Dear Dr. Muthen,
I have a path model of observed variables only. One of the outcomes is categorical and was declared so in CATEGORICAL section. To get the model run, I had to put parameterization=theta; Can I ask if this is to use a probit link for the categorical outcome variable? And how I should interpreate the STDYX or STD coefficient?
Both Delta and Theta are probit regression. Theta must be used for certain models. The interpretation of the standardized coefficients does not change.
Xu, Man posted on Friday, April 01, 2011 - 3:22 am
Many thanks! Do mean that it should be interpreted like: one standard deviation increase in X is assocated with say 0.3 standard deviation increase of Y, or 0.3 standard deviation increase of in the probit index of Y?