Sample Weights for Coverage: a Stochastic Interpretation

practice in sampling, placed Abstract in the context of probability theory. We hope to show the The derivation and correspondence between the two development of weights for approaches to estimation and sample surveys has come almost how for certain classes of exclusively from the sample distributions the two survey tradition. Yet the approaches lead to the same forms of sampling that are results, but via different most commonly used have direct paths. The paths are what interpretations as realizations of stochastic events arising from well defined probability distributions. This paper attempts to develop weights for estimators from sample surveys using maximum likelihood methods applied to the likelihood functions for make this investigation interesting, because they force the statistician to consider what assumptions are being made regarding how the data has arisen and whether the procedures used adequately reflect real life.