Advances and Applications in Statistics
Volume 2, Issue 3, Pages 233 - 248
(December 2002)
|
|
POWER-DIVERGENCE TYPE MEASURE OF DEPARTURE FROM UNIFORM ASSOCIATION FOR CROSS-CLASSIFICATIONS HAVING ORDERED CATEGORIES
Sadao Tomizawa (Japan) and Tomoyuki Hashimoto (Japan)
|
Abstract: For the analysis of
cross-classifications having ordered categories,
Tomizawa [Statistics and Probability Letters 11
(1991), 547-550] proposed a Kullback-Leibler
information type or Shannon entropy type measure
to represent the degree of departure from
uniform association. This paper proposes a
generalization of this measure. The measure
proposed is expressed by using the Cressie-Read
power-divergence or Patil-Taillie diversity
index based on the local odds ratios. It
includes Tomizawa’s [Statistics and Probability
Letters 11 (1991), 547-550] measure in a special
case. It would be useful when one wants to use a
single summary measure to measure distance of
the local odds ratios from uniformity, and also
useful for comparing the degrees of departure
from uniformity in several cross-classification
tables. The eye color-hair color data and three
kinds of unaided vision data are analyzed by
using the proposed measure. |
Keywords and phrases: local odds ratio, power-divergence, Shannon entropy. |
|
Number of Downloads: 7 | Number of Views: 1207 |
|