Beginners question about the Man-whitney test

Moderators: statman, Analyst Techy, andris, Fierce, GerineL, Smash

dalej
Posts: 2
Joined: Wed Apr 13, 2016 2:39 pm

Beginners question about the Man-whitney test

Postby dalej » Wed Apr 13, 2016 2:47 pm

Im comparing the differences between 2 values from to different groups. The results i retrieved from the MW test gave a significant P value for when comparing the values of 8.1 and 30.1 BUT NO SIGNIFICANT P VALUE for 2 and 42.9 which is clearly a increased difference. These results im putting forward are important and I just wanted to know if anyone can explain these results and if this is normal for this kind of test. Thanks in advance for any help!

PS if this helps this was the data used for the 2 groups:

Group A
0% 0% 19.4% 8% 6% 12% 11.5% MEAN 8.1%

0% 0% 13.8% 0% 0% 0% 0% MEAN 2%


Group B
23% 14.2% 20% 60% 0% 41.6% 52.6% MEAN 30.2%

100% 0% 0% 100% 0% 100% 0% MEAN 42.9%
GerineL
Moderator
Posts: 1477
Joined: Tue Jun 10, 2008 4:50 pm

Re: Beginners question about the Man-whitney test

Postby GerineL » Wed Apr 13, 2016 3:23 pm

The mann whitney u test does not compare means but compares distributions (more or less) between groups.
dalej
Posts: 2
Joined: Wed Apr 13, 2016 2:39 pm

Re: Beginners question about the Man-whitney test

Postby dalej » Wed Apr 13, 2016 3:26 pm

What test do you recommend for comparing 2 percentages from 2 different groups?

PS. I didnt put the means into spss, I just put them there to reference the data to my original question!
GerineL
Moderator
Posts: 1477
Joined: Tue Jun 10, 2008 4:50 pm

Re: Beginners question about the Man-whitney test

Postby GerineL » Thu Apr 14, 2016 6:58 am

This may just be a t-test, but depends on the distribution within your 2 vars.

Who is online

Users browsing this forum: No registered users and 4 guests

cron