But there are studies correlating low socioeconomic status with low IQ. And there are studies correlating race with socioeconomic status. So connect the two and you have a correlation between race and low IQ. But to make assertions about the IQ of a race is backwards. Because if your goal is to solve IQ, then you would solve socioeconomic status not race, as race is just some other third variable that we don't care about. Causality could never be proven between race and IQ because socioeconomic status (and education opportunities and cultural biases in IQ tests and historical disadvantages for different races... woops that list is exhaustive) is a confounding factor. Such a study would be unscientific.