I try to perform a multinomial regression with svy_vglm function from svyVGAM package after propensity score matching. I used matchthem function from MatchThem package in R.
model <- with(matchthem object, svy_vglm(outcome~exposure, family = multinomial(refLevel="0")), cluster = TRUE)
The script runs well I think but I get a warning message :
In vglm.fitter(x = x, y = y, w = w, offset = offset, Xm2 = Xm2, : some quantities such as z, residuals, SEs may be inaccurate due to convergence at a half-step
The problem is that: I can't obtain the summarized results.
a summary(pool(model)) isn't applicable for objects of class svy_vglm, and I get error message when I try a function MIcombine(model) or MIextract(model, fun=vcov) :
Error in vcov.default(X[[i]], ...) : object does not have variance-covariance matrix
Do you know how I can pooled or summarized the estimations ? May be should I used others functions to pool or correct the variance-covariance matrix ?
Thanks in advance