Hi all,
I am doing ANOVA to compare tree different groups, ( n= 22,25,23). each group have normal distribution and ANOVA result significative (p= 0.03). ANOVA residuals are no- significatives so resiuals have homogeneus variances and errors have normal distribution.
but when I run tukeyHSD all groups result equal, despite ANOVA significances. So my question is why it would be?
I really appreciate any suggestion to have a reliable result.
TukeyHSD is more strict than other methods to find the pairwise differences. Your ANOVA result with a p-value of 0.03 is not overly enthusiastic wrt significance, so TukeyHSD is not seeing them. With only 3 groups (and thus 3 pairwise comparisons) use Fisher's LSD, that should make it a bit easier to see significant differences.
- Badges
- Science topic
- Similar topics
- Biological Science
- Kinesiology
- Running