Hi! I'm doing one and two sample nonparametric tests for the median using wilcox test. For a one-sample test I use:> wilcox.test(x, mu =50 (or whatever), y=NULL,correct=TRUE)For two-sample test I use:> wilcox.test(x,y,correct=TRUE)The problem is when I try to duplicate problems from textbooks, I get p-values that are much different from the examples from the literature. They are off by as much as 30% to 40%. Not even close. Using an "exact" argument doesn't change the p-value. What am I doing wrong? Thanks for any help. cvandy Charles H Van deZande [[alternative HTML version deleted]]
Charles Van deZande <cvandy26 <at> gmail.com> writes:> I'm doing one and two sample nonparametric tests for the median using wilcox > test. For a one-sample test I use: > > wilcox.test(x, mu =50 (or whatever), y=NULL,correct=TRUE) > For two-sample test I use: > > wilcox.test(x,y,correct=TRUE) > The problem is when I try to duplicate problems from textbooks, I get > p-values that are much different from the examples from the literature. > They are off by as much as 30% to 40%. Not even close. > Using an "exact" argument doesn't change the p-value.> What am I doing wrong?You have not posted the textbook example you tested. You can use dput to compactly post the data samples. Also check package coin. Dieter