A Contingent Valuation Test for Measuring the Construct Validity of Willingness-to-Pay Estimates Derived from Choice Experiments
提出一个基于两轮调查的协议，通过比较不同混合Logit模型估计的支付意愿分布，检验其建构效度，帮助研究者判断哪种模型更可靠。
We introduce a protocol for measuring the construct validity of competing willingness-to-pay (WTP) distributions derived from mixed logit models. The protocol is based on a two-round survey. Round 1 consists in a standard discrete choice experiment (DCE) sample survey, at the end of which competing models are estimated. Round 2 introduces additional respondents who, in addition to taking part in the DCE, are asked to identify the money value interval that better reflects their preferences using a set of values drawn from the estimated range of WTPs obtained from competing models estimated on the previous sample.