PISA tests are not IQ tests.

Cultural and socio-economic factors are just as good an explanation for the effects you mention than genetic ones tied to race.

I also very much doubt that the effects you mention are studied in the data on PISA test takers - there is simply not that level of demographic information being collected when a PISA test is administered. No one asks each child in the classroom where their parents or grand-parents were born, nor is their race or skin color noted down. So until you cite some papers, I'm going to be extremely skeptical of your claim.