Using Man Whitney U Test Statistic in Research
This guide will explain, step by step, how to run the Mann-Whitney U Test in SPSS software by using an example.
The Mann–Whitney U test is a popular test for comparing two independent samples. That is to say, It is a nonparametric test, as the analysis is undertaken on the rank order of the scores and so does not require the assumptions of a parametric test. (Source)
The Mann-Whitney U test is a nonparametric test (data are not normally distributed). In other words, we use a Mann-Whitney test to determine whether there is a difference in a continuous variable between two independent groups (categorical variable). For example, we want to compare whether there is a difference in IQ score (continuous variable) between male and female students (categorical variable).
Assumptions for the Mann-Whitney U Test
The only assumptions for carrying out a Mann-Whitney test are that the two groups must be independent so that the dependent
variable is ordinal or numerical (continuous). However, in order to report the difference between groups as medians, the shape of the distributions of the dependent variable by the group must be similar. Certainly, it doesn’t matter if the distributions have a different location on the x-axis, they just have to be a similar shape. (Source-PDF)
Your variables should have:
The Mann Whitney U test is often used when the assumptions of the t-test have been violated. Thus it is useful if:
- The dependent variable is ordinally scaled instead of interval or ratio.
- The assumption of normality has been violated in a t-test (especially if the sample size is small.)
- The assumption of homogeneity of variance has been violated in a t-test
An Example for Man Whitney U Test
We wanted to examine whether there is a difference between a male and a female student on the Math test score. Therefore, we collected data from 25 students. The math test score is our dependent variable; gender is our independent variable with two groups (male, female).
Null hypothesis:
There is no difference in the Math test score between males and females.
Alternative hypothesis:
There is a difference in Math test scores between males and females.
Please watch the SPSS video Tutorial on how to run the Mann-Whitney U Test in SPSS.
