| Previous: Counterfactual Evaluation | Up: Step 2: Conduct Analyses | Next: district.report: Examine the estimation |
In order to recode the data for counterfactual analysis or prediction, the option new.predictors must be added to a judgeit analysis command.
As an example, suppose we wish to find out what were happened if term limits were imposed and no incumbent could run for re-election, and that the incumbency indicator is labelled INC. The option
new.covariates=list("INC",0)
would instruct each routine to use this counterfactual data.
Suppose it came to pass that, instead, a group of representatives were forced to resign shortly before the election. If the new incumbency indicators are stored in a variable new.incs,
new.covariates=list("INC",new.incs)
would make that substitution. Make sure, however, that the substitution variable has the same number of districts as the model for that year; to confirm this, run the command
summary(judgeit.object,year).