Statistical Power
The probability that a study will detect a true effect if one exists (i.e., correctly rejecting a false null hypothesis).
What is Statistical Power?
Statistical power is the likelihood that a hypothesis test will avoid a Type II error (false negative). It is influenced by the sample size, the effect size, and the chosen significance level (alpha).
Why Statistical Power Matters
An underpowered study is fundamentally flawed because it is unlikely to find the very effect it is looking for. High power ensures that the study is capable of detecting meaningful differences.
How to Interpret
Researchers typically aim for a power of 0.80 (80%) or higher when designing a study. This is determined a priori using a power analysis.
Example
A researcher wants to detect a small reduction in blood pressure. If their sample size is only 10 people, the statistical power might be 20%, meaning they only have a 1 in 5 chance of detecting the true effect.
Common Mistakes
- Conducting a study without running an a priori power analysis to determine the required sample size.
- Running 'post-hoc' power analysis using the observed effect size, which is mathematically circular and uninformative.