How to design an ecologic study while minimizing potential for ecologic fallacy?

[MisfitsFiend]

any suggestions? Once two factors are determined, can go another route and do a case-control, but is there any way of eliminating this step, and just creating a good ecologic study minimizing potential for ecologic fallacy?

-mf


EDIT: can't spell

 

[z0mb13]

my brain just exploded