WHEN MEASUREMENT ERRORS CORRELATE WITH TRUTH - SURPRISING EFFECTS OF NONDIFFERENTIAL MISCLASSIFICATION
WHEN MEASUREMENT ERRORS CORRELATE WITH TRUTH - SURPRISING EFFECTS OF NONDIFFERENTIAL MISCLASSIFICATION
复制标题
DOI：
10.1097/00001648-199503000-00012
复制
发表时间：
1995-03-01
期刊：
影响因子：
5.4
通讯作者：
WACHOLDER, S
中科院分区：
文献类型：
--
作者：
WACHOLDER, S
大多数文献的影响，非微分误分类和错误的变量，要么地址二进制暴露变量或讨论连续变量的经典误差模型，其中的错误被假定为不相关的真实值。在这两种情况下，不完全测量的暴露总是衰减的关系，至少在单变量设置。此外，测量具有与暴露无关的误差的混杂因素，即使在完美地测量感兴趣的暴露的同时，也导致对混杂的部分控制。然而，对于流行病学中测量的许多变量，特别是那些基于自我报告的变量，误差通常与真实值相关，这些规则可能不适用。流行病学家需要警惕经典误差模型的偏差，因为即使误差不因疾病状态而异，测量不佳也可能偶尔解释阳性结果。
Most of the literature on the effect of nondifferential misclassification and errors in variables either addresses binary exposure variables or discusses continuous variables in the classical error model, where the error is assumed to be uncorrelated with the true value. In both of these situations, an imperfectly measured exposure always attenuates the relation, at least in the univariate setting. Furthermore, measuring a confounder with error independent of the exposure, even while measuring the exposure of interest perfectly, leads to partial control of the confounding. For many variables measured in epidemiology, particularly those based on self-report, however, errors are often correlated with the true value, and these rules may not apply. Epidemiologists need to be wary of deviations from the classical error model, since poor measurement might occasion ally explain a positive finding even when the error does not differ by disease status.