Gian Carlo Diluvi
In the context of hypothesis testing, statistical power is defined as the probability of rejecting the null hypothesis when it is indeed not true. Multiple factors come into play when calculating power: sensitivity, specificity, the size of the effect that is to be detected, and sample size.
This dashboard helps to visualize how these factors affect the power in a Normal hypothesis test, and includes an interactive sample size calculator.
The dashboard is available here. I also developed some slides talking about the relationship about power and sample size.
The folder dashboard
includes the R
code used for generating the dashboard.