Tjur’s R2for logistic regression models is the same as Youden’s index for a 2×2 diagnostic table

The linear logisticmodel [1] provides a basis for statistical analysis of the dependence of a binary outcome variable on one or more explanatory variables. As such, logistic regression is often deployed in epidemiological applications where the relationship of interest is betweentrue (i.e., gold-standard assessed) disease statusclassified in two categories (denoted non-cases 'nc', numerically coded 0; and cases 'c', coded 1), and outcomes calculated on the basis of one or more risk factors. In practice, such relationships may be usedto provide support for a decision either to make or to withhold an intervention whencarrying out a gold-standard assessment is deemed too risky, too time-consuming or too expensive and instead, data for previously identified and quantified risk factorscan be deployed in the decision process.

from ! ORL Sfakianakis via paythelady.61 on Inoreader http://ift.tt/2zpBe2q
via IFTTT