Front cover image for A small sample theory for post stratification

A small sample theory for post stratification

Paul William Guy, H. O. Hartley (Degree supervisor), R. J. Freund, D. J. Hartfiel, L. J. Ringer, Texas A & M University (Degree granting institution)
When a sample is stratified after the units are selected, the process is called post stratification. Several new estimators using post stratification are considered for simple random sampling and for stratified random sampling that can be used even when the sample size is small, which usually leads to some post strata being empty. The problem of non-response, where some sampled units do not respond and thus can not be post stratified or used in the estimators, is also considered. Computer simulations are used to compare estimators on the basis of mean square error. A post stratified estimator is also developed for use in general designs and its variance estimate is given for the case of stratified random sampling

Thesis, Dissertation, English, 1979
Guy, [College Station, Tex.], 1979