I am a novice and I am confused about whether aregImpute draws a new bootstrap sample from the original data prior to each time it creates one of the "n.impute" multiply imputed datasets?
I know that I need to capture the variability in the original sampling distribution and I am not sure if aregImpute is resampling the original distribution with replacement (e.g. doing a bootstrap prior to doing imputations).
If I need to capture that variability, then do I need to use a function like boot from the boot package to pass the bootstrap resample datasets to aregImpute?
Thanks!