[Update: I repeated my previous analysis using Pisa 2006 Science data (age 15) and using International Explorer. In that data — screen shot here — there were 72 self identified mixed race students. The following averages were found: (non-hispanic) White 522, (non-hispanic) Mixed White-Black 470, (non-hispanic) Black 408.]
[Update: I repeated my previous analysis using Pisa 2006 Math data (age 15) and using International Explorer. In that data — screen shot here — there were 67 self identified mixed race students. The following averages were found: (non-hispanic) White 502, (non-hispanic) Mixed White-Black 464, (non-hispanic) Black 403.]
[Update: I repeated my previous analysis using the Pisa 2003 US data. In that data set, there were 785 (non-hispanic) Whites, 52 (non-hispanic) White-Blacks, and 3190 (non-hispanic) Blacks. The difference between Whites and Blacks on math, verbal, and science tests ranged from (about) .8-1 SD. White-Blacks were only .15-.25 SD below Whites. Averaging the Pisa 2003 and 2009 results together, (nonhispanic) Black-Whites (N=132) score about .30 SD below Whites. As I noted below, I’m taking a lot of shortcuts in my analysis, so the results are not too robust. But they are about what one would expect, given the aggregated “mixed race” performance reported in the released reports.]
[Note to Steve Sailer: If you’re still interested in US immigrant IQ by generation, you might peruse the Pisa data.]
I looked at the effect of racial mixture on international achievement scores in the Program for International Student Assessment (PISA) 2009 US data. (You can download the Pisa (2009, 2006, 2003, 2000) public use data files here.) In the Pisa data, 5 plausible values are given for each of the three assessed areas (Math, Reading, and Science) per student. They represent probable latent value scores. When comparing groups, one is supposed to average the means of the 5 (x 3) values. To save myself time, I compared the means of the first two values for each area. Also, I did not add weights to the values (so the numbers are a little different from those listed in the Pisa manual), as I wanted the printout to show actual Ns; in this instance, doing so does not substantially change the results. Race/ethnicity was self-reported. As can be seen, there were 727 non-Hispanic Blacks, 80 non-Hispanic mixed Black-Whites, 97 non-Hispanic + Hispanic Black-Whites, 65 US born non-Hispanic Blacks with at least one parent born outside the US, and 3022 non-Hispanic Whites. In math, reading, and science, Biracial (Hispanic/non-Hispanic) Black-Whites fell in between Whites and Blacks. As did 2nd generation Blacks. Overall, these findings are in agreement with others discussed previously. In light of the accumulated evidence, claims of the intellectual equality between mixed and non mixed Black/Whites (e.g., Nisbett, 2011) are no longer tenable; this category of evidence can no longer be said to support a wholly environmental etiology to the Black-White achievement gap.