What is a post-hoc test?

• A. A test to determine overall effect
• B. A test to adjust for multiple comparisons
• C. A follow-up test to determine which specific group differences are significant
• D. A test to replace ANOVA

Answer: (C)

Post-hoc tests are follow-up analyses performed after you run an ANOVA with more than two groups when the overall test shows a significant difference. They’re used to identify exactly which pairwise group means differ, not just that some difference exists somewhere among the groups. Because you’re making multiple comparisons, these tests adjust the error rate to keep the chance of a false positive under control. Common methods include Tukey’s HSD and Bonferroni adjustments. They are not used to assess the overall effect (that’s the role of the initial ANOVA) and they’re not meant to replace ANOVA; they clarify which specific groups differ. They’re usually not needed if the overall test is not significant. In short, a post-hoc test is a follow-up to determine which specific group differences are significant.