Wilcoxon Single Sample T-test
In performing a nonparametric single sample t-test in Stata, you are comparing a group that violates the standard assumptions for a t-test to a known test-value. The syntax is simply signrank [depvar]=[testvalue]
For instance if I had a sample of words written per minute and I wanted to know if my sample (an Administrative training class, made-up data) typed faster than the general us population (average of about 35 WPM according to a highly inaccurate google search), the command would be signrank wpm=35 which produces the following output:
The output includes how the data fell in distance from the median and, at the bottom, the two-tailed value of the test. In this case, my secretary sample was significantly different from the general population.
If I had used a directional hypothesis, like, my secretaries will type significantly faster than average, then I would use the corresponding One-sided test as show below:
Back to Nonparametric Tests
Back to Home