MCHANGE after GSEM

[Rohit Sonika]

Hello statalisters,

I am trying to estimate the marginal effects after a gsem command but am unsure how to frame the expression. The mchange command runs pretty much out-of-the-box after xttobit but doesn't seem o accept the predict(ystar) option after gsem. Would appreciate you help in solving this issue. Below is the code that I am trying after xttobit.

webuse nlswork3, clear
xtset idcode year
xttobit ln_wage union age grade not_smsa, ul(1.9)
mchange, centered predict(ystar(.,1.9))

Here is the corresponding gsem command. Would like to run mchange after this.

gsem ln_wage <- union age grade not_smsa RE[idcode], family(gaussian, rcensored(1.9))

I am using Stata 15.1 on MacOS 10.14.

Thanks!
Rohit