Calculates the Shapiro-Wilk test for Normality (R.W. Payne).
Option
PRINT = string tokens |
What to print (test); default test |
|---|
Parameters
DATA = variates |
Samples of data to be tested for Normality |
|---|---|
W = scalars |
Saves the Shapiro-Wilk W statistic for each sample |
PROBABILITY = scalars |
Saves the probability for W under the assumption that the data are Normal |
Description
WSTATISTIC uses the Shapiro-Wilk test to assess whether a sample of data comes from a Normal distribution. The data values must be supplied, in a variate, using the DATA parameter. By default WSTATISTIC prints the statistic, W, with its probability value under the assumption that the data are Normal. (So a low probability indicates that the data are unlikely to be from a Normal distribution.) The printed output can be supressed by setting option PRINT=*. The test statistic can be saved, in a scalar, using the W parameter, and its probability can similarly be saved using the PROBABILITY parameter.
Option: PRINT.
Parameters: DATA, W, PROBABILITY.
Method
WSTATISTIC calculates the statistic and its probability using the methods of Royston (1993, 1995).
Action with RESTRICT
The DATA variate can be restricted to assess a subset of the data.
References
Royston, P. (1993). A toolkit for testing for non-normality in complete and censored samples. The Statistician, 42, 37-43.
Royston, P. (1995). A remark on Algorithm AS 181: the W-test for Normality. Applied Statistics, 44, 547-551.
See also
Directive: DISTRIBUTION.
Procedures: EDFTEST, NORMTEST.
Commands for: Basic and nonparametric statistics.
Example
CAPTION 'WSTATISTIC example','Data from Royston (1995)'; STYLE=meta,plain
VARIATE [VALUES=4.2,4.9,5.2,5.3,6.7,6.7,7.2,7.5,8.1,8.6,8.8,9.3,\
9.5,10.3,10.8,11.1,12.2,12.5,13.3,15.1,15.3,16.1,19.0,19.5] Glucose
WSTATISTIC Glucose