I am using survey data which uses jackknife replicate weights for variance estimation. I was hoping that I could run a logistic regression, since my outcome variables are either dichotomous or ordinal.
I am using TYPE = COMPLEX; REPSE = JACKKNIFE1;
My understanding is that the default analysis type is WLSMV. In order to do a logistic regression, in previous analyses I switched the estimator to MLR. However, this option doesn't seem possible with jackknife replicates. Is there another way to get logistic regression (versus probit regression which is what I believe to be estimated with the WLSMV estimation)?