- Their culture is a dumpster fire, the collective product of bad, personal decisions made, in a great part, as a result of cultural pressures. We'll call this the Traditional hypothesis.
- They are victims of oppression. An oppressor class exists and takes all kinds of steps to make sure the oppressed fail. We'll call this the SJW hypothesis.
You can test these two hypotheses relatively simply. Take your data, any data you like, and divide it according to group populations.
In one group are the homogeneous communities where there is almost no mixing the oppressors and the oppressed. Discard any sets where the dominant group does not also control the levers of power. That is, you don't want a set where the oppressed are the vast majority, but the oppressors still hold most of the political offices.
In the other group are the heterogeneous communities where the oppressed and oppressors live together. It's best if the oppressors outnumber the oppressed and they control the levers of power.
Take a look at the data scoreboard for these two sets.
Under the Traditional hypothesis, you would expect performance to be roughly the same for both sets or superior for the heterogeneous communities. That would indicate that the underlying group cultures can be modified by the collective culture of the mixed sets or that the underlying group culture dominates. If either condition holds, it shows that oppression is not the dominant factor.
Under the SJW hypothesis, you would expect performance of the oppressed groups to excel in those sets that are numerically dominated and run by the oppressed group. This would indicate that, in the absence of oppressors oppressing, the oppressed would flourish.
You can do this with historical data, too. Going back in time, America was worse for the oppressed groups. I don't think anyone disputes that. If the SJW hypothesis holds, then the performance differences between the homogeneous and heterogeneous groups would grow as you went back in time.