El test de la T de Student, ¿sólo en poblaciones normales?
Loading...
Identifiers
Publication date
Authors
Advisors
Editors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
[ES] El test de la T de Student sirve para estimar la media de una población normalmente distribuida cuando el tamaño muestral es pequeño y la varianza es desconocida. Cuando los datos no son normales la potencia del test de la T de Student puede ser inferior a otros tests, ya sean paramétricos o no paramétricos. En este trabajo estudiaremos en qué contrastes, tanto en una población como en dos, la prueba T se ve superada o no por otros tests alternativos. Para ello, nos apoyaremos en ejemplos, donde se aportarán las representaciones de las funciones de potencia para dichos tests. En algunos casos será un problema complejo saber la distribución que seguirá el estadístico, pudiendo ser éste también difícil de calcular. Por este motivo, para realizar buenas aproximaciones de las funciones de potencia nos ayudaremos con simulaciones realizadas en R
[EN] The Student’s T test is used to estimate the mean of a normally distributed population when the sample size is small and the variance is unknown. When the data are not normal, the power of the Student’s T test may be lower than in other tests, whether parametric or non-parametric. In this project we will study in which contrasts, whether in one or two populations, the T test is surpassed or not by other alternative tests. To do this, we will rely on examples, where the representations of the power functions of these tests will be provided. Occasionally it will be a complex problem to know which distribution the statistic will follow, and it may also be difficult to calculate. Therefore, we will simulate in R to achieve an approximate representation of the power functions
[EN] The Student’s T test is used to estimate the mean of a normally distributed population when the sample size is small and the variance is unknown. When the data are not normal, the power of the Student’s T test may be lower than in other tests, whether parametric or non-parametric. In this project we will study in which contrasts, whether in one or two populations, the T test is surpassed or not by other alternative tests. To do this, we will rely on examples, where the representations of the power functions of these tests will be provided. Occasionally it will be a complex problem to know which distribution the statistic will follow, and it may also be difficult to calculate. Therefore, we will simulate in R to achieve an approximate representation of the power functions
Description
Traballo Fin de Grao en Matemáticas. Curso 2020-2021
Keywords
Bibliographic citation
Relation
Has part
Has version
Is based on
Is part of
Is referenced by
Is version of
Requires
Sponsors
Rights
Atribución-NoComercial-CompartirIgual 4.0 Internacional



