Lynn's study uses regional differences in the performance of Italian secondary school children on Organisation for Economic Co-operation and Development tests of educational achievement to assess regional IQ differences. However, data on Italian regional differences in educational achievement obtained in a much larger INVALSI study of 2,089,829 Italian schoolchildren provide unequivocal evidence that Lynn's educational achievement measure is not a valid index of IQ differences. More generally, the lengthy literature review in Lynn's article reveals uncritical acceptance of reported correlations between any putative index of IQ and socio-economic variables. Any measure of cognitive performance that is correlated with IQ is considered a measure of IQ, even if there is only a weak correlation. All correlations between such measures and socio-economic or public health variables are viewed as evidence of direct causal relationships. In all cases, causality is assumed to be in the direction that supports Lynn's doctrine when it would be equally valid to argue that socio-economic and public health differences cause differences in the performance of IQ tests. In addition to these fundamental logical and statistical errors the present report records numerous other data processing, methodological, and conceptual errors.
Robinson et al. "The case against Lynn's doctrine that population IQ determines levels of socio-economic development and public health status". J Pub Ment Health, 2011.