> This US-government provided table of average SAT scores in the United States in 2023
If you look at their source[0], there's no information about how they controlled for confounders (because it's impossible as they acknowledge[1].)
There's a strong correlation between "education of parents" and "SAT score"[2] which implies that family wealth is a strong contribution to a child's SAT score (something we all know anyway); that's also backed up by [3].
(I'd suggest that [4] also contributes to the positive correlation between familial wealth and test scores but perhaps in a more oblique "the higher goals are aimed at by kids who have the backing to contemplate them because of family support structure, tutoring, ability to pay for the degree(s), etc." way.)
(Similarly for [5], I suppose - there's a distinct correlation between what I'd say was "perceived difficulty of major" and the mean SAT scores. Again probably down to familial wealth, support, tutoring, etc.)
Someone who's an actual statistician would probably rip this apart much more thoroughly and with more rigour than I, of course.
[0] https://reports.collegeboard.org/media/pdf/2023-total-group-...
[1] "Relationships between test scores and other background or Evidence-Based
Reading and Writing (ERW) and Math contextual factors are complex and
interdependent. Caution is sections of each assessment in the SAT Suite:
warranted when using scores to compare or evaluate teachers, schools,
districts, or states, because of differences in participation and test
taker populations."
[2] Bottom of page 4: "Highest Level of Parental Education"
[3] Bottom of page 5: "Median Family Income"
[4] Bottom of page 8: "Degree-Level Goal"
[5] Top of page 8: "Intended College Major"