The link of the theory of standard inference for finite population and the theory of
sample survey, especially their common foundation of the theory of probability, provides
in depth understanding in the theory of sample survey and would therefore enhance the
skills in situation analysis as well as creative problems solving in both practical
applications and theoretical domains of statistics science. In practice, based on the
properties of the estimators as discussed, when simple random sampling without
replacement is employed, for a large population, the sample proportion is still
recommended, and for a relatively large population, the estimator Pwor
ˆ with bias
adjusted may be considered as a better alternative in estimating population proportion.