Dear allstat,
I hope you can help, and all comments greatly appreciated.
In their book streiner and norman (Health Measurement Scales: chapter 8)
describe how to use Fishers ZR transform to estimate the sample size for ICC
based on the desired width of the confidence interval for the ICC.
Could I ask if anyone has any references to this approach as the book just
references Fisher 1925. I’m struggling to understand this as there is no
account of power in the fisher approach. If we base the sample size on the
width of the 95% CI for an assumed ICC then once we get the 95% CI we can
use this to do a significance test against an alternative. So I’m
struggling to reconcile this appraoch to the power based calculations of
Donner and Eliaswiz.
Many thanks
Malcolm Morrissey
_________________________________________________________________
The next generation of Hotmail is here! http://www.newhotmail.co.uk
|