Power Analysis Calculator: One-Way ANOVA
Plan the sample size for a one-way ANOVA the way G*Power does — in the browser. Enter the number of groups, the expected effect size Cohen's f, your alpha and target power, and get the required sample size per group computed from the exact noncentral F distribution (identical parameters to G*Power's 'ANOVA: fixed effects, omnibus' module and R's pwr.anova.test). Switch to 'find power' to check what an already-collected sample can detect. If you only have an eta squared (η²) from the literature, convert it to f with our Effect Size Converter first: f = √(η²/(1 − η²)) — for example η² = .06 gives f = 0.25, the conventional 'medium' effect.
AI Report
Let AI interpret your results: a downloadable Word document in APA 7 / business report format.
Frequently asked questions
How many participants do I need for a one-way ANOVA?
For three groups, a medium effect (f = 0.25), α = .05 and 80% power, you need 53 participants per group (159 total). More groups or smaller effects increase the requirement; the calculator gives the exact number for your design.
How do I convert eta squared to Cohen's f?
f = √(η² / (1 − η²)). So η² = .01 → f ≈ 0.10 (small), η² = .06 → f ≈ 0.25 (medium), η² = .14 → f ≈ 0.40 (large). Our Effect Size Converter tool does this automatically, including the reverse direction.
How do I report an ANOVA power analysis in APA format?
For example: "An a priori power analysis (one-way ANOVA, three groups, f = 0.25, α = .05, power = .80) indicated a required sample size of 53 per group (N = 159)." The AI Report writes this methods-section paragraph for you.
Does this work for repeated-measures or factorial ANOVA?
This module covers the between-subjects one-way (omnibus) design. Repeated-measures and factorial designs need additional parameters (correlation among measures, number of measurements); those modules are on our roadmap.