American Board of Surgery Qualifying Exam (ABS QE) Practice Test

For non-parametric, unpaired data analysis involving two groups, what statistical test is recommended?

1. Paired t test

2. ANOVA

3. Wilcoxon rank sum

4. Mann-Whitney

The correct answer is: Mann-Whitney

When analyzing non-parametric, unpaired data from two groups, the Mann-Whitney test is a suitable choice because it is specifically designed to compare differences between two independent samples. This test assesses whether the distribution of the two groups differs in a statistically significant way without assuming a normal distribution, which is an important consideration for non-parametric data. The Mann-Whitney test ranks all the data from both groups together, then evaluates the sum of the ranks for each group to determine if there is a significant difference between them. It is particularly useful when the sample sizes are small or when the data does not meet the assumptions required for parametric tests, such as the t-test. The Wilcoxon rank sum test is actually another name for the Mann-Whitney test, so while both terms might be interchangeable, the direct answer to the question posed is more about understanding the context of when and why the Mann-Whitney test is appropriate for unpaired, non-parametric data analysis. In summary, the Mann-Whitney test effectively compares two groups with independent samples when the data does not conform to parametric test assumptions.