Resources

Power analysis through simulations in Stata: a guide for dummies

You want to run an experiment for which you expect to find a treatment effect. How big should the sample size be for you to have a reasonable chance of finding significant results? In this manuscript, I share and explain the Stata code that I use to answer this question. The code uses simulations to perform power analyses. In contrast to the already widely used analytical tools to perform power analyses, this method can accommodate any experimental design and statistical test that Stata can perform. The code is very simple, easy to use and will be useful for researchers without much coding experience.

To access the document, please follow this link.