Monte Carlo Benchmarks for Discrete Response Valuation Methods: Reply
回应Poe和Vossler的评论，承认先前研究中Turnbull下界均值与英式拍卖均值相等性检验有误，但进一步分析表明离散响应模型对支付意愿的估计仍存在负偏倚。
<i>This response to a comment by Poe and Vossler on our earlier research (Huang and Smith 1998) agrees that our test of the equality of the Turnbull lower bound mean and English Auction mean was incorrect. However, this result was part of the review of literature and did not affect any of the primary conclusions of our earlier research. Moreover, further analysis of the sampling experiments suggests that the selection of true model specifications, hypothesized values for key preference parameters, and choice of estimating equations did not preclude negative biases in the estimates of willingness to pay from discrete response models.</i>