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.
Reviewed for accuracy by Manoj Kumar, Mathematics Educator
ANOVA Post-Hoc Calculator
Calculator
Adjust values & calculateEnter your values below. Every result is computed in your browser — no data is sent to any server.
Formula: F = MSB / MSW = (SSB / (k-1)) / (SSW / (N-k))
Worked example — F(2,12) = 32.66, p < 0.05 | All three teaching methods produce significantly different scores
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 SSB = 5(74-79.67)^2 + 5(80-79.67)^2 + 5(85-79.67)^2 = 304.93 SSW = (sum of squared deviations within each group) = 56.0 MSB = 304.93/2 = 152.47, MSW = 56.0/12 = 4.67 F = 152.47/4.67 = 32.66 Tukey 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 SSB = 5(6-10.67)^2 + 5(11-10.67)^2 + 5(15-10.67)^2 = 204.93 SSW = 40.0 F = 102.47/3.33 = 30.74 Post-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 is eta-squared and what does it tell me about effect size?
Eta-squared is a measure of effect size in ANOVA that represents the proportion of total variance in the dependent variable that is explained by the independent variable (group membership). It is calculated as SSB/SST, where SSB is the sum of squares between groups and SST is the total sum of squares. Values range from 0 to 1, with conventional benchmarks being 0.01 for small effects, 0.06 for medium effects, and 0.14 for large effects. While statistical significance tells you whether an effect exists, eta-squared tells you how large that effect is. A statistically significant result with a tiny eta-squared may not be practically meaningful, while a large eta-squared in a small sample might not reach significance.
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.
What does a non-significant ANOVA result mean?
A non-significant ANOVA result (p-value greater than alpha) means you fail to reject the null hypothesis that all group means are equal. However, this does not prove the means are identical. Non-significance could indicate truly equal means, insufficient sample size to detect a real difference (low statistical power), high within-group variability that masks between-group differences, or violation of ANOVA assumptions. Before concluding no effect exists, calculate statistical power to determine if your sample was large enough to detect a meaningful difference. A power analysis can tell you how many observations per group you would need. Also consider the effect size, as even non-significant results with moderate effect sizes may warrant further investigation with larger samples.
How do I report ANOVA results in a research paper?
Standard APA format for reporting ANOVA results includes the F-statistic, degrees of freedom, p-value, and effect size. For example: F(2, 27) = 15.34, p < .001, eta-squared = 0.53. The first degree of freedom is df-between (k-1) and the second is df-within (N-k). If post-hoc tests were conducted, report which test was used and the significant pairwise differences with their mean differences and confidence intervals. Include descriptive statistics (means and standard deviations) for each group, preferably in a table. Also report whether assumptions were checked and met. A complete results section helps readers evaluate the validity and practical significance of your findings beyond just statistical significance.
References
Reviewed for accuracy by Manoj Kumar, Mathematics Educator · Editorial policy
Related Calculators
🧮ANOVA Calculator
Calculate anova with inputs, formulas, and instant results.
🧮Annulus Area Calculator
Calculate annulus area with inputs, formulas, and instant results.
🧮Area Calculator
Calculate area with inputs, formulas, and instant results.
🧮Area of a Rectangle Calculator
Calculate the area, perimeter, and diagonal of a rectangle. Find missing sides from known area. Convert between metric and imperial area units.
🧮Area of Crescent Calculator
Calculate area of crescent with inputs, formulas, and instant results.
🧮Center of Mass Calculator
Calculate center of mass with inputs, formulas, and instant results.
🧮Centroid Calculator
Calculate centroid with inputs, formulas, and instant results.
🧮Chord Length Calculator
Calculate chord length with inputs, formulas, and instant results.