I am pretty new to SEM, so when I first saw a SEM model with 2 DVs, I wanted to know the difference between having 2 DVs and running SEM process twice for each DV.

From what I understand, having more than 1 dependent variable in SEM (say 2) can control existing correlation between DVs, which then allows us to parcel out the difference (or lack thereof) in each individual IV's effect on DVs that originates from the existing correlation between the DVs.

so my questions are:

  1. is the above description correct? if not, what would be the correct understanding / implication of adding more than 1 DV to SEM in terms of testing the partial effects of IVs?

  2. are there any other purposes to adding more than 1 DV to SEM instead of running the SEM per DV?


There are several reasons you might add more than one outcome (endogenous) variable to SEM.

  1. You can examine the partial correlation.
  2. You can carry out multivariate tests of two (or more parameters simultaneusly).
  3. You can compare the coefficients predicting each outcome.

As well as the reason you suggest.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.