Inference Concerning Partially Dominated Ordering in a r by k Contingency Table

• 발행기관: 한국자료분석학회
• 저자명: Myongsik Oh
• 간행물 정보: 『Journal of The Korean Data Analysis Society (JKDAS)』Vol.6 No.1, 89~97쪽, 전체 9쪽
• 주제분류: 자연과학 > 통계학
• 파일형태: PDF
• 발행일자: 2004.02.28

영문 초록

We consider the statistical inference concerning partially dominated ordering which is imposed on two rows (or possibly column) in a r by k contingency table. This type of association occurs frequently in many practical situations. Suppose we collect data and classify them into k categories according to their certain attributes twice. These form two probability vectors which may be dependent. We are interested in detecting whether two probability vectors are partially dominated ordered. Order restricted inferential approach is used to solve this problem and a real example is analyzed.

목차

I. Introduction

II. Constrained Maximum Likelihood Estimation

III. Likelihood Ratio Tests

IV. An Example

V. Discussion

References

키워드

chi-bar-square constrained maximum likelihood estimation likelihood ratio test multinomial parameter partially dominated ordering