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

Which statistical procedure cannot be used for analyzing ordinal data?

1. Mann-Whitney

2. Chi-squared

3. Unpaired t test

4. Wilcoxon rank sum

The correct answer is: Unpaired t test

The unpaired t-test is specifically designed for analyzing continuous data that is normally distributed and compares the means of two independent groups. Ordinal data, on the other hand, represents ranks or ordered categories without assuming equal intervals between values. Since the unpaired t-test relies on the calculation of means and assumes a normal distribution, it is not appropriate for ordinal data, where the focus should be on medians or ranks. In contrast, procedures such as the Mann-Whitney and Wilcoxon rank sum tests are non-parametric tests that are suitable for ordinal data because they rank the data instead of working with the means. The Chi-squared test also handles categorical data effectively, making it applicable to ordinal data when assessing relationships or frequencies. These methods respect the ranked nature of ordinal data, thereby making them appropriate choices for analysis.