I am using "mixed" with three levels (two nested random effects), and I wish to analyze the sensitivity of the results to individual observations, for example using Cook's D. The
mltcooksd command of the mlt package returns "mltcooksd works only after xtmixed, xtmelogit or xtmepoisson." Thus, I ran the same model using xtmixed, which returned "mltcooksd works only with 2-level models." Does anyone know a work-around or of another package with Cook's D or similar (e.g., DFBeta)? Thank you for your help.
mltcooksd command of the mlt package returns "mltcooksd works only after xtmixed, xtmelogit or xtmepoisson." Thus, I ran the same model using xtmixed, which returned "mltcooksd works only with 2-level models." Does anyone know a work-around or of another package with Cook's D or similar (e.g., DFBeta)? Thank you for your help.
