New notebook:
Sample Size Calculations in #RStats : Money and Power
https://arelbundock.com/posts/money_and_power/index.html
Sample Size Calculations in #RStats : Money and Power
https://arelbundock.com/posts/money_and_power/index.html
Comments
I think the way you do cost makes sense and like that it is stochastic here (though with complete randomization you do as a function of the design rather than realisation)
If you now placed an objective function over cost and power you could do a contour plot
We go a bit in this direction here https://book.declaredesign.org/lifecycle/planning.html#fig-ch21num1
@aecoppock.bsky.social @graemeblair.bsky.social I do remember we had a contour plot at some point showing indifference curves
Key idea though is that costs is a “diagnosis” just like power or MSE