Total CV for parallel study from CI [Power / Sample Size]
Dear MGR,
it is only possible to get the total (inter+intra) CV from a parallel group study.
With this in mind and assuming 90% CI you can do in the software R
:
gives you a CV = 0.258 (25.8%).
If you don't have R you can do the algebra behind by hand. It's not so hard. See Helmut's lectures about "sample size estimation". Change degrees of freedom and design constant according to the parallel group design.
Better install R, it's free. And use the "gold standard" package
.
it is only possible to get the total (inter+intra) CV from a parallel group study.
With this in mind and assuming 90% CI you can do in the software R

library(PowerTOST)
CVfromCI(lower=0.876, upper=1.02, n=122, design="parallel")
gives you a CV = 0.258 (25.8%).
If you don't have R you can do the algebra behind by hand. It's not so hard. See Helmut's lectures about "sample size estimation". Change degrees of freedom and design constant according to the parallel group design.
Better install R, it's free. And use the "gold standard" package
PowerTOST

—
Regards,
Detlew
Regards,
Detlew
Complete thread:
- Inter CV calculation for parallel study MGR 2015-11-13 07:43 [Power / Sample Size]
- Total CV for parallel study from CId_labes 2015-11-13 11:13