16.1 Statistical Power

• Statistical power is the probability that a particular comparison will achieve “statistical significance” at some predetermined level before a study is performed.

• A power analysis is performed by hypothesizing an effect size, making assumptions about the variation in the data and the sample size of the study to be conducted, and using probability calculations to determine the chance of the p-value being below the threshold.

• Low-power but inexpensive studies seem tempting, but if they succeed they are likely to have Type-M or even Type-S errors. Winners Curse

• To estimate effect size, consider a range of values, or decide what magnitude of effect would be practically interesting.