Skip to main content

5.3 Normal test

The normaltest command runs a selection of normal distribution tests for specified variables, or entire datasets if no variables are listed. For each test, the parameter value and p-value are stated. The tests that are run are skewness, kurtosis, s-k (not adjusted), Jarque-Bera and Shapiro-Wilk.

Syntax:

normaltest <variable list> [if <condition>]

Example of a test for normal distribution for the variable innt19 (occupational income measured in 2019). P-values ​​lower than 0.05 mean that the distribution is not normally distributed, and vice versa: