Skip to content

Wilcoxon V = 0

4 messages · Naomikb, David L Carlson, Peter Ehlers

#
I am running a pairwise wilcoxon signed rank test, and I am not sure how to
interpret the result.  I would like to see if there is a difference between
the values in conditions a and b. It doesn't seem possible to have a V = 0,
but a significant p value. Am I doing something wrong? 

The command I used is this:
 wilcox.test(x=a$x,y=b$x,paired=TRUE)


The output looks like this:
	Wilcoxon signed rank test

data:  a$x and b$x
V = 0, p-value = 0.0009766
alternative hypothesis: true location shift is not equal to 0 

The raw data is here:

a
   participant        condition            x
34           1 fused.mu.control           315.5677
35           2 fused.mu.control           516.8781
36           3 fused.mu.control           362.8108
37           4 fused.mu.control           482.6022
38           5 fused.mu.control           423.6593
39           6 fused.mu.control           404.9358
40           8 fused.mu.control           342.7314
41           9 fused.mu.control           466.7561
42          10 fused.mu.control           467.5002
43          11 fused.mu.control           434.0221
44          12 fused.mu.control           480.2990


b
   participant   condition      x
45           1 fused.multi           373.0648
46           2 fused.multi           517.5159
47           3 fused.multi           390.2658
48           4 fused.multi           497.2303
49           5 fused.multi           519.7867
50           6 fused.multi           499.7950
51           8 fused.multi           345.8077
52           9 fused.multi           479.1607
53          10 fused.multi           539.5856
54          11 fused.multi           446.9569
55          12 fused.multi           569.5171




--
View this message in context: http://r.789695.n4.nabble.com/Wilcoxon-V-0-tp4637675.html
Sent from the R help mailing list archive at Nabble.com.
#
It all depends on which order you put the values. Try this

wilcox.test(x=b$x,y=a$x,paired=TRUE)

Either one is an extreme. A non significant value would be in the middle. 


----------------------------------------------
David L Carlson
Associate Professor of Anthropology
Texas A&M University
College Station, TX 77843-4352
#
On 2012-07-24 14:36, Naomikb wrote:
Not understanding the test?
V is the sum of the ranks[*] of those pairs for which a$x > b$x;
there are none, so V = 0. You can check the p-value with the
[dp]signedrank() function.

[*] ranks of absolute differences.

Peter Ehlers