Analysis and Experimentation, Microsoft Corporation, Redmond, Washington.
Department of Statistics, North Carolina State University, Raleigh, North Carolina.
Biometrics. 2020 Jun;76(2):664-669. doi: 10.1111/biom.13148. Epub 2019 Nov 19.
For ordinal outcomes, the average treatment effect is often ill-defined and hard to interpret. Echoing Agresti and Kateri, we argue that the relative treatment effect can be a useful measure, especially for ordinal outcomes, which is defined as , with and being the potential outcomes of unit under treatment and control, respectively. Given the marginal distributions of the potential outcomes, we derive the sharp bounds on which are identifiable parameters based on the observed data. Agresti and Kateri focused on modeling strategies under the assumption of independent potential outcomes, but we allow for arbitrary dependence.
对于有序结果,平均处理效应通常难以定义和解释。我们赞同 Agresti 和 Kateri 的观点,认为相对处理效应是一种有用的度量方法,尤其是对于有序结果而言。相对处理效应定义为 ,其中 和 分别是单位 在治疗和对照下的潜在结果。给定潜在结果的边缘分布,我们根据观测数据推导出 的有界估计值,这些边界估计值是可识别的参数。Agresti 和 Kateri 专注于在独立潜在结果假设下的建模策略,但我们允许任意依赖关系。