Hi, all

I am hoping for some insights into an appropriate independent goodness-of-fit test for GSEM as well as information on/Stata code for mediation analysis.

The model I am working with is pictured below where:
X1: Ordinal (1-4)
X2: Binary, factor variable (5 categories)
X3: Continuous (mediator)
X4: Binary, factor variable (7 categories)
X5: Continuous
Y1: Categorical, multinomial logit
Y2: Binary, bernoulli logit
All observed, no latent.

Array