Meng's Z-test for correlated correlation coefficients

This function implements Meng's z-test for correlated correlations.

This function implements Meng's z-test for correlated correlations (Meng,
Rubin, & Rosenthal (1992), Comparing Correlated Correlation Coefficients,
Psych Bulletin 111(1), 172-175.)

mengz(r1, r2, r12, n) compares two correlations r1 and r2:
r1: correlation between X and Y
r2: correlation between X and Z
rx: correlation between Y and Z
n: number of observations used to compute correlations

mengz(R, k, n) tests the heterogeneity of a correlation matrix with
respect to correlating with the variable indicated by index k. This test
is a chi-squared test, so output argument z is actually a chi-squared
statistic.

mengz(R, k, n, lambda) tests the contrast indicated by vector lambda.

h: hypothesis outcome (1 - reject null hypothesis of equal correlations
under alpha level of 0.05 (one-tailed))
p: chance of falsely rejecting null hypothesis
z: computed z- or chi-square score of Meng's test

Copyright (C) 2012 Eelke Spaak, Donders Institute for Brain, Cognition,
and Behaviour, Nijmegen, The Netherlands