Proc Calis for Nonnormal Data

[DocChinnerson]

Hey gang! Long time user, first time caller. 

I'm trying to confirm my suspicion that proc calis does not provide estimators for nonnormal data. All the lit I've seen suggests the only option is to transform the data or delete outlliers. Any guidance would be appreciated. 

I have about 336 complete records after imputation. Outcome data are counts so a link function would also be helpful. 

Thanks!

Bill

[PaigeMiller]

Here's what the documentation says:

In structural models, as opposed to functional models, all variables are taken to be random rather than having fixed levels. For maximum likelihood (ML, the default) and generalized least squares (GLS) estimation in PROC CALIS, the random variables are assumed to have an approximately multivariate normal distribution. Nonnormality, especially high kurtosis, can produce poor estimates and grossly incorrect standard errors and hypothesis tests, even in large samples. Consequently, the assumption of normality is much more important than in models with nonstochastic exogenous variables. You should remove outliers and consider transformations of nonnormal variables before using PROC CALIS with maximum likelihood (default) or generalized least squares estimation.

[DocChinnerson]

Thank you!

Yes, that is what I've found in the SAS docs. Mplus provides a scaled chi square and robust standard errors option (per the documentation) that can generate estimates in SEM when data are not normal. Just wanted to confirm proc calis does not have similar options.