The Effect of Errors in Diagnosis and Measurement on the Estimation of the Probability of an Event
研究了基础数据中的误分类和测量误差如何影响逻辑回归和正态判别分类方法的渐近偏差与效率，发现误差会增大偏差并降低效率，且逻辑回归在有误差时表现相对更好。
Abstract This article investigates the effect of misclassification and measurement error in the basic data on the asymptotic bias and efficiency of the logistic regression (LR) and normal discrimination (ND) classification procedures. The effect of misclassification in a single binary independent variable on the bias and efficiency of both procedures is also presented. Typically, asymptotic bias increases and efficiency decreases as misclassification and measurement error increase. The performance of LR relative to ND is shown to be better in the presence of error than without error.