The nonparametric counterpart to the independent-samples t-test is which test?

• A. The Mann-Whitney U test
• B. The Sign test
• C. The Wilcoxon matched pairs test
• D. The Chi-square test

Answer: (A)

The main concept here is choosing a nonparametric test when comparing two independent groups. In a parametric independent-samples t-test, you’re looking at differences in means and you assume the data are roughly normally distributed within each group. When that normality assumption can’t be met, a nonparametric approach that doesn’t rely on the data’s shape is used. The Mann-Whitney U test does exactly this: it works with ranks rather than raw values and compares the overall distribution of scores between the two independent groups. If one group tends to yield higher values than the other, the ranks will reflect that difference, and the test will indicate a significant difference. This method is especially useful with ordinal data or when distributions are skewed or contain outliers.

The other options don’t fit this scenario. The Sign test and the Wilcoxon matched-pairs test are designed for paired or matched data, not independent groups. The Chi-square test is for categorical data, not continuous or ordinal data.