Testing for Equiprobability Using a Multi-Nomial or Poisson Sample.
Abstract
The well-known chi-square test is discussed for testing equiprobability when the expected number of observations per cell is not large. The results are used to give a justification for the Poisson Index of Dispersion test of fit for the Poisson distribution. A chi-square type statistic is studied for testing equiprobability and Poisson-fit when the frequencies associated with zero are missing and thus the sample size is unknown. Some applications are discussed in detail where the samples are incomplete. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Jul 01, 1976
- Accession Number
- ADA031971
Entities
People
- Lars Holst
Organizations
- University of Wisconsin–Madison