期刊名称：Journal of Modern Applied Statistical Methods
出版年度：2011
卷号：10
期号：1
页码：14
出版社：Wayne State University
摘要：Ecole Supérieure de Commerce de Tunis. Fisher’s exact test is adapted to handle the misclassified data arising from comparing two binomial populations. The bias-adjusted odds ratio is proposed to account for misclassification errors. Its expected power depends in a nonlinear way on the true sensitivity and specificity of the classification method. The data taken from the no conviction rate of criminality for two types of twin populations was used to illustrate how to calculate true sensitivity and specificity and the expected power of the adjusted odds ratio.
关键词：Fisher’s exact test; misclassification; power function; odds ratio; sensitivity; specificity
