Principal investigator: James F. Troendle, Ph.D.
Nonparametric methods that use empirical likelihood are extremely robust and require few assumptions. Methods that use empirical likelihood are typically quite computational and require careful numerical strategies to be successful. By employing such methods, researchers have found likelihood ratio tests for many important statistical problems that have resultedin robust hypothesis-testing procedures.
Past work includes application of nonparametric and semiparametric models to test the nonparametric Behrens-Fisher hypothesis. More recently, studies have found a robust test for group mean effect in a 2Xk heteroscedastic ANOVA model.
DESPR Collaborators
· Kai Fun Yu, Ph.D.
Selected Publications
Troendle JF. (In press). Testing for group effect in a 2Xk heteroscedastic ANOVA model. Biometrical Journal.
Fokianos K & Troendle JF. (2007). Inference for the relative treatment effect with the Density Ratio Model. Statistical Modeling: An International Journal, 7, 155-173.
Troendle JF & Yu KF. (2006). Likelihood approaches to the nonparametric two-sample problem with right censored data. Statistics in Medicine, 25, 2284-2298. [Abstract]
Troendle JF. (2002). A likelihood ratio test for the nonparametric Behrens-Fisher Problem. Biometrical Journal, 44, 813-824.