Estimating total of wealth

When the conventional statistical methods are used to get an accurate estimator of total (or mean) wealth, some problems get rise. For example, in conventional statistical methods we assume that the response values are crisp. But in practice, people don't like to speak about their economic state clearly. Usually their response about their wealth is vague and unclear, like at most 10001, or approximately 700$. Another problem get rise when we need to define a subpopulation like the set of more wealthy people. Subpopulation sets and rare subpopulation sets are used frequently in survey sampling methods. The set of more wealthy people is not a clear and well-defined set. Here, we encountered to fuzzy sets and fuzzy numbers. A lot of statistical methods are generalized based on fuzzy theory, but fuzzy theory is not too extended in survey sampling field. In this paper, some survey sampling methods are generalized based on fuzzy theory to get a more appropriate estimator of total wealth in a population

کلیدواژه ها:

Crisp numbers ، Fuzzy sets ، Fuzzy numbers ، Stratified population ، Possibility theory