Descriptive measures for nominal categorical variables

被引：2
作者：
Biswas, Atanu [1 ]
Mandal, Saumen [2 ]
机构：
[1] Indian Stat Inst, Appl Stat Unit, Kolkata 700108, India
[2] Univ Manitoba, Dept Stat, Winnipeg, MB R3T 2N2, Canada
来源：
STATISTICS & PROBABILITY LETTERS | 2010年 / 80卷 / 11-12期
基金：
加拿大自然科学与工程研究理事会;
关键词：
Association; Entropy; Probability of incorrect guess; Variability;
D O I：
10.1016/j.spl.2010.02.012
中图分类号：
O21 [概率论与数理统计]; C8 [统计学];
学科分类号：
020208 ; 070103 ; 0714 ;
摘要：
Descriptive measures/statistics are tools to understand the features of a probability/frequency distribution. Not many descriptive measures/statistics have been attempted for nominal categorical data. The present paper is motivated to provide some sensible descriptive statistics, analogous to the variance, skewness, and kurtosis, for univariate nominal categorical data. The measures are illustrated by some real data. Some intuitive test procedures are also described in this context. (C) 2010 Elsevier B.V. All rights reserved.
引用
收藏
页码：982 / 989
页数：8
相关论文
共 39 条
1234