Accession Number:

ADA030790

Title:

Goodness of Fit Tests for Composite Hypotheses Based on an Increasing Number of Order Statistics

Descriptive Note:

Technical rept.

Corporate Author:

CORNELL UNIV ITHACA NY SCHOOL OF OPERATIONS RESEARCH AND INFORMATION ENGINEERING

Personal Author(s):

• Jakobovits, Ray H.

Report Date:

1976-09-01

Pagination or Media Count:

167.0

Abstract:

Concern is shown with sequences of tests of the composite null hypothesis that the distribution function of an observed sample is a member of a specified scale-location parameter family, where the scale and location parameters are unknown and unspecified. First, a test of this hypothesis proposed by L. Weiss is studied. Then several tests for the most common special case--that of normality are analyzed.

Descriptors:

• *ASYMPTOTIC NORMALITY
• *FITTING FUNCTIONS(MATHEMATICS)
• *STATISTICAL TESTS
• DISTRIBUTION FUNCTIONS
• CHI SQUARE TEST
• ORDER STATISTICS
• STATISTICAL DISTRIBUTIONS
• NORMALIZING(STATISTICS)

Subject Categories:

• Statistics and Probability

Distribution Statement:

APPROVED FOR PUBLIC RELEASE