ANOVA F Test - Analysis of Variance

The ANOVA F-test compares the means across three or more independent groups to test whether at least one group's mean differs from the others. Example: comparing pain scores across three different treatment protocols.


Step 1 — Add your groups

The Data Input panel starts with two empty groups and lets you Add Group as many times as you need. Each group has:

  • A group name field (defaults to Group 1, Group 2, etc.)
  • A data text area where you paste or type the values

A Remove Group button lets you delete a group you no longer need.


Step 2 — Load data

You have three paths:

  • Manual entry in the group text areas
  • Excel Import — upload an Excel file; the system maps columns to groups automatically
  • Sample Data Generator — pick from four built-in scenarios:
    • Drug Dosage Comparison (Pain Reduction, Symptom Relief)
    • Teaching Methods (Exam Score, Satisfaction)
    • Exercise Type & Cholesterol (LDL Cholesterol, HDL Cholesterol)
    • Antibiotic Efficacy (4 Groups)

The Sample Data Generator can also download the chosen scenario as an Excel file.


Step 3 — Calculate ANOVA

Click Calculate ANOVA.


Step 4 — Read the results

The ANOVA Results panel shows:

  • F-value
  • p-value
  • Effect size — Eta-squared (η²) with an interpretation label
  • Descriptive Statistics — mean, SD, and n for every group
  • ANOVA Summary Table — Source / Sum of Squares / df / Mean Square for Between Groups, Within Groups, and Total

Step 5 — Visualisations

An AnovaCharts panel provides box plots for all groups side by side. Each box plot shows the median, the IQR (Q1–Q3), the whiskers, and individual outliers (using the standard 1.5×IQR fence rule).


Step 6 — Post-hoc comparisons

If ANOVA is significant (p < 0.05) and there are more than two groups, the panel includes a post-hoc comparisons table with each pair of groups compared using two parallel methods:

Comparison Tukey HSD Bonferroni
Group 1 vs Group 2 p-value p-value
Group 1 vs Group 3 p-value p-value
... ... ...

Step 7 — Discuss with the chatbot

A UniversalChatBot is available at the bottom for interpretation.


Statistical methods used

Test statistic

The one-way ANOVA decomposes total variance into between-groups and within-groups components:

Component Formula
Grand mean x̄ = Σ(nᵢ × x̄ᵢ) / N
SS_between Σ nᵢ × (x̄ᵢ − x̄)²
SS_within Σᵢ Σⱼ (xᵢⱼ − x̄ᵢ)²
SS_total SS_between + SS_within
df_between k − 1 (k = number of groups)
df_within N − k
MS_between SS_between / df_between
MS_within SS_within / df_within
F-statistic MS_between / MS_within

Hypothesis statement

  • H₀: μ₁ = μ₂ = ... = μₖ (all group means are equal)
  • H₁: At least one μᵢ differs
  • Significance level: α = 0.05

Effect size — Eta-squared (η²)

Computed as η² = SS_between / SS_total. Interpretation:

η² value Label
< 0.06 Small effect
0.06 – 0.14 Medium effect
≥ 0.14 Large effect

Post-hoc analysis

Two methods run in parallel when ANOVA is significant and there are more than two groups:

Method Approach
Tukey HSD Converts standard t-statistic to studentised range q via q = t × √2, computes p from studentised-range distribution
Bonferroni Multiplies uncorrected pairwise p-value by number of comparisons m = k(k−1)/2, capped at 1.0

Error handling

Condition Error message
Fewer than 2 valid groups "At least 2 groups with valid data are required"
A group has fewer than 2 values "Each group must have at least two values"
Within-group variance is 0 "Within-group variance is zero. All values in groups are identical."
Invalid input value "Invalid number found in group N"