Atribución-NoComercial 4.0 InternacionalTrujillo Oyola, LeonardoAponte Nonzoque, Leonardo2019-06-242019-06-242012-02https://repositorio.unal.edu.co/handle/unal/11248Se exploró esta propiedad en cada una de las propuestas, mediante el estudio del desempeño en términos del comportamiento de sus tasas de error tipo I, estimadas con métodos empíricos bajo diferentes condiciones experimentales. Los resultados obtenidos en las simulaciones realizadas para cada propuesta, mostraron una tendencia favorable hacia el cumplimiento de la propiedad conservativa en una de las modificaciones realizadas. / Abstract. In this thesis, two permutation-based tests are proposed in order to assess the equality of means under heteroscedasticity. The proposed statistic test correspond to two modifications to one very well known in the literature in order to fulfill the conservation property of a hypothesis test. This particular property was analytically studied through the analysis of the type-I error rates being estimated under different experimental conditions. The obtained results in several simulations for each proposed statistic shown this property is preserved under the two proposed modifications.Abstract. In this thesis, two permutation-based tests are proposed in order to assess the equality of means under heteroscedasticity. The proposed statistic test correspond to two modifications to one very well known in the literature in order to fulfill the conservation property of a hypothesis test. This particular property was analytically studied through the analysis of the type-I error rates being estimated under different experimental conditions. The obtained results in several simulations for each proposed statistic shown this property is preserved under the two proposed modifications.application/pdfspaDerechos reservados - Universidad Nacional de Colombiahttp://creativecommons.org/licenses/by-nc/4.0/51 Matemáticas / Mathematics61 Ciencias médicas; Medicina / Medicine and health62 Ingeniería y operaciones afines / Engineering63 Agricultura y tecnologías relacionadas / AgricultureTrabajo de grado - Maestríahttp://bdigital.unal.edu.co/8656/info:eu-repo/semantics/openAccessTasas de error tipo IHeterogeneidad de varianzasPrueba de JamesSimulación Monte CarloConservación / Type I error ratesHeterogeneity of variancesJames’ testMonte Carlo SimulationConservative