Data Mining for Predictive Value of
Discarded Individuals with Missing Data
Bruce Ratner, Ph.D.
The problem of analyzing data with missing values is well known to data analysts. Data analysts know that almost all standard statistical analyses require complete data, and consequently discard individuals with missing data. They make every effort to impute the missing data values, but are mindful that their first intention can leave a sizable sample of discarded individuals. It is not common practice to recognize and assess the predictive value of the sample of discarded individuals because there is no standard methodology for doing so. This article presents a non-imputation methodology that uncovers predictive value of the discarded individuals with missing data.
1 800 DM STAT-1, or e-mail at firstname.lastname@example.org.