[R] Finding 'ncp' for t

[Peter Dalgaard]

(Ted Harding) <Ted.Harding at nessie.mcc.ac.uk> writes:

> Hi Folks,
> 
> I'm looking for a neat procedure for the following:
> 
> Given t0 such that
> 
>   pt(t0,df,ncp=0) = alpha (given)
> 
> find ncp0 such that for given beta
> 
>   pt(t0,df,ncp=ncp0) = (1 - beta)
> 
> (In other words, what's the ncp such that you get power (1-beta)
> to detect it, using a 1-sided test with size alpha when ncp = 0?)
> 
> In the past I've done the groping by hand, but this time it
> needs to be done many times over, so a good solver for ncp0
> in this situation would be handy.

You might want to peek inside power.t.test(). Or just use it...

-- 
   O__  ---- Peter Dalgaard             Blegdamsvej 3  
  c/ /'_ --- Dept. of Biostatistics     2200 Cph. N   
 (*) \(*) -- University of Copenhagen   Denmark      Ph: (+45) 35327918
~~~~~~~~~~ - (p.dalgaard at biostat.ku.dk)             FAX: (+45) 35327907