Which test is appropriate for comparing two related samples, such as measurements on the same individuals under two conditions?

Which test is appropriate for comparing two related samples, such as measurements on the same individuals under two conditions?

• A. Two-Sample T-Test
• B. Paired T-Test
• C. Chi-Square Test
• D. ANOVA

Answer: (B)

When measurements come from the same individuals under two conditions, the observations are paired and you should analyze the differences within each pair. The paired t-test is the right choice because it directly tests whether the average difference between the two related measurements is zero, effectively using each person as their own control and reducing variability from individual differences. This provides more power to detect a real effect than treating the two conditions as independent groups. The two-sample t-test assumes independence between groups, which isn’t the case with related samples. The chi-square test handles categorical data, not mean differences. ANOVA compares means across three or more groups (and a repeated-measures version can handle related samples with more conditions), but for exactly two related measurements, the paired t-test is the simplest and most appropriate. If the difference scores aren’t roughly normally distributed, a nonparametric alternative like the Wilcoxon signed-rank test can be used.

When measurements come from the same individuals under two conditions, the observations are paired and you should analyze the differences within each pair. The paired t-test is the right choice because it directly tests whether the average difference between the two related measurements is zero, effectively using each person as their own control and reducing variability from individual differences. This provides more power to detect a real effect than treating the two conditions as independent groups. The two-sample t-test assumes independence between groups, which isn’t the case with related samples. The chi-square test handles categorical data, not mean differences. ANOVA compares means across three or more groups (and a repeated-measures version can handle related samples with more conditions), but for exactly two related measurements, the paired t-test is the simplest and most appropriate. If the difference scores aren’t roughly normally distributed, a nonparametric alternative like the Wilcoxon signed-rank test can be used.