Test of suitability of normal distribution based on Cramer-von Mises statistics
Since normal distributions are the most important ones in statistics, there are large number of tests for normality. However they have less some drawbacks. Some of these tests are simple but suitable for some situations. In this study, the traditional Cramer-von Mises test statistics is modified bas...
Saved in:
Main Authors: | , |
---|---|
Format: | Article |
Published: |
UKM
2007
|
Subjects: | |
Online Access: | http://eprints.utm.my/id/eprint/7641/ |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Summary: | Since normal distributions are the most important ones in statistics, there are large number of tests for normality. However they have less some drawbacks. Some of these tests are simple but suitable for some situations. In this study, the traditional Cramer-von Mises test statistics is modified based on Weibull formula. The new goodness-of-fit test is compared with the traditional Anderson-Darling (AD), Cramer von-Mises (CR), Kolmogorov-Smirnov (KS) and Shapiro-Wilk (SW) test statistics. A simulation study using several different distributions shows that the proposed test is very powerful for testing normality. |
---|