In Stata, how do I test the normality of a variable?
In Stata, you can test normality by either graphical or numerical methods. The former include drawing a stem-and-leaf plot, scatterplot, box-plot, histogram, probability-probability (P-P) plot, and quantile-quantile (Q-Q) plot. The latter involve computing the Shapiro-Wilk, Shapiro-Francia, and Skewness/Kurtosis tests.
The examples below are for the variable score:
| Graphical methods | |
|---|---|
| Command | Plot drawn |
. stem score |
stem-and-leaf |
. dotplot score |
scatterplot |
. graph box score |
box-plot |
. histogram score |
histogram |
. pnorm score |
P-P plot |
. qnorm score |
Q-Q plot |
| Numerical methods | |
| Command | Test conducted |
. swilk score |
Shapiro-Wilk |
. sfrancia score |
Shapiro-Francia |
. sktest score |
Skewness/Kurtosis |
Be aware that in these tests, the null hypothesis states that the variable is normally distributed.
If you have questions about using statistical and mathematical software at Indiana University, email UITS Research Analytics (formerly known as the Stat/Math Center). Research Analytics is located on the IU Bloomington campus at 410 N. Park Avenue and is open for consultation by appointment Monday-Friday 9am-5pm. For more, visit Research Analytics on the web, or call 812-855-4724 (IUB) or 317-278-4740 (IUPUI).
Last modified on October 18, 2012.







