Articles
Get started
Install powerape and run a first power analysis, in R or in Stata.
- Powering the minimum-effect claim
- Get started with powerape in Stata
Install the command, run a first power analysis for an average partial effect, and read its output: the six steps of the R article, in the same world.
Worked example
Five study designs taken through both languages, from a treatment effect worth acting on to an instrumented treatment.
- Worked example: five designs, from a treatment effect to an instrument
Five study designs taken through powerape in R and in Stata: an effect worth acting on, an interaction, a design built from a pilot, a firm-year panel and an instrumented treatment.
Workflows
One design question per article, with the R code and the Stata code side by side.
- The classic SESOI power analysis, and what it cannot tell you
- Powering for equivalence: showing an effect is too small to matter
- Power for average interaction effects (AIEs)
- Panel designs: correlated random effects
- IV designs: endogenous focal variables via control functions
- Justifying the error rate: alpha by sample size, or by error costs
Guides
How the two implementations correspond, and how both are checked.
- R and Stata side by side
The functions and arguments of the R package and the subcommands and options of the Stata command, matched, with the places where the two differ.
- How powerape is validated
The independent implementations and exact results the R package is checked against, and the four tiers in which the Stata command is checked against the R package.
Stata reference
- Stata help file
The help file of the Stata command, as displayed by
help powerape.- News: the Stata command
Versions of the Stata command powerape. The R package’s changes are under News, R package.