The Problem Of Assessing And Achieving Normality: An Application To Environmental Data
Price
Free (open access)
Transaction
Volume
10
Pages
8
Published
1995
Size
592 kb
Paper DOI
10.2495/AIR950302
Copyright
WIT Press
Author(s)
J. Mateu
Abstract
It is well known that many statistical works in environmental analysis are based on methods that need normal variables. Data transformation methods capable of achieving normality of distributions have a crucial role in statistical analysis, especially towards an efficient application of techniques such as analysis of variance and multiple regression analysis. Different techniques to test the normality of distributions have been studied in this paper. Since normal distribution has value zero in skewness and kurtosis, we develope confidence intervals over them to contrast their nullity. Moreover different nonparametric tests are applied such as Chi-square and K-S. On the other hand we suggest Box-Cox transformations as a method to achieve normal variab
Keywords