topic Re: SEM in Statistical Procedures

SEM

SudeepNeupane — Fri, 09 Sep 2022 18:07:42 GMT

What proc should I use for a binary categorical outcome with all categorical predictors collected ten times?

I figured we could get outputs using any SEM procs; however, I am scared I would violate the statistical assumptions.

Re: SEM

ballardw — Fri, 09 Sep 2022 18:55:19 GMT

What is/are the research question(s) asked about the data?

There are multiple procedures but without knowing what kind of question needs to be answered it is hard to make a suggestion.

When you say "collected ten times" are the "times" independent of each other or related?

Re: SEM

SudeepNeupane — Fri, 09 Sep 2022 19:32:00 GMT

The data looks at the development of a disorder (binary outcome variable) over time mediated by some socio-ecological factors.

I meant to say we have the data of the same sample collected at ten different time points. Times are independent of each other.

Re: SEM

PaigeMiller — Fri, 09 Sep 2022 23:20:13 GMT

I don't see where you answered this question:

What is/are the research question(s) asked about the data?

Re: SEM

SteveDenham — Mon, 12 Sep 2022 12:46:00 GMT

If you want marginal estimates, try PROC GENMOD. If you want conditional estimates, try PROC GLIMMIX. If you know the structural equations/path model, try PROC CALIS. All of these handle repeated measures of a binary response. Selecting which to use depends very much on what @ballardw and @PaigeMiller are asking for - a concise, but complete, statement of the research question.

SteveDenham