plotCorrPrecision {Hmisc} | R Documentation |
This function plots the precision (margin of error) of the
product-moment linear
correlation coefficient r vs. sample size, for a given vector of
correlation coefficients rho
. Precision is defined as the larger
of the upper confidence limit minus rho and rho minus the lower confidence
limit. labcurve
is used to automatically label the curves.
plotCorrPrecision(rho = c(0, 0.5), n = seq(10, 400, length = 100), conf.int = 0.95)
rho |
single or vector of true correlations. A worst-case precision graph results from rho=0 |
n |
vector of sample sizes to use on the x-axis |
conf.int |
confidence coefficient; default uses 0.95 confidence limits |
Xing Wang and Frank Harrell
plotCorrPrecision() plotCorrPrecision(rho=0)