What is the parametric test for comparing three or more related samples?

What is the parametric test for comparing three or more related samples?

• A. One-way repeated measures ANOVA
• B. Two-way ANOVA
• C. Kruskal-Wallis
• D. Friedman test

Answer: (A)

Comparing three or more related samples with a parametric approach is accomplished with one-way repeated measures ANOVA. Because the same subjects are measured under multiple conditions, the observations are not independent, so the analysis accounts for this within-subject correlation. The test extends the idea of a paired t-test to more than two conditions by partitioning the total variance into within-subject (condition-related) and between-subject components, and it uses an F statistic to determine if the condition means are all equal. If the overall result is significant, you can follow with post-hoc comparisons to see which specific conditions differ. Key assumptions include that the differences between conditions are approximately normally distributed and that sphericity holds—the variances of the differences between all pairs of conditions are equal. When sphericity is violated, corrections like Greenhouse-Geisser or Huynh-Feldt are applied. For contrast, Kruskal-Wallis is for independent samples, Friedman is a nonparametric alternative for related samples, and two-way ANOVA involves two factors.

Comparing three or more related samples with a parametric approach is accomplished with one-way repeated measures ANOVA. Because the same subjects are measured under multiple conditions, the observations are not independent, so the analysis accounts for this within-subject correlation. The test extends the idea of a paired t-test to more than two conditions by partitioning the total variance into within-subject (condition-related) and between-subject components, and it uses an F statistic to determine if the condition means are all equal.

If the overall result is significant, you can follow with post-hoc comparisons to see which specific conditions differ. Key assumptions include that the differences between conditions are approximately normally distributed and that sphericity holds—the variances of the differences between all pairs of conditions are equal. When sphericity is violated, corrections like Greenhouse-Geisser or Huynh-Feldt are applied. For contrast, Kruskal-Wallis is for independent samples, Friedman is a nonparametric alternative for related samples, and two-way ANOVA involves two factors.