Equivalent of power for presize functions

[aghaynes]

Power is currently ca 50% - half the trials will find a larger CI

Could abuse the closed formulae to estimate the power...

> mu <- 5
> sd <- 2
> n <- 20
> presize::prec_mean(mu, sd, n)

     precision for mean 

  mean sd  n conf.width conf.level      lwr      upr
  1    5  2 20   1.872058       0.95 4.063971 5.936029

> power.t.test(delta = 0.94, sd = 2, power = .5, sig.level = .05, type = "one.sample")

     One-sample t test power calculation 

              n = 19.3697
          delta = 0.94
             sd = 2
      sig.level = 0.05
          power = 0.5
    alternative = two.sided

> power.t.test(delta = 0.94, sd = 2, power = .8, sig.level = .05, type = "one.sample")

     One-sample t test power calculation 

              n = 37.49745
          delta = 0.94
             sd = 2
      sig.level = 0.05
          power = 0.8
    alternative = two.sided

[aghaynes]

Paper: Probabilistic precision calculations for the planning of studies assessing negative binomial rates
https://onlinelibrary.wiley.com/doi/10.1002/pds.5750
R code as supplement to the paper

Potential new rate method too?

[lindeloev]

+1 for this feature. The probability of obtaining a CI of at most the desired width is often called "assurance". It's built into many SAS PROC POWER procedures where it's called "probwidth". I like "assurance" better, though "power" may be OK too.

[aghaynes]

Thanks @lindeloev! I'm not a SAS user, so this is news to me. Assurance does seem to be a nice term for it. Power just brings the whole baggage from testing with it, which is not ideal.