Medium
Post-hoc Tests
SPSS
Free
You run a one-way ANOVA in SPSS comparing average weekly caffeine intake (mg) across 3 groups: No Coffee, 1–2 Cups/Day, 3+ Cups/Day. The omnibus ANOVA is significant (p < .001). SPSS Tukey HSD output shows:
Multiple Comparisons (Tukey HSD)
- No Coffee vs 1–2 Cups/Day: Mean Difference = -85, Sig. = .012, 95% CI [-150, -20]
- No Coffee vs 3+ Cups/Day: Mean Difference = -210, Sig. < .001, 95% CI [-280, -140]
- 1–2 Cups/Day vs 3+ Cups/Day: Mean Difference = -125, Sig. = .003, 95% CI [-190, -60]
Which statement is correct?
A
Only No Coffee differs significantly from 3+ Cups/Day.
B
All three pairwise group differences are statistically significant.
C
None of the pairwise differences are significant because Tukey is too conservative.
D
You should ignore the Sig. column and use the mean differences only.