In 1970 a most remarkable study was done. 35 Australian Aboriginal kids (57% female) raised by white families in Adelaide, South Australia since they were 1.5-years-old (on average), were intelligence tested. The kids ranged in age from five to 14 and ranged in ancestry from 100% Australoid to only 13% Australoid (about 58% on average). Unlike the Minnesota transracial adoption study where the adoptive parents were college educated, the “socio-economic status of the adopting families exhibited no obvious characteristics: they appeared to be representative of a wide cross-section of the South Australian population.”

This is good because it lets us see how a different race would score if reared in the typical white environment.

One of the tests given was the highly g loaded Peabody Picture Vocabulary test where the examiner says a word, and the subject points to one of several pictures that best represents that word. Below we see a chart comparing the mixed race adopted Adelaide Australoids to both white norms (standard scores from the test manual) and presumably unmixed and non-adopted Australoids tested at Hermannsburg Mission.

Dasen, P. R., de Lacey, P. R., & Seagrim, G. N. (1973). Reasoning ability in adopted and fostered Aboriginal children. In G. E. Kearney, P. R. de Lacey, & G. R. Davidson (Eds.), The psychology of Aboriginal Australians, (pp. 97-104). Sydney: Wiley.

The chart above suggests that unmixed, unadopted Australoids at age nine, 10, and 11 would have obtained raw scores on the PPVT (form A) of 37, 42, and 42. Using table 2 below I equate these to IQs of 44, 51, and 45, respectively (U.S. white norms). If these IQs sound absurdly low, keep in mind this is a vocabulary test and many of the unmixed unadopted Australoids perhaps did not speak English as their primary language.

Meanwhile at ages five, nine, 10 and 11, the mixed adopted Australoids have raw scores of about 44, 63, 65 and 95, corresponding to IQs of 88, 86, 85 and 118.

Osicka, C. J. (1976). Peabody Picture Vocabulary Test: Comments on administration and two methods of scoring. Psychological Reports, 38(3, Pt 2), 1135–1144. Using the first 4 rows I was able to roughly equate raw score mean and SD to IQ equivalent mean and SD, thus estimate IQ equivalent for any raw score in the given age groups.

On average the mixed adopted Australoids probably averaged verbal IQs of 94 (6 points below the white mean of the U.S., the U.K. and Australia) and considering they’re only 58% Australoid on average, unmixed adopted Australoids would likely have scored 90 (6 points below white mean/0.58 = 10 points below the white mean).

Conclusion

When raised in their own communities, unmixed Australoids seem to average about IQ 47 on English Picture Vocabulary (3.53 standard deviations below the white mean). But when raised in white homes they likely average IQ 90 (0.66 SD below the white mean). Thus this racial gap appears at most, only 19% genetic.

Challenges

Were the norms from the original PPVT valid for English speaking communities in 1970 Australia?

Given that the original PPVT was normed on several thousand white kids in and around Nashville, Tennessee in 1959, one might question its validity in 1970 Australia. A 1972 paper by L. J. Taylor , P. R. De Lacey & B. Nurcombe specifically set out to answer this question. They tested 60 Australian kids (mostly using form B) from age 4 to 10. Aside from age 4 where they included a lot of mixed Australoids in the sample, the average IQ at every age ranged from 91 to 112 or about 104 on average. So if anything the 1959 white norms from Nashville were perhaps 4 points too generous for Australians in the early 1970s, though even this slight inflation might be explained by the fact that the study translated three words on the test from American English to Australian English.

Did the IQs of the mixed race adopted kids decline in adulthood?

There was no follow-up study so we don’t how the adopted kids would have scored in adulthood, but in the Minnesota transracial adoption study, the IQs of adopted kids dropped between five to 13 points from age seven to 17. This drop was likely exaggerated by old norms in an at least one of the test sessions (and poor attempts to correct for the Flynn Effect), but to the extent that it was real, it was likely caused by the adoptive parents being upper-middle class. By contrast, in this study, the adoptive parents were more representative of the white population so the adoptees enjoyed the same environment as the average white kid, not benefitting from an upper class IQ boost that quickly fades.

[Update December 12, 2021: Thanks to a correction from an alert reader, some numbers in this article were substantially revised]