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

For parametric data analysis involving two groups where pairs are measured, which test should be used?

1. Unpaired t test

2. ANOVA

3. Paired t test

4. Mann-Whitney

The correct answer is: Paired t test

When analyzing parametric data involving two groups with paired measurements, the paired t test is the appropriate choice. This test is specifically designed to compare the means of two related groups. In scenarios where the same subjects are measured under two different conditions (for example, before and after treatment), the paired t test accounts for the fact that the data points are not independent. By focusing on the differences between paired observations, the paired t test leverages the inherent relationship within the data, providing more reliable results compared to tests that assume independent samples. This statistical test evaluates whether the mean difference between the pairs is significantly different from zero, which is crucial for determining the effect of interventions or treatments in clinical studies. In contrast, other tests like the unpaired t test and ANOVA are not suitable for paired data. The unpaired t test compares means from two independent groups, while ANOVA is used for comparing means across three or more groups or for analyzing multiple factors simultaneously. The Mann-Whitney test assesses differences between two independent groups and is a non-parametric alternative but does not cater to the paired nature of the measurements either. Hence, for situations involving paired measurements, the paired t test is the correct framework to apply.