A practical primer to power analysis for simple experimental designs

Abstract

Power analysis is an important tool to use when planning studies. This contribution aims to remind read-ers what power analysis is, emphasize why it matters, and articulate when and how it should be used. The focus is on applications of power analysis for experimental designs often encountered in psychology, starting from simple two-group independent and paired groups and moving to one-way analysis of variance, factorial designs, contrast analysis, trend analysis, regression analysis, analysis of covariance, and mediation analysis. Special attention is given to the application of power analysis to moderation designs, considering both dichotomous and continuous predictors and moderators. Illustrative practical examples based on G∗Power and R packages are provided throughout the article. Annotated code for the examples with R and dedicated computational tools are made freely available at a dedicated web page (https://github.com/mcfanda/primerPowerIRSP). Applications of power analysis for more complex designs are briefly mentioned, and some important general issues related to power analysis are discussed.