CorDiffCI

CorDiffCI calculates a confidence interval for the difference of two independent correlation coefficients.

CorDiffCI(R1, R2, N1, N2 = N1, CONF.LEVEL = 0.95, DIGITS = 3)

Arguments

R1

estimate of first correlation coefficient
R2

estimate of second correlation coefficient
N1

number of valid values for R1
N2

=N1 number of valid values for R2
CONF.LEVEL

=0.95 desired confidence level
DIGITS

=3 Number of digits for correlation difference

Value

data.frame of R1, OR, CorDiff, Lower and Upper

Examples

if (FALSE) {
CorDiffCI(.49,.36,145,87,CONF.LEVEL=.95)  ## Example from Zou 2007 reproduced
x<-c(rnorm(50),rnorm(50,5))
y<-c(rnorm(50),rnorm(50,5)+10*x[51:100])*2
Split<-factor(paste("Group",rep(LETTERS[1:2],each=50),sep="-"))
ScatterPlot(x,y,SPLIT=Split)
CorDiffCI(cor(x[Split=="Group-A"],y[Split=="Group-A"]),
           cor(x[Split=="Group-B"],y[Split=="Group-B"]),
           50,50)
}