期刊
AMERICAN STATISTICIAN
卷 57, 期 4, 页码 285-288出版社
AMER STATISTICAL ASSOC
DOI: 10.1198/0003130032341
关键词
Anderson-Darling test; Cramer-von-Mises test; kernel density estimation; Monte Carlo test; Shapiro-Wilk test
Interpretation of normal probability plots is not always straightforward for the inexperienced data analyst. In the finance literature a plot of empirical and fitted normal densities on the log scale is frequently preferred as a graphical diagnostic for normality. This article describes the construction of this type of plot, and suggests a refinement that can facilitate its interpretability with small samples. A Monte Carlo test for normality arises naturally as a by-product of this methodology.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据