Criterion Power Analyses

α (and the associated decision criterion) is computed as a function of

• 1-β,
• the effect size, and
• a given sample size.

Criterion analyses are alternatives to post-hoc power analyses after a study has already been conducted. They may be reasonable whenever the control of α is less important than the control of β.

In case of goodness-of-fit tests for statistical models, for example, it is most important to minimize the β error risk of wrong decisions in favor of the model (H₀). Researchers could thus use criterion analyses to compute the significance level α compatible with β = .05 for a small effect size.