Distinguishing 'missing at random'' and ''missing completely at random''
Distinguishing 'missing at random'' and ''missing completely at random''
复制标题
DOI：
10.2307/2684656
复制
发表时间：
1996-08-01
影响因子：
1.8
通讯作者：
Basu, S
中科院分区：
文献类型：
--
作者：
Heitjan, DF;Basu, S
随机缺失（MAR）和完全随机缺失（MCAR）是可验证性条件-当它们成立时，它们保证可以进行某些类型的推断，而无需求助于复杂的缺失数据建模。在这篇文章中，我们回顾MAR，MCAR的定义，以及他们最近的推广。我们应用的定义，在三个常见的不完整的数据的例子，通过模拟演示的后果，偏离可验证性。我们认为，从业者谁面对潜在的nonconciliable不完整的数据必须考虑的推理模式和条件的性质时，决定调用哪一个conciliability条件。
Missing at random (MAR) and missing completely at random (MCAR) are ignorability conditions-when they hold, they guarantee that certain kinds of inferences may be made without recourse to complicated missing-data modeling. In this article we review the definitions of MAR, MCAR, and their recent generalizations. We apply the definitions in three common incomplete-data examples, demonstrating by simulation the consequences of departures from ignorability. We argue that practitioners who face potentially nonignorable incomplete data must consider both the mode of inference and the nature of the conditioning when deciding which ignorability condition to invoke.