You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The current robust methods and those in planning do not handle dummy and categorical variables (in exog) differently from continuous variables.
Our methods will run into problems if there are many dummy variables in exog, especially if some have small cell counts.
For example, subsampling will run into empty cell problems.
Small cell counts might also create influential points, that, however, we will need to include.
I briefly looked at part of the literature that includes specific categorical handling, but I did not look at the details.
Categorical exog are not a problem in current RLM, AFAIU, or we never ran into them, because there is no subsampling and similar methods.
The text was updated successfully, but these errors were encountered:
(there is not specific issue for this yet)
The current robust methods and those in planning do not handle dummy and categorical variables (in exog) differently from continuous variables.
Our methods will run into problems if there are many dummy variables in exog, especially if some have small cell counts.
For example, subsampling will run into empty cell problems.
Small cell counts might also create influential points, that, however, we will need to include.
I briefly looked at part of the literature that includes specific categorical handling, but I did not look at the details.
Categorical exog are not a problem in current RLM, AFAIU, or we never ran into them, because there is no subsampling and similar methods.
The text was updated successfully, but these errors were encountered: