CVCL {PowerTOST}

Confidence limits of a CV for log-normal data

Description

The function calculates the 1–α confidence limits (either 1-sided or 2-sided) via the χ² distribution of the error variance the CV is based on.

Usage

CVCL(CV, df, side = c("upper", "lower", "2-sided"), alpha = 0.05)

Arguments

CV

Coefficient of variation as ratio (not percent)

df

degrees of freedom of the CV (error variance)

side

Side(s) to calculate the confidence limits for, defaults to upper

alpha

Type I error probability, aka significance level

Value

Numeric vector of the confidence limits named as lower CL and upper CL.
In case of the one-sided upper confidence limit the lower CL is = 0.
In case of the one-sided lower confidence limit the upper CL is = Inf.

Author(s)

D. Labes

Examples

# upper one-sided 95% CL of a CV=0.3 
# from a study with df=22 (f.i. a 2x2 crossover with n=24)
# default side="upper" since not explicitly given
CVCL(0.3, df = 22)
# should give:
#  lower CL  upper CL 
# 0.0000000 0.4075525