Look at ravens progressive matrices, these are as far from relying on culture as you can get- they are too abstract and tend to show reasonable distributions of results in all groups. They also show poor results for some groups, including africans.
This comment makes my point on this better than the one above did: http://lesswrong.com/lw/kk/why_are_individual_iq_differences_ok/776g
Look at ravens progressive matrices, these are as far from relying on culture as you can get- they are too abstract and tend to show reasonable distributions of results in all groups. They also show poor results for some groups, including africans.
This comment makes my point on this better than the one above did: http://lesswrong.com/lw/kk/why_are_individual_iq_differences_ok/776g