ANOVA Post-Hoc Calculator
Free Anovapost hoc Calculator for statistics. Enter values to get step-by-step solutions with formulas and graphs. Includes formulas and worked examples.
Formula
F = MSB / MSW = (SSB / (k-1)) / (SSW / (N-k))
Where F is the F-statistic, MSB is Mean Square Between groups, MSW is Mean Square Within groups, SSB is Sum of Squares Between, SSW is Sum of Squares Within, k is the number of groups, and N is the total number of observations.
Worked Examples
Example 1: Comparing Three Teaching Methods
Problem: Test scores from three teaching methods: Group 1 (Traditional): 72, 75, 78, 71, 74. Group 2 (Online): 80, 82, 78, 81, 79. Group 3 (Hybrid): 85, 87, 83, 86, 84. Test at alpha = 0.05.
Solution: Means: G1 = 74.0, G2 = 80.0, G3 = 85.0. Grand Mean = 79.67\nSSB = 5(74-79.67)^2 + 5(80-79.67)^2 + 5(85-79.67)^2 = 304.93\nSSW = (sum of squared deviations within each group) = 56.0\nMSB = 304.93/2 = 152.47, MSW = 56.0/12 = 4.67\nF = 152.47/4.67 = 32.66\nTukey HSD comparisons identify all pairs as significantly different.
Result: F(2,12) = 32.66, p < 0.05 | All three teaching methods produce significantly different scores
Example 2: Drug Dosage Effectiveness
Problem: Blood pressure reduction for three dosages: Low (5, 7, 6, 4, 8), Medium (10, 12, 11, 9, 13), High (14, 16, 15, 13, 17). Compare at alpha = 0.05.
Solution: Means: Low = 6.0, Medium = 11.0, High = 15.0. Grand Mean = 10.67\nSSB = 5(6-10.67)^2 + 5(11-10.67)^2 + 5(15-10.67)^2 = 204.93\nSSW = 40.0\nF = 102.47/3.33 = 30.74\nPost-hoc: All pairwise differences exceed HSD threshold.
Result: F(2,12) = 30.74, p < 0.05 | All dosage levels differ significantly in blood pressure reduction
Frequently Asked Questions
What is ANOVA and when should I use it?
ANOVA (Analysis of Variance) is a statistical test used to compare means across three or more groups simultaneously. Instead of running multiple t-tests between pairs (which inflates the Type I error rate), ANOVA tests whether at least one group mean differs significantly from the others using a single test. It works by partitioning total variance into between-group variance and within-group variance, then comparing them via the F-statistic. ANOVA is appropriate when you have a continuous dependent variable and a categorical independent variable with three or more levels. Common applications include comparing treatment effects in clinical trials, product performance across manufacturing processes, and student test scores across teaching methods.
What is a post-hoc test and why is it needed after ANOVA?
A post-hoc test is conducted after a significant ANOVA result to determine exactly which group means differ from each other. ANOVA only tells you that at least one group is different, but not which specific pairs differ. Post-hoc tests perform pairwise comparisons while controlling for the family-wise error rate, which is the probability of making at least one Type I error across all comparisons. Without post-hoc correction, comparing three groups would involve three pairwise tests, each at alpha = 0.05, giving a combined error rate of about 14.3%. Popular post-hoc methods include Tukey HSD, Bonferroni, Scheffe, and Dunnett tests, each with different strengths depending on your research design.
How does the Tukey HSD post-hoc test work?
The Tukey Honestly Significant Difference (HSD) test compares all possible pairs of group means while maintaining the family-wise error rate at the desired alpha level. It calculates a critical difference threshold using the studentized range distribution (q-distribution), the mean square error from the ANOVA, and the sample sizes. If the absolute difference between any two group means exceeds this HSD value, those groups are declared significantly different. Tukey HSD is most appropriate when group sample sizes are equal and you want to test all pairwise comparisons. It is considered moderately conservative, falling between the liberal Fisher LSD test and the more conservative Bonferroni correction in terms of statistical power.
What assumptions must be met for ANOVA to be valid?
ANOVA requires three main assumptions to produce valid results. First, independence means observations within and between groups must be independent of each other, typically ensured through random sampling and assignment. Second, normality means the dependent variable should be approximately normally distributed within each group. ANOVA is fairly robust to moderate violations of normality, especially with larger sample sizes due to the Central Limit Theorem. Third, homogeneity of variances (homoscedasticity) means the variance of the dependent variable should be roughly equal across all groups. This can be tested using Levene test or Bartlett test. If this assumption is violated, you can use Welch ANOVA or non-parametric alternatives like the Kruskal-Wallis test.
How do I choose between different post-hoc tests?
The choice of post-hoc test depends on your specific research situation. Tukey HSD is the standard choice when you want to compare all possible pairs with equal sample sizes. Bonferroni correction is more flexible and works well with unequal sample sizes but becomes very conservative with many groups. Scheffe test is the most conservative and is appropriate when testing complex contrasts beyond simple pairwise comparisons. Dunnett test is optimal when you only need to compare each treatment group to a single control group. Games-Howell is preferred when the homogeneity of variances assumption is violated. For unequal sample sizes, Tukey-Kramer or Games-Howell are better choices than standard Tukey HSD. Consider your error tolerance and the number of comparisons being made.
What is the difference between one-way and two-way ANOVA?
One-way ANOVA examines the effect of a single categorical independent variable on a continuous dependent variable. Two-way ANOVA examines the effects of two categorical independent variables simultaneously, plus their interaction effect. For example, one-way ANOVA might test whether drug type affects blood pressure, while two-way ANOVA could test whether both drug type and dosage level (and their interaction) affect blood pressure. Two-way ANOVA is more efficient because it can detect interaction effects that one-way ANOVA would miss entirely. ANOVA Post-Hoc Calculator implements one-way ANOVA with post-hoc tests, which is the most commonly used form in introductory statistics and many practical research scenarios.