Testing the normality of errors in regression models with a forward approach

被引:2
|
作者
Coin, Daniele [1 ]
机构
[1] Univ Turin, Dept Appl Math & Stat, I-10100 Turin, Italy
关键词
Forward Search; test for normality; outlier; Least Median of Squares;
D O I
10.1016/j.econlet.2006.03.009
中图分类号
F [经济];
学科分类号
02 ;
摘要
This paper presents a new robust method for testing the normality of errors in classical linear model. It gives information about both the true distribution of the errors of most of the data and the percentage of contaminated data. (c) 2006 Elsevier B.V All rights reserved.
引用
收藏
页码:323 / 329
页数:7
相关论文
共 50 条