Bounds on Distributions Arising in Order Restricted Inference: The Partially Ordered Case.

Abstract

In testing hypotheses involving order restrictions on a collection of parameters, distributions arise which are mixtures of chi-squared or beta distributions. In general, the mixing coefficients are quite intractable even for a moderate number of populations. Stochastic upper and lower bounds are obtained for mixtures which arise in Bartholomew's tests for homogeneity of normal means with the alternative restricted by a quasi ordering. These bounds are applicable in the dual-testing situation, that is in testing the order restriction as a null hypothesis; in testing order restrictions in exponential families, Poisson intensities and multinomial parameters; and in some nonparametric settings. These can also be applied to obtain the least favorable configuration for testing equality of two multinomial populations with a stochastic ordering alternative. (Author)

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Sep 01, 1985
Accession Number
ADA159872

Entities

People

  • F. T. Wright

Tags

Communities of Interest

  • C4I

DTIC Thesaurus Topics

  • Algorithms
  • Binomials
  • Computations
  • Convolution
  • Data Science
  • Decomposition
  • Homogeneity
  • Hypotheses
  • Information Science
  • Mathematics
  • Probability
  • Random Variables
  • Sequences
  • Statistical Algorithms
  • Statistics
  • Theorems
  • Two Dimensional

Fields of Study

  • Mathematics

Readers

  • Statistical inference.

Technology Areas

  • AI & ML
  • AI & ML - Bayesian Inference
  • AI & ML - Machine Learning Algorithms