A New Method of Approximating Normal and t-Tail Areas.

Abstract

The bounds of Boyd and Soms for the tail areas of the normal and t-distributions are used to obtain a new method of approximating the tail areas. The absolute and relative errors are given. A reasonable accurate and simple method of evaluating normal and t-tail areas and percentiles is described. This approach controls both absolute and relative errors, unlike the usual methods which only control the absolute error. The control of both errors is of importance in calculating Bonferroni percentiles and Bonferroni descriptive probability levels, which are useful in simultaneous inference. For example, a nonspecialist in statistics, unaware of the limitations of standard routines, might report a Bonferroni descriptive probability level that had a relative error of 100% or more. (Authr).

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Sep 01, 1984
Accession Number
ADA149457

Entities

People

  • A. P. Soms

Organizations

  • University of Wisconsin–Madison

Tags

Communities of Interest

  • C4I

DTIC Thesaurus Topics

  • Computers
  • Continents
  • Contracts
  • Geographic Regions
  • Mathematics
  • Military Research
  • North America
  • North Carolina
  • Probability
  • Procedures (Computers)
  • Standards
  • Statistics
  • United States
  • Universities
  • Wisconsin

Fields of Study

  • Mathematics

Readers

  • Regression Analysis.
  • Statistical inference.
  • Systems Analysis and Design

Technology Areas

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