Previous: Purpose:

Up: IMPUTE

Next: Operators

Recommended Use:

To create a dataset with your imputed variable, JudgeIt must erase the original file and save a new file with the transformed variables appended. This process is risky; if something terminates the program between erasing the original datafile and recording the new one you could lose the GAUSS version of your dataset. (Of course, you will still have the ASCII version and can run ATOG again to regenerate the GAUSS version.) For this reason, we recommend that you place the variables for which transformations will be imputed into a separate datafile, so that fewer will be jeopardized by the process.

Since the imputation may take significant computer time, you do not want to have your JudgeIt command file run through this process each time an analysis is conducted. Once the imputed variable is created, the procedure is not needed again. For this reason, we recommend you create a separate JudgeIt command file just to perform all imputations and save all imputed variables in the same dataset. The first impute command will create a new dataset; the following impute commands will then append the new variables to this file. For subsequent analysis commands, you will then presumably use the variables in this imputed dataset for YVOTE.