Thank you, Val! That was what I was recalling, as well-- but what I meant by "small numbers" was actually something that I left out. Immigrant children who were adopted. That's a much smaller group. Presumably the same facts would hold true, but there is little way to evaluate the effect of SES versus educational attainment in adoptive households among particular immigrant groups, nevermind among groups with a known IQ range. To be fair, though-- the sibling meta-study above suggests that the effect is what I'd assume-- that environment matters a very great deal.

I think that is an important consideration if Bostonian (and others') premise is to be taken seriously and evaluated purely upon its own merits. Can you take a birth cohort which is normed for IQ, and determine outcomes on the basis of two different measures in adoptive homes? One for SES, and one for EdAtt? I would bet that the LATTER is the more robust effect, but I don't know.
The other thing that adoption studies cannot tease apart is the impact of "wanted" versus "unplanned" childrearing. It's a problem.


My point is that most immigrants of "successful" groups tend not to come from the lowest level working class, in terms of educational attainment.

Well, they do from Latin America. But not from overseas. The groups that do tend not to produce "highly successful" high SES children in large numbers.

I'm pretty sure that such a thing does NOT go to supporting the notion that lack of success is related to ethnic differences in intellect, however. Because WHO those immigrants are (culturally, educationally, in terms of SES) matters a great deal. It's a vastly different slice of the demographic when you look at war refugees versus those immigrating for purely SES reasons.

Schrödinger's cat walks into a bar. And doesn't.