TESTS OF NORMALITY.

reportActive / Technical Report | Accession Number: AD0697289 | Need Help?

Abstract:

The paper gives percentage points for four goodness-of-fit statistics, of Kolmogorov-type i.e., based on the comparison of empirical and theoretical distribution functions. The points are to be used for a test of normality, when mean and variances must be estimated from the data. Simple modifications of the statistics are also given, which make the tests very easy to apply. The statistics are compared for power with the W statistic they compare much better than previously reported, and for certain long-tailed alternatives seem to be as good as the W-statistic. Author

Author(s):
Stephens, M. A.
Author Organization(s):
STANFORD UNIV CALIF DEPT OF STATISTICS
Descriptive Note:
Technical rept.,
Supplementary Note:
Prepared in cooperation with McGill Univ., Montreal (Quebec).
Pagination:
0026

Security Markings

DOCUMENT & CONTEXTUAL SUMMARY

Distribution:
Approved For Public Release

RECORD

Collection: TR
Identifying Numbers
Report Number(s):
TR-152
Contract/Grant Number(s):
Nonr-225(52)
Project Number(s):
NR-342-022
 Subject Terms
Communities of Interest:
No COI(s) Identified
Descriptor(s):
(*STATISTICAL TESTS, POWER), STATISTICAL DISTRIBUTIONS, DISTRIBUTION FUNCTIONS, SAMPLING, TABLES(DATA)
Field(s)/Group(s):
Statistics and Probability
Keyword(s):
ASYMPTOTIC DISTRIBUTIONS, CHI SQUARE TEST, KOLMOGOROV SMIRNOV TEST, NORMAL DENSITY FUNCTIONS, NORMALITY
Report Date:
1969 Nov 10