FAQ/fishrpow - CBU statistics Wiki

Please enter your password of your account at the remote wiki below.
/!\ You should trust both wikis because the password could be read by the particular administrators.

Clear message
location: FAQ / fishrpow

Comparing correlations form two separate samples using fisher's transformation in R

Copy and paste the below into R.

fn <- function(r1,r2,n1,n2,alpha) {

a <- abs((1+r1)/(1-r1))
b <- abs((1+r2)/(1-r2))
a <- 0.5*log(a)
b <- 0.5*log(b)
c <- 1/(n1-3)
d <- 1/(n2-3)
chisq <- abs((a-b)/sqrt(c+d))

pow <- 1 - pchisq(abs(qchisq(1-alpha,1)),1,chisq*chisq)
cat("Power for Fisher's test is")
print(pow)
 }

Input the (two-tailed) type one error, alpha, correlations, r1 and r2, and sample sizes, n1 and n2 as below into R.

alpha <- 0.05
r1 <- 0.7
r2 <- 0.8
n1 <- 67
n2 <- 10000

then copy and paste, or type in, the below to obtain the power.

fn(r1,r2,n1,n2,alpha)