The relationship between power, alpha, beta, P-Values and clinical importance

When being introduced to frequentist power calculations invariably you will be taught using a figure showing two normal distributions. If you are not familiar with the properties of distributions the concepts may seem quite strange, here I try to explain what’s going on using RShiny. The sliders below are used to select the true population parameters for a normally distributed continuous response and the probabilities for making two types of error. Though the concepts apply equally to proportions as well.