When running the Legacy Dialogs for two independent samples using SPSS, there are many tests like Mann-Whitney U test, Moses extreme reactions and Wald-Wolfowitz runs. What is the difference between these tests? Can be used for any two independent samples?
So far I have performed a Mann-whitney on my (non-normal, unequal variances) data. This has come out insignificant (not what I was hoping for) but when I perform a Moses test it came out significant. So basically which one I should use?
Dear Dr Abdualkader, If you go to the favourite links page of my website there is a link to an electronic statistical textbook which is quite good so you may find definitions there. I know there are definitions for the Whitney U test and Wald-Wolfowitz, however I cannot recall seeing the Moses or Legacy mentioned but they maybe there, I just have not looked these up previously. Thanks Deborah Hilton Statistics Online
http://sites.google.com/site/deborahhilton/
The Mann-Whitney U test is also known as the Wilcoxon rank-sum test. The MW was formulated to test samples from populations A and B: Ho: P(A >= x) = P(B >=x) vs. Ha: P(A >= x) >= P(B >=x). Put another way, let D = abs(A - X). The assumptions are that 1) the distribution of the D is symmetric; Di is independed of Dj; all of the Ds have the same mean; and, the measurement scale of the Ds interval or continuous.
Given two independent samples, the Wald-Wolfowitz test evaluates whether the samples come from populations with the same distribution or from different distributions for at least some x. Conceptually, this is similar to the Mann-Whitney U test.
The Moses extreme reaction test is for two independent samples on a continuous scale, and tests this:
Ho: Extreme values are equally likely in both populations vs. Ha: Extreme values are more likely to occur in the population from which the sample with the larger range was drawn.
Which test to use depends on what you want to know.
Hello Dr Abdualkader,
I've got exactly the same problem with my data (left skewed distribution)=> Mann-whithney nonsignificant and Moses test significant.
How do you interpret and write out results from the Moses test?
Thanks!
When performing a Mann-whitney you need to test for homogeneity of variance in your sample. If the assumption of homogeneity is not met, then you must use a different test.
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS