Deriving welfare measures from discrete choice experiments: a response to Ryan and Santos Silva
回应评论，论证在离散选择实验（DCE）中，当选择存在不确定性时，使用Small和Rosen修正的补偿变差方法比传统卫生经济学方法更通用、更合适，后者依赖不现实的假设。
Abstract In this response we start by highlighting the key area of agreement between the commentaries and our original paper: if there is uncertainty regarding which alternative will be chosen, in a DCE or in the real world, then the compensating variation as modified for discrete data by Small and Rosen is the appropriate method of deriving welfare measures from DCEs. Both commentators point out circumstances in which the method traditionally used in the health economics arena may be consistent with the compensating variation. We show that these circumstances require a number of potentially unrealistic and ad hoc assumptions, and argue that using the traditional method could produce erroneous welfare estimates if these assumptions fail to hold in practice. We show that the compensating variation method can accommodate each of the special cases raised by the commentators and therefore is the more general and appropriate approach to deriving welfare measures from DCEs. We also respond to issues raised regarding the estimation of DCEs in general and our application to asthma medication in particular. Copyright © 2004 John Wiley & Sons, Ltd.