Gerhard Rudolph
Abstract
Exact and asymptotic tests of independence are derived for a contingency table based on a half-multinomial, half-negative-multinomial model. These tests are shown to be identical to the well-known corresponding tests for a multinomial model. Some properties are also considered of the probability distribution on which the present model is based.
OpenAlex reports 2 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Exact and asymptotic tests of independence are derived for a contingency table based on a half-multinomial, half-negative-multinomial model. These tests are shown to be identical to the well-known corresponding tests for a multinomial model. Some properties are also considered of the probability distribution on which the present model is based.
Key concepts: Multinomial distribution, Contingency table, Mathematics, Independence (probability theory), Table (database), Statistics, Negative multinomial distribution, Multinomial probit