Could someone have some insights?
Suppose that based on one measurement X, a sample size of N will be
enough to give me sufficient precision of the population mean with a
certain power. Now, suppose we can only measure Y, but Y is highly
correlated with X, say with .9 correlation, is there any easy way to
compute the needed sample size based on Y to give similar precision of
the mean at the same power?
Thanks!
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================
