Question

5. In researching the effect of working in a cubicle as opposed to a private office on the productivity of middle managers in a large public agency, you have collected data on individuals who are cubicle users and others who have private offices. If you realize that you cannot assume homogeneity of variance, what test should you use to analyze your results?

A. A Wilcoxon T Test

B. A related-samples T test

C. A Mann-Whitney U Test

D. An independent-samples T test

Answer 1

Solution:

The correct option is C. A Mann-Whitney U Test

Explanation:

In the given question, we are given that in researching the effect of working in a cubicle as opposed to a private office on the productivity of middle managers in a large public agency, you have collected data on individuals who are cubicle users and others who have private offices. If you realize that you cannot assume homogeneity of variance.

Since the samples are independent so we could have used an Independent-samples T-test but since we cannot assume the homogeneity of variance, so we cannot use an independent-samples T-test. Therefore, option D is eliminated.

Looking at the option A. a Wilcoxon T-test, it is used when we dependent samples under consideration and when the assumptions of normality are not met. So we cannot use Wilcoxon T-test.

Looking at option B. a related-samples T-test, it is a parametric test and is used when the samples are dependent. So this cannot be used here.

The option C. A Mann-Whitney U test is a non-parametric test and is an alternative test for an independent-samples t-test, therefore, we can use a Mann-Whitney U test here.

Hence the option C. Mann-Whitney U Test is correct