Misclassification Errors in 2 x 2 Tables with One Margin Fixed: Some Further Comments

Publisher Website
Google Scholar
Abstract
It is demonstrated that, in 2 x 2 tables with one margin fixed, misclassification errors can either increase or decrease the expected values of commonly used measures of association. It is shown that the usual assumption of equal misclassification probabilities for cases and controls is most likely not true in realistic situations. It is suggested that in any study of this kind misclassification probabilities be determined separately for cases and controls on the basis of sub samples.

This publication has 3 references indexed in Scilit: