Inference Concerning an Association in a 2 by k Contingency Table when Poisson Sampling Model is Applied

• 발행기관: 한국자료분석학회
• 저자명: Myongsik Oh
• 간행물 정보: 『Journal of The Korean Data Analysis Society (JKDAS)』Vol.5 No.4, 777~785쪽, 전체 9쪽
• 주제분류: 자연과학 > 통계학
• 파일형태: PDF
• 발행일자: 2003.12.30

영문 초록

We consider the statistical inference concerning an association in a 2 by k contingency table whose cell frequencies follow Poisson distribution. The association is related to weak majorization between two sets of Poisson parameters. A simple transformation of parameter space enables us to solve this seemingly intractable problem. Maximum likelihood estimates of Poisson parameters and likelihood ratio test for testing against weak majorization are discussed.

목차

I. Introduction

II. Constrained Maximum Likelihood Estimation

III. Likelihood Ratio Tests

IV. An Illustrative Example

V. Remarks

References

키워드

Chi-bar-square isotonic regression Poisson weak majorization