|
Sample Balancing for
Database Response Models Bruce Ratner, Ph.D. Data analysts typically build a response model with a convenience sample, in which the distribution of responders and nonresponders does not reflect the population response rate. The procedure for adjusting the sample to match the population response rate is called sample balancing. Sample balancing is easy to understand; however, it is arithmetically cumbersome to perform. The purpose of this article is to provide a simple SAS-code program for sample balancing, which should be a welcomed entry in the tool kit of data analysts who frequently work on building response models. 1 800 DM STAT-1, or e-mail at br@dmstat1.com. |
|