CVM数据分析中的半参数模型及实证研究

条件价值评估法（CVM）是目前国际上评价景观价值最广泛的方法. 二分式选择问卷是最有效的CVM问卷格式，但二分式选择问卷数据不能提供支付意愿（WTP）的直接估计，其参数模型不能够充分拟合WTP，且不能充分挖掘样本数据信息. 针对上述问题，本文提出一种基于生存分析的CVM 数据分析方法，构建基于半参数比例风险模型的支付意愿影响因素分析模型，为CVM 数据分析提供新的路径. 以三江平原湿地景观价值评价为例进行实证分析，得到黑龙江省居民平均支付意愿为90.53元/人/年，个人收入、环保意识以及受访者所处地理位置是影响其支付意愿的主要危险因素. 研究证明了半参数分析方法在处理二分式CVM数据分析中的可行性及有效性.

The semi-parametric model and empirical research in CVM

The contingent valuation method (CVM) is a availably, flexible non-market valuation method which is widely used to evaluate the environmental goods, especially the landscape evaluation. The double bounded dichotomous choice is the most effective format of contingent valuation models except that it can not provide a direct estimate of the willingness to pay (WTP) and its parameter model can not fully fit the WTP as well as exploit the sample data information. This research provides a new path by addressing an analysis method of the CVM based on survival analysis and builds a semi-parametric proportional hazards model to analyse the impact factors of willingness to pay. Taking Sanjiang plain wetland landscape evaluation as an example, the willingness to pay of Sanjiang wetland landscape is RMB 90.53 per year. Individual average annual income, educational background, respondents' location and the attitudes to environment protect are the main risk factors which affect the willingness to pay. This research shows the feasibility and effectiveness of the semi-parametric analysis methods to deal with the data of dichotomous choice CVM.