Faradj Koliev
2016-Apr-28 13:38 UTC
[R] Robust clustered errors for probit ordinal regression analysis
Dear all, I?ll need your help with obtaining robust clustered errors. I use polr command in MASS package m<?porl(y~x1+x2,data=mydata, method=probit). In the rms package, this is as simple as: clusterSE<?robcov(m, mydata$id). Is it possible to do something similar for polr object as well? Thank you very much Best, Faradj [[alternative HTML version deleted]]