Back to main page

Shapiro-Wilk Expanded test for normally distributed sample

The test is accurate for sample sizes between 12 and 5000, although sample sizes of at least 20 are recommended. You can run the test for 10 or more samples. You can read more about the test at Real Statistics. You can also read about skewness and kurtosis at Real Statistics.
Tests the hypotheses:
H0 The sample is normally distributed  
H1 The sample is not normally distributed  

 Data Entry

Sample A:
Significance level α: 
Upload CSV file: