More info
Description / Abstract:
Scope and Purpose
This standard describes statistical hypothesis-testing
techniques that are designed to assist in the assessment of the
assumption of normality by making use of all the individual
observed values in a sample. There is no requirement that the
number of observed values be "large', nor is it required that the
observations be grouped into intervals with the attendant loss of
information about the inter-value spacings. The procedures are
called "omnibus" because they are appropriate for detecting
departures from normality caused by any of several characteristics,
including skewness (nonsystmetry) and kurtosis ("peakedness").
Empirical studies have demonstrated that these procedures generally
are more effective than several others in wide currency (further
details appear in Appendix A).