CONFOUNDING AND MISCLASSIFICATION
CONFOUNDING AND MISCLASSIFICATION
复制标题
DOI：
10.1093/oxfordjournals.aje.a114131
复制
发表时间：
1985-01-01
影响因子：
5
通讯作者：
ROBINS, JM
中科院分区：
文献类型：
--
作者：
GREENLAND, S;ROBINS, JM
作者检查了一些最近提出的用于确定何时调整与 mlssassification 相关的协变量的标准，并表明这些标准是不正确的。特别是，他们表明，当存在错误分类时，协变量控制有时会增加净偏差，即使协变量在完美分类下会成为混杂因素，即使协变量是分类的决定因素。因此，用于控制混杂的方法不能充分处理由于错误分类而导致的偏差。所提供的例子还表明，当存在错误分类时，决定是否控制协变量的“估计变化”标准可能会系统性地产生误导。这些结果表明，在决定是否控制协变量时，有必要考虑错误分类的程度。
The authors examine some recently proposed criteria for determining when to adjust for covariates related to mlsdassification, and show these criteria to be incorrect In particular, they show that when misdassffication is present, covariate control can sometimes increase net bias, even when the covariate would have been a confounder under perfect classification, and even if the covariate is a determinant of classification. Thus, bias due to misclasstflcation cannot be adequately dealt with by the methods used for control of confounding. The examples presented also show that the “change-in-estimate” criterion for deciding whether to control a covariate can be systematically misleading when mis-classification is present These results demonstrate that it is necessary to consider the degree of misdassiflcation when deciding whether to control a covariate.