Seconder of the vote of thanks to Evans & Didelez and contribution to the Discussion of 'Parameterizing and Simulating from Causal Models'
Seconder of the vote of thanks to Evans & Didelez and contribution to the Discussion of 'Parameterizing and Simulating from Causal Models'
复制标题
对埃文斯表示感谢的附议票
DOI：
10.1093/jrsssb/qkae013
复制
发表时间：
2024
影响因子：
--
通讯作者：
Silva R
中科院分区：
文献类型：
--
作者：
Silva R
因果推断中的许多统计问题涉及的概率分布与实际观察数据的概率分布不同；更复杂的是，感兴趣的对象通常是这种另一种概率分布的边际数量。这为统计推断带来了许多实际的复杂性，即使在问题是非参数识别的情况下也是如此。特别是，很难进行基于似然的推理，甚至很难从模型中进行一般的模拟。我们引入了“节俭的参数化”，它把兴趣的因果效应放在它的中心，然后围绕它建立模型的其余部分。我们这样做的方式提供了一种使用感兴趣的因果量来构建规则的、非冗余的参数化的方法。在离散变量的情况下，我们可以使用赔率比来完成参数化，而在连续的情况下，Copula是自然的选择；还讨论了其他的可能性。我们的方法允许我们构建和模拟具有参数指定的因果分布的模型，并使用基于似然的方法来拟合它们，包括完全贝叶斯方法。我们的建议包括平均因果效应和治疗对受试者的影响的参数，以及其他感兴趣的因果量。
Many statistical problems in causal inference involve a probability distribution other than the one from which data are actually observed; as an additional complication, the object of interest is often a marginal quantity of this other probability distribution. This creates many practical complications for statistical inference, even where the problem is non-parametrically identified. In particular, it is difficult to perform likelihood-based inference, or even to simulate from the model in a general way. We introduce the ‘frugal parameterization’, which places the causal effect of interest at its centre, and then builds the rest of the model around it. We do this in a way that provides a recipe for constructing a regular, non-redundant parameterization using causal quantities of interest. In the case of discrete variables, we can use odds ratios to complete the parameterization, while in the continuous case copulas are the natural choice; other possibilities are also discussed. Our methods allow us to construct and simulate from models with parametrically specified causal distributions, and fit them using likelihood-based methods, including fully Bayesian approaches. Our proposal includes parameterizations for the average causal effect and effect of treatment on the treated, as well as other causal quantities of interest.