Summary
The probability of detecting a true effect, determined by sample size, effect size and variability.
In depth
Power = 1 - ฮฒ, usually aimed at 0.8. Larger samples, bigger effect sizes and lower variability raise power. Underpowered studies miss true effects and exaggerate effect sizes. Power analysis is done before experiments.
Increase power by reducing noise, using paired designs or larger samples. Report effect sizes. Underpowered studies waste resources and animals.
Common mistake: 'no significant difference' in a small study is not evidence of no effect.
Statistical power is the probability that a test will detect an effect that really exists (1 minus the type II error rate); conventionally 80 percent is a target. It depends on sample size, the true effect size, variability (SD), and the significance level.
Low-powered studies are likely to miss real effects, and when they do find an effect it tends to be exaggerated, and results are less likely to replicate. Researchers should do a power analysis before an experiment to calculate how large a sample is needed. Larger samples, less variability, and bigger effects all increase power.
Example: to detect a small difference in blood pressure between two drugs needs hundreds of patients per group.
Common mistake: concluding no effect from a non-significant result in a small study.
Frequently asked questions
What is Statistical Power?
The probability of detecting a true effect, determined by sample size, effect size and variability.
Which area of biology covers Statistical Power, and how advanced is it?
Statistical Power is one of the 51 topics in Lab Methods & Biostatistics. It is rated Advanced, so it suits learners who want specialist depth. Related topics in this area include Null Hypothesis, t-Test, Experimental Design.
Can you give an example of Statistical Power?
A clinical trial must recruit enough participants.
What should I learn before Statistical Power?
Work through the Beginner topics in Lab Methods & Biostatistics first, such as Scientific Method, Hypothesis, Variables, then return to Statistical Power. Advanced topics assume the Intermediate vocabulary.
What topics are related to Statistical Power?
The closest topics to Statistical Power are Null Hypothesis, t-Test, Experimental Design, Meta-Analysis. Comparing them side by side is a good way to see how the ideas connect.
How can I check that I understand Statistical Power?
Read the summary and the in-depth section, then use the quick check on the interactive lesson for Statistical Power. To practise the whole of Lab Methods & Biostatistics, use the Lab Methods & Biostatistics quiz.
How is Statistical Power different from Null Hypothesis?
Statistical Power: The probability of detecting a true effect, determined by sample size, effect size and variability. Null Hypothesis: The default assumption of no effect or difference, which data may lead us to reject. Read both side by side and note what each one explains.
What is a common mistake about Statistical Power?
'no significant difference' in a small study is not evidence of no effect.
How should I study Statistical Power step by step?
First, read the summary and write the definition in your own words. Second, read the in-depth section and note any example. Third, compare it with Null Hypothesis and explain the difference aloud. Finally, take the Lab Methods & Biostatistics quiz and come back to this page for any question you miss.