5 continuous variables: A,B,C,D,E. (These are my "IV" or confounding variable depending on the scenario. Will explain below.)

1 categorical variable, F. (This will only be used as a confound.)

1 continuous variable, G. This is my "DV."

I am basically after three separate tests:

1-relationship between each of the first 5 continuous variables (A,B,C,D,E) and G. This is easy. Pearson Correlation. But I want it in regression b/c the below tests will probably require regression. I want to put the results of all these in one big table preferably, and also compare them with each other perhaps; that's why I would

*prefer*for the tests to match.

2-relationship between each of the first 5 continuous variables (A,B,C,D,E) and G

*whilst controlling for the categorical confounder, F.*Apparently I cannot do a partial correlation for this b/c the confounder is categorical. So, linear regression?

3-relationship between each of the first 5 continuous variables (A,B,C,D,E) and G

*whilst controlling for the categorical confounder F, and the rest of the continuous variables.*Specifically;

+the relationship between A and G whilst controlling for B,C,D,E, and F

+relationship between B and G whilst controlling for A,C,D,E, and F

+relationship between C and G whilst controlling for A,B,D,E, and F

+relationship between D and G whilst controlling for A,B,C,E, and F

+relationship between E and G whilst controlling for A,B,C,D, and F

I assume I do a multiple regression for these. I did. In the results screen of SPSS, under the coefficients table, there are the p values along Zero-order, Partial, and Part correlations. So Q: Can I use these correlation coefficients and the p values for each of the 5 correlations above in my 3rd (set of) test(s)? What if the p value of the model is above 0,05? Say one of them is significant but the model is not, what do I do?