The Application of Information Entropy on the Sample Amount and Content in Contingent Valuation Survey
运用信息熵理论研究条件估值调查中的抽样问题

The contingent valuation method is a survey-based approach that directly elicits an individual's willingness to pay for environmental quality. Usually, large cost accompany with contingent valuation survey. In this paper, taking the survey of restoring Ejina banner ecological system as a case, the appropriate sample amount and survey content were discussed by adopting the theory of information entropy. In this paper, the marginal information gained and marginal cost incurred on increasing sample amount were calculated by taking Ejina banner's survey as a case. The calculated results showed that choosing 400 sample can get information equal distributing 646 sample in our empirical survey. At the same time , the calculated average mutual information and statistical test showed that we should investigated related social economic information with interviewee. The knowledge, income, registered permanent residence and residence site of interviewee should be preferred investigated, the age and sex can be excluded from the survey on interviewee's relating information. Usually, the higher average mutual information of interviewee's social economic information, the higher the influence of social economic information on the willingness to pay. ;At last, the empirical experience on sample in contingent valuation survey was summed up.?