4 ms·
I think there are two dimensions (likely correlated) 1. Race. 2. Socioeconomic status. Callibrating a score on race would be a problem, because you are expli
by uptownfunk 7y ago
I think there are two dimensions (likely correlated)
1. Race.
2. Socioeconomic status.
Callibrating a score on race would be a problem, because you are explicitly reinforcing virtues that lead to suboptimal performance. However calibrating a score on (2) makes sense because kids with unequal resources should have access to the same opportunities, and it is reasonable to me that a top performer in a lower socioeconomic neighborhood would do better than someone who performed similarly were they raised in a more affluent neighborhood.
Where you run into problems is that lower socioeconomic neighborhoods are associated strongly with black/Latino populations and so it is indirectly biasing for select racial groups. I don’t think there’s a perfect solution, or one that is 100% equitable to all parties but I think we can do better than we are now, in fact I think it is crucial.