Key points are not available for this paper at this time.
Standardization is discussed as a technique to control for extraneous variables in survey analysis. The weights for the standardization are chosen, in the case of estimation, in order to minimize the variance of the estimator and, in the case of a significance test, in order to maximize the power of the test. Examples are given to illustrate the uses and limitations of standardization in survey analysis.
Graham Kalton (Mon,) studied this question.