Statistical power

Definition: Statistical power is the long-run probability that a statistical test correctly rejects the null hypothesis if the alternative hypothesis is true. It ranges from 0 to 1, but is often expressed as a percentage. Power can be estimated using the significance criterion (alpha), effect size, and sample size used for a specific analysis technique. There are two main applications of statistical power. A priori power where the researcher asks the question “given an effect size, how many participants would I need for X% power?”. Sensitivity power asks the question “given a known sample size, what effect size could I detect with X% power?”.

Related terms: Effect Size, Meta-analysis, Null Hypothesis Significance Testing (NHST), Positive Predictive Value, Power Analysis, Quantitative research, Sample size, Significance criterion (alpha), Type I error, Type II error

Related term to alternative definition: Type II Error

References: Carter et al. (2021), Cohen (1962), Cohen (1988), Dienes (2008), Giner-Sorolla et al. (2019), Ioannidis (2005), & Lakens (2021a)

Drafted and Reviewed by: Thomas Rhys Evans, James E. Bartlett, Jamie P. Cockcroft, Adrien Fillon, Emma Henderson, Tamara Kalandadze, William Ngiam, Catia M. Oliveira, Charlotte R. Pennington, Graham Reid, Martin Vasilev, Qinyu Xiao, Flávio Azevedo

Note that we are currently working on an automated mechanism to link references cited above with their full-length version that can be found at https://forrt.org/glossary/references with all references used so far.