Wilcoxon Signed Ranks Test
If you feel a bit insecure about calculating and interpreting the inferential statistics required for this assessment, do not panic. Vassarstats is a really easy program to use, and it does not require you to go back and review the calculus.
First, go to the Vassarstats site.
To calculate the Wilcoxon Signed Ranks Test, click on "Ordinal Data" in the left-hand margin.
The Wilcoxon Signed Ranks Test is used for repeated measures designs. Click on the test. The program will ask for the size (n) of the sample. In order to practice using the test, enter a size of 9.
Enter the following data in the boxes for Xa: 15, 12, 13, 15, 17, 18, 14, 19, 20. Then enter the following data for Xb: 25, 27, 29, 30, 30, 22, 28, 29, 30.
After you have entered the data, click on the "calculate" button.
Now for the final step. The IA’s require that students predict the direction of the hypothesis – that is, their hypothesis argues that x increases the effect on y or decreases the effect on y. Therefore, you are going to use the values under "directional test."
To interpret the results, you need to compare the W score – in this case -45 to the values in the table below. You will see that for a directional test at a significance level of p<0.05 for a sample of 9 a W of 29 or higher is required. You should always use the absolute value of the W score – that is, ignore the fact that in this case it is negative. Since our value is greater than 29, we can say that our data is significant at p< 0.05. If we then move across the chart, we see that our data is greater than 35, 39, and 43. Therefore, we can conclude that our data is significant at p< 0.005. That is, there is less than a .05 percent chance that our results are due only to chance
not have exceeded the value for p<0.05. Then we would have concluded that our data is not significant, and that our null hypothesis is retained.