Anonymous posted on Tuesday, December 20, 2011 - 11:31 am
How does Mplus compute standardized coefficients in a multiple group context? For example, if the model is
Y on X (linear regression)
and the coefficient for X is YX standardized and reported for each group separately, is it multipled by the standard error of X calculated across all the observations or the standard error of X calculated in that group only, and is it divided by the standard error of the outcome variable calculated across all observations or calculated within that group only?