An example of a research setting where the Mann-Whitney U would be both appropriate and informative would be looking at pain scale numbers in two different groups of patients on a med-surg floor. Group 1 could have a new pain medication on board treating the pain while group 2 has the standard pain medication on board. Measuring these two groups on a numerical scale to see how the pain rates, you could use the Mann-Whitney U test to determine the difference in pain levels between the two: Refer to one your peer's example of a research setting for a non-parametric test. Which assumptions of a non-parametric test do you see it meeting? Are any of those questionable? If there are, how might you modify the design/setup to mitigate these issues