Variations on kurtosis estimation with different statistics software

Kurtosis, or a distribution’s fourth moment, is used to describe distributions and belongs to some normality contrast tests. Most of the statistical softwares include kurtosis, which makes the estimation to be relatively easy. Nevertheless, for a same data set, statistical softwares can provide a different result. The objective of this work is to depict kurtosis estimation differences between statistical softwares that are most commonly used among agricultural economists. Two samples were used to compare kurtosis coefficient estimation differences with nine different statistic softwares. Results shows that these differences are not due to a mistaken estimation procedure, but mainly because the term kurtosis is used wrongly. In conclusion, when working with a small sample size and adjustment factor is not considered, there is a 20% probability of making a mistaken conclusion.

Saved in:
Bibliographic Details
Main Author: Barrantes Aguilar, Luz Elena
Format: Digital revista
Language:spa
Published: Escuela de Agronegocios, Tecnológico de Costa Rica 2019
Online Access:https://revistas.tec.ac.cr/index.php/eagronegocios/article/view/4456
Tags: Add Tag
No Tags, Be the first to tag this record!
id oai:ojs.pkp.sfu.ca:article4456
record_format ojs
spelling oai:ojs.pkp.sfu.ca:article44562020-08-25T20:46:11Z Variations on kurtosis estimation with different statistics software Diferencias en la estimación del coeficiente de curtosis en diferentes softwares estadísticos Barrantes Aguilar, Luz Elena cuarto momento exceso de curtosis exceso de curtosis ajustado software estadístico fourth momentum excess kurtosis adjusted excess kurtosis statistic software Kurtosis, or a distribution’s fourth moment, is used to describe distributions and belongs to some normality contrast tests. Most of the statistical softwares include kurtosis, which makes the estimation to be relatively easy. Nevertheless, for a same data set, statistical softwares can provide a different result. The objective of this work is to depict kurtosis estimation differences between statistical softwares that are most commonly used among agricultural economists. Two samples were used to compare kurtosis coefficient estimation differences with nine different statistic softwares. Results shows that these differences are not due to a mistaken estimation procedure, but mainly because the term kurtosis is used wrongly. In conclusion, when working with a small sample size and adjustment factor is not considered, there is a 20% probability of making a mistaken conclusion. La curtosis, o cuarto momento de una distribución, se emplea para describir una distribución y forma parte de algunos contrastes de normalidad. La mayoría de los paquetes estadísticos la incluyen, por lo que su cálculo es sencillo. Sin embargo, para un mismo grupo de datos, los resultados que proporcionan los diferentes programas estadísticos pueden diferir. Con el objetivo de explicar las diferencias en la estimación de la curtosis entre los paquetes estadísticos de mayor uso para los economistas agrícolas se emplearon dos muestras y se realizó una comparación en la estimación del coeficiente curtosis con nueve softwares estadísticos diferentes. Los resultados mostraron que las diferencias no se deben a errores de cálculo, sino a que el término se emplea de manera incorrecta por la mayoría de programas y se estimó que existe aproximadamente un 20% de probabilidad de llegar a una conclusión equivocada al trabajar con muestras pequeñas y no considerar el factor de ajuste. Escuela de Agronegocios, Tecnológico de Costa Rica 2019-07-01 info:eu-repo/semantics/article info:eu-repo/semantics/publishedVersion paper artículo application/pdf application/epub+zip text/html https://revistas.tec.ac.cr/index.php/eagronegocios/article/view/4456 10.18845/rea.v5i2.4456 e-Agronegocios; Vol. 5 Núm. 2 (2019): Julio-Diciembre, 2019 2215-3462 2215-3462 spa https://revistas.tec.ac.cr/index.php/eagronegocios/article/view/4456/4953 https://revistas.tec.ac.cr/index.php/eagronegocios/article/view/4456/4954 https://revistas.tec.ac.cr/index.php/eagronegocios/article/view/4456/4046 Derechos de autor 2019 Luz Elena Barrantes Aguilar, Licda. http://creativecommons.org/licenses/by-nc/4.0
institution TEC CR
collection OJS
country Costa Rica
countrycode CR
component Revista
access En linea
databasecode rev-eagronegocios-cr
tag revista
region America Central
libraryname Sistema de Bibliotecas del TEC Costa Rica
language spa
format Digital
author Barrantes Aguilar, Luz Elena
spellingShingle Barrantes Aguilar, Luz Elena
Variations on kurtosis estimation with different statistics software
author_facet Barrantes Aguilar, Luz Elena
author_sort Barrantes Aguilar, Luz Elena
title Variations on kurtosis estimation with different statistics software
title_short Variations on kurtosis estimation with different statistics software
title_full Variations on kurtosis estimation with different statistics software
title_fullStr Variations on kurtosis estimation with different statistics software
title_full_unstemmed Variations on kurtosis estimation with different statistics software
title_sort variations on kurtosis estimation with different statistics software
description Kurtosis, or a distribution’s fourth moment, is used to describe distributions and belongs to some normality contrast tests. Most of the statistical softwares include kurtosis, which makes the estimation to be relatively easy. Nevertheless, for a same data set, statistical softwares can provide a different result. The objective of this work is to depict kurtosis estimation differences between statistical softwares that are most commonly used among agricultural economists. Two samples were used to compare kurtosis coefficient estimation differences with nine different statistic softwares. Results shows that these differences are not due to a mistaken estimation procedure, but mainly because the term kurtosis is used wrongly. In conclusion, when working with a small sample size and adjustment factor is not considered, there is a 20% probability of making a mistaken conclusion.
publisher Escuela de Agronegocios, Tecnológico de Costa Rica
publishDate 2019
url https://revistas.tec.ac.cr/index.php/eagronegocios/article/view/4456
work_keys_str_mv AT barrantesaguilarluzelena variationsonkurtosisestimationwithdifferentstatisticssoftware
AT barrantesaguilarluzelena diferenciasenlaestimaciondelcoeficientedecurtosisendiferentessoftwaresestadisticos
_version_ 1763178549161754624