shira mitchell 3 months ago
blog post about imputation:
With nonresponse worsening, we want to adjust for a lot of covariates.
This often means handling many missing covariates.
In theory, fit one big model for everything. But how can practitioners handle this ?
add a skeleton here at some point
