testing differences between correlated effect sizes

#1
Please help!

I would like to test for significant differences in effect sizes between three measures of the same construct (relationship satisfaction) generated in the same dataset. These three measures are highly correlated (r's of about .90) but demonstrate markedly different effects due to differing levels of noise in repeated measurement for each scale.

I have been able to find equations to test for differences between effect sizes generated in different samples (presumably using the same measure), but I really want to test for differences in the same sample using correlated measures.

Does anyone know the appropriate equation(s) to do this??? Your help would be greatly appreciated!!! Thanks!

Ron :confused:


Specifics of the study:
I have tested how responsive three different measures of relationship satisfaction are to real-world changes in relationship happiness in a series of three different samples.

I have calculated effect sizes for each scale for detecting small amounts of improvement or deterioration in contrast to a no-change group. (These were all significant contrasts). I then calculated weighted averages across the three studies and would really like to test for differences among those summary effect sizes.
 
#2
I think I figured out how to do this

My solution: I converted my effect sizes (Cohen's d's) into correlation coefficients. That way, I could use the test for differences in dependent correlation coefficients suggested by Cohen & Cohen!