Documentation: I don't know which group is meant to be "greater"

[hughjonesd]

In pwr.t2n.test, does alternative = "greater" mean that the n1 group is greater or the n2 group is greater? Similar questions wherever this argument is used.

(This mystery is never explained in the base R documentation either. Perhaps it's an in-joke among statisticians.)

[heliosdrm]

Thanks for the comment. The functions of the package are based on Cohen's book "Statistical power analysis for the behavioral science", which is referred in the documentation, and all details are given there. In this particular case, "greater" means the alternative hypothesis of "effect size d > 0".

[hughjonesd]

Maybe just adding that to the docs will help people use it who don't have the book.