03-13-2018 12:28 AM
I am using proc calis statement to run Structural Equation Model. I am using a nationally representative survey data, so I have individual-level weights as well as variables reflecting clustered stratified sampling of the survey. I am not sure how to account for the survey design in the analysis; all I found is a weight statement which could account for individual weights but not for the type of the sampling design. I use strata and cluster statements in proc surveylogistic, but haven't come across such statements in proc calis. Is there any way to incorporate the survey design information in the proc calis statement? Thank you.