| 多元分布下单边备择假设的两步检验 |
| |
| 引用本文： | 谢民育,吴茗,熊明.多元分布下单边备择假设的两步检验[J].数理统计与管理,2011,30(2):227-233. |
| |
| 作者姓名： | 谢民育 吴茗 熊明 |
| |
| 作者单位： | 华中师范大学数学与统计学学院,湖北武汉,430079 |
| |
| 基金项目： | 国家自然科学基金资助项目(项目编号10571070) |
| |
| 摘 要： | x~2检验是熟知的拟合优度检验.它适用于任意的备择假设,尤其是适用于一般的备择假设"H_0不真"的情况.但是,当备择假设是一个单边假设时,x~2检验没有用到单边假设所提供的信息.因此,一个充分利用单边备择假设信息的检验是有可能优于传统的x~2检验的.本文就是利用单边备择假设的信息给出了一个两步检验,并通过一个例子,详细介绍了两步检验法的实施,说明了其应用.
|
| 关 键 词： | x~2检验 单边备择假设 两步检验 |
A Two-step Test with theOne-sideAlternative in Multinomial Distribution |
| |
| XIE Min-yu,WU Ming,XIONG Ming.A Two-step Test with theOne-sideAlternative in Multinomial Distribution[J].Application of Statistics and Management,2011,30(2):227-233. |
| |
| Authors: | XIE Min-yu WU Ming XIONG Ming |
| |
| Institution: | XIE Min-yu WU Ming XIONG Ming (Mathematics and Statistics Department of Hua Zhong Normal University,Hubei Wuhan 430079,China) |
| |
| Abstract: | The chi-square test is a well-known goodness-of-fit test.It is available for arbitrary alternative hypothesis,particularly for a general alternativeH_0 is not true.However,when the alternative is aone-side 8hypothesis,the chi-square test does not use the information offered by the one-side hypothesis. Therefore it is possible that an appropriate test,which uses the information of the one-side alternative, may be better.This note gives a two-step test with the information offered by the one-side alternative.... |
| |
| Keywords: | chi-square test one-side alternative hypothesis two-step test |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
