The Joint Distribution of Pearson's X-Squared Statistics when Maximum Likelihood Estimated are Used.
Abstract
Limiting joint distributions are studied in Pearson's tests for goodness of fit when parameters are estimated from the original data before grouping. The joint distribution is bounded in terms of distributions appropriate when the parameters are known or are estimated from grouped data. The several distributions are characterized in the limit in terms of partitioned Wishart matrices of unit rank having parameters as appropriate.
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 01, 1974
- Accession Number
- ADA002241
Entities
People
- Donald R. Jensen
Organizations
- Virginia Tech