I have read that a multiple regression works best when the IVs strongly correlate with the DV and do not correlate much with each other.
My multiple regression model is designed to test a theory which proposes that my IVs are predictive of a DV value. However, the IVs have a small correlation with the DV (which is not significant). Also some of the IVs are moderately correlated (all r <.20). Is it still ok to run the regression model as it is testing a theory?
I have been told that I need to justify the use of a regression in this case. I was wondering if anyone knew of any papers that suggest using a regression model in this way is acceptable.
Thanks very much for any help you can give me. I have been scouring books and journals and the internet, but can't seem to find an answer to my problem.