How to run SEM mediation analysis with ordinal variables?

[Evan Carron-Kee]

Hi,

I'm an undergrad doing a summer research project that I'm statistically underprepared for - I'm really learning as I go! I've taken CenterStat's free introductory SEM course which was very helpful but only dealt with continuous variables.

I'm at a loss as to how I should treat ordinal variables - all of my observed exogenous and endogenous variables are ordinal, and although I haven't fully decided what my model will look like, I want to first figure out how to deal with these variables in Stata appropriately.

All of the literature I've come across so far is either not specific enough or too complex - I've heard that polychoric matrices might be helpful but I'm not sure how to implement this. I've also seen some say that in some cases, ordinal variables can be treated as continuous variables.

Any help or direction towards some introductory material would be very helpful! Mainly though, any help/resources that would give me practical tips on how to implement this in Stata would be perfect - I don't have a burning desire to understand all the maths behind it. (As long as I have a general idea what's going on!)

Thanks,
Evan