Statistical Power Measure

A Statistical Power Measure is a statistical test performance measure that quantifies the probability that a statistical hypothesis test will correctly reject a false null hypothesis.

• AKA: Test Power, Statistical Test Power, Power of a Test, 1-β.
• Context:
  • It can typically equal one minus the Type II Error probability (1-β).
  • It can typically increase with larger sample sizes and larger effect sizes.
  • It can often be calculated using power analysis before conducting experiments.
  • It can often guide sample size determination for achieving desired test sensitivity.
  • It can range from being a Low Statistical Power Measure to being a High Statistical Power Measure, depending on its probability value.
  • It can range from being a Fixed Statistical Power Measure to being a Adaptive Statistical Power Measure, depending on its calculation timing.
  • It can range from being a Univariate Statistical Power Measure to being a Multivariate Statistical Power Measure, depending on its test dimensionality.
  • It can range from being a Exact Statistical Power Measure to being a Approximate Statistical Power Measure, depending on its calculation method.
  • ...
• Example(s):
  • Clinical Trial Power Measures, such as:
    • An 80% Power standard for detecting clinically meaningful effects.
    • A 90% Power requirement for regulatory approval studies.
  • A/B Test Power Measures, such as:
    • A Minimum Detectable Effect calculation for website experiments.
    • A Sequential Test Power for early stopping decisions.
  • Research Study Power Measures, such as:
    • A Post-Hoc Power Analysis for published studies.
    • A Prospective Power Calculation for grant proposals.
  • ...
• Counter-Example(s):
  • A Significance Level, which controls Type I Error rather than Type II Error.
  • A P-Value, which measures evidence against the null rather than test sensitivity.
  • An Effect Size, which measures magnitude rather than detection probability.
• See: Type II Statistical Hypothesis Testing Error, Statistical Hypothesis Testing Task, Sample Size Determination, Effect Size, Power Analysis, Statistical Significance Level, Neyman-Pearson Framework, Optimal Test, Test Sensitivity.

---