Another question about regression

[Diane]

First, thank you to John and Quark for responding to my question about multiple regression!!!!

I have another question. I am using SPSS and working with dummy and effect coding. In order to determine the total variance in a variable determined by the different groups, would I square the part correlation of each group in order to find out how much each of the groups contribute.

Thanks a bunch!!!

Diane

 

[JohnM]

Yes, that's correct. If you go to the following link, it will give a detailed explanation of part correlation:

http://www2.chass.ncsu.edu/garson/pa765/partialr.htm#part

Scroll down to the section heading:
What is part correlation (semi-partial correlation)? Compare to partial correlation.