Significance versus hypothesis testing

I really like the precision basis for N. I think we also need some new vocabulary around these: * maximum budgeted sample size * expected sample size under a certain prior distribution of treatment effects * minimum sample size if effectiveness is declared (usually precision-guided)