Hi, Marie. I rely on Bob Rosenthal for answering that question. It's essentially the same thing as comparing effect sizes reported as correlations. Transform your raw rs to Fisher z(sub)r. Z of difference is zr1 minus zr2 divided by the square root of 1 over N1-3 plus 1 over N2-3). I use Chapter 21 of Rosenthal & Rosnow's Essentials of Behavioral Research as a quick reference. It also contains the table for finding your Zr (though Excel will do the same).
HTH, Cheryl R. --- Cheryl A. Rickabaugh University of Redlands -----Original Message----- From: Helweg-Larsen, Marie [mailto:[EMAIL PROTECTED] Sent: Sun 11/4/2007 5:02 PM To: Teaching in the Psychological Sciences (TIPS) Subject: Comparing correlations Hi Tipsters I have a simple correlational question. What is the formula for comparing two significant correlation coefficients? Among one group r(44) = .37, p = .01 and among a different group r(115) = .19, p = .04. My question is whether the first corr is sig. greater than the second? Marie ---
<<winmail.dat>>
