How to calculate null model in Structural Equation Modeling (SEM)

[Rob Belanger]

Does anyone know how I calculate a null model in path analysis?

I have 4 variables in my model (1 dependent variable)

Thanks

[Joseph Coveney]

How about something along these lines?

version 15.1

clear *

set seed `=strreverse("1425895")'

tempname Corr
matrix define `Corr' = J(4, 4, 0.5) + I(4) * 0.5

quietly drawnorm q1 q2 q3 q4, double corr(`Corr') n(500)

// Constrained model
sem (q1 <- q2) (q2 <- q3) (q3 <- q4), nocnsreport nodescribe nolog

// Saturated model
sem (q1 <- q2 q3 q4) (q2 <- q3 q4) (q3 <- q4), nocnsreport nodescribe nolog

exit

[Marcos Almeida]

I fear your query is not quite clear to me, hence I had a different interpretation. I wonder what you mean by "calculate a null model". Hazarding a guess, I wish to underline that the GOF tests are basically a comparison with the null model. Therefore, by getting them, you're calculating the difference between your model and the null model. Hopefully that helps.

[Rob Belanger]

Thank you both for helping. Joseph's model is what I was looking for. Thanks a lot.

-Rob