Power analysis¶
Works out the relationship between significance level, statistical power, effect size, and sample size for a chosen test — fixing three of them to solve for the fourth. No data set is needed.
When to use it¶
Before a study, to plan the sample size needed to detect an effect.
After (or while planning), to check the power of a given design, or the effect size it can detect.
Inputs (all numbers you type)¶
Test type — Two-sample t-test, Paired / one-sample t-test, One-way ANOVA, or Correlation.
Solve for — Sample size, Power, or Effect size.
Alpha (e.g. 0.05), Power (e.g. 0.80), Effect size, Sample size — provide the three you know.
Number of groups — for ANOVA.
Tails — two-sided or one-sided.
Output¶
The solved quantity with the inputs that produced it.
A power-vs-sample-size curve for the chosen test.
Notes¶
Effect sizes follow Cohen’s conventions: d for t-tests, f for ANOVA, r for correlation.
t-tests and ANOVA use the standard non-central distributions; correlation uses the Fisher-z approximation.