Bahadur Efficiencies of the Student's t Tests.

Abstract

In the one-sample problem of testing for the location of a symmetric distribution, previous computations of the Bahadur efficiency of the t-test were based on the slope of the t-test under the restricted null hypothesis of normality. The author rectifies this shortcoming by finding the slope of the t-test under the full nonparametric null hypothesis. This is done by solving an extremal problem in large deviations for the t-statistic. It is also shown that the maximum slope in this one-sample problem is attained by a linear signed-rank test.

Document Details

Document Type
Technical Report
Publication Date
Mar 01, 1975
Accession Number
ADA012216

Entities

People

  • Douglas H. Jones
  • Jayaram Sethuraman

Organizations

  • Florida State University

Tags

DTIC Thesaurus Topics

  • Computations
  • Computing-Related Activities
  • Data Science
  • Efficiency
  • Information Science
  • Interdisciplinary Science
  • Mathematical Analysis
  • Mathematics
  • Normality
  • Statistical Analysis

Fields of Study

  • Mathematics

Readers

  • Regression Analysis.