Estimating p-value for Wilcoxon Signed-Rank Test

1. Apr 14, 2007

Ford

I'm trying to estimate the p-value range for a Wilcoxon Signed-Rank Test using a W-alpha table. For example:

Ho: M1 = M2
Ha: M1 < M2
alpha = 0.05
n = 17

I calculated W = 62 and the rejection region W <= 41. So I would reject Ho.

But I'm also required to know how to come to a conclusion using the estimated p-value that's supposed to come from the W-alpha table. In a t-test using a t-table I would look at the t-table at the row with the right degrees of freedom and see where my test statistic would fall in. When I try to do this with a Wilcoxon test and a W-alpha table it gives me conclusion that contradicts the one I get from the test itself.

Any ideas on how to estimate the p-value properly from the W-alpha table? Thanks.

2. Apr 25, 2007

EnumaElish

How do you know that the rejection region is W <= 41? "41" is nowhere on the table on row "n=17".

Last edited: Apr 25, 2007