Structural Equation Modelling (SEM) using STATA

[Ainun Nisa]

Hi,
I want to analyze my data using SEM in stata. This is my model:

Health variable consist of chategorical data with value 1 to 4. Here are my data example:

* Example generated by -dataex-. To install: ssc install dataex
clear
input float(health distance income asset)
2 0 17.661394 20.64536
3 0 18.379858 20.89301
2 0 16.562782 18.49829
4 0 15.789592 16.213587
4 0 15.789592 11.512925
3 0 16.300417 18.569101
3 0 18.092176 20.09253
3 0 15.76142 15.955577
2 0 17.399029 20.09253
2 0 16.206106 20.64536
3 0 14.69098 19.84017
2 0 17.843716 16.118095
2 0 13.81551 17.58135
3 0 17.175886 19.84017
3 0 16.993565 19.879295
3 0 15.60727 20.055786
3 0 16.333206 16.33320

I already run my data using SEM and i got this result:

My question is about how to interpret the result from income and asset to health. Because health is a dependen variable and its value are chategorical. I tried to interpret just for the coefficient sign, can i do that? or is there another solution to interpret my result using SEM?

[Sven-Kristjan Bormann]

To get a better answer you should show us the exact command that you have run and the output (no picture). If your dependent variable is categorical or ordinal scaled, then you need to use Generalised SEM and set your dependent variable as a generalised response variable with for example "Family/link: Ordinal, Logit" or "Multinomial, Logit" in the SEM builder.
Without knowing exactly what your variables represent, we cannot give you good guidance how to interpret the results.