Power Test

A Power Test is a statistical calculation performed before a study to determine the minimum sample size needed for the study to have enough power. In other words, the minimum numbers of participants you need to have in your study. To make this more understandable, let's discuss "Power".

Power is the probability that a statistically significant effect can be found when it actually exists. Without adequate power you might commit a Type II error, meaning that you fail to reject the null hypothesis when it is false. The general consensus is that power should be 0.8 or greater; if it is less than 0.8 then the same size is too small. The exact formula for a power test depends on what type of analysis you are running (such as a t-test), but power formulas take into account the desired alpha or significance level, the effect size or expected difference you wish to detect, and known variation in the population.

Add flashcard Cite Random