Gardner IA, Greiner M (2006) Receiver-operating characteristic curves and likelihood ratios: improvements over traditional methods for the evaluation and application of veterinary clinical pathology tests.Altman DG, Machin D, Bryant TN, Gardner MJ (Eds) (2000) Statistics with confidence, 2 nd ed.2007 except when the predicitive value is 0 or 100%, in which case a Clopper-Pearson confidence interval is reported. 2000.Ĭonfidence intervals for the predictive values are the standard logit confidence intervals given by Mercaldo et al. Sensitivity, specificity, disease prevalence, positive and negative predictive value as well as accuracy are expressed as percentages.Ĭonfidence intervals for sensitivity, specificity and accuracy are "exact" Clopper-Pearson confidence intervals.Ĭonfidence intervals for the likelihood ratios are calculated using the "Log method" as given on page 109 of Altman et al. $$ Accuracy = Sensitivity \times Prevalence + Specificity \times (1 - Prevalence) $$
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |