I will review the classical sampling theories from finite population and formalize the concepts of design for sampling with unequal probabilities. Define sampling design that supplies a unified framework within which to work for a search for optimum e...
I will review the classical sampling theories from finite population and formalize the concepts of design for sampling with unequal probabilities. Define sampling design that supplies a unified framework within which to work for a search for optimum estimators of a given parametric function. In section2, introduce the definition of generalized sampling design for finite population and propose the sampling selection algorithm to associated with the design. In section 3, discuss estimation theory for inference from sampled data to finite population. In particular, the Horvitz Thompson Estimator is in detail discussed.