When countries are compared on intellectual measures, significant differences are found. Leaving aside why these differences exist, this has considerable implications for the economies of those nations, and also for the way in which their institutions function. Although those intellectual differences have been established, the pressing question is whether they are amenable to change. If the gaps can be closed by whatever means then the effects will be diminished and eventually annulled.
It is with that background in mind that it is particularly interesting to have a preview of the publication in the special issue of Intelligence on the Flynn Effect.
“Cognitive ability differences between countries can be large, with average IQs ranging from approximately 70 in sub-Saharan Africa to 105 in the countries of north-east Asia. A likely reason for the great magnitude of these differences is the Flynn effect, which massively raised average IQs in economically advanced countries during the 20th century. The present study tests the prediction that international IQ differences are diminishing again because substantial Flynn effects are now under way in the less developed “low-IQ countries” while intelligence is stagnating in the economically advanced “high-IQ countries.” The hypothesis is examined with two periodically administered scholastic assessment programs. TIMSS has tested 8th-grade students periodically between 1995 and 2011 in mathematics and science, and PISA has administered tests of mathematics, science and reading between 2000 and 2009. In both TIMSS and PISA, low-scoring countries tend to show a rising trend relative to higher-scoring countries. Despite the short time series of only 9 and 16 years, the results indicate that differences between high-scoring and low-scoring countries are diminishing on these scholastic achievement tests. The results support the prediction that through a combination of substantial Flynn effects in low-scoring countries and diminished (or even negative) Flynn effects in high-scoring countries, cognitive differences between countries are getting smaller on a worldwide scale.”
In the discussion they add:
“The magnitude of test score convergence is less certain. In PISA, continuation of the current trends is calculated to erase the differences between high-scoring and low-scoring countries in only 40 years, with a 95% confidence interval of 27 to 77 years. In TIMSS, complete convergence would result after 341 years, with a 95% confidence interval of 70 years to never. These calculations are based on the prediction of the trend measured by the averaged performance on the first and last assessments. We do not know whether test score convergence will ever be complete. It might not if, as is frequently assumed (e.g., Jensen, 1998), biological limits for the development of high intelligence are different in different countries. One possible outcome is partial convergence leading to smaller but persistent gaps, similar to test score convergence between racial groups in the United States during the last three decades of the 20th century (National Center for Education Statistics, 2009).”
This is an interesting observation. PISA and TIMSS are similar, though the latter has a greater emphasis on maths and science, so could conceivably be seen as the harder of the two.
As I had already discussed in the previous post on the Rindermann and Thompson paper about narrowing ethnic gaps in the US, that what has happened is a partial convergence, which then seems to have stopped converging. One response to this finding is the “one more push” policy, which is to keep administering the special programs in the not unreasonable hope that they will eventually be effective “all the way”. This might happen. Another possibility is that, even if the compensatory education policies are redoubled in their scope and intensity, not much new convergence is achieved. One mildly absurd possibility is that such programs overcome the environmental half of the gap, but cannot touch the genetic component. Another 15 years of data might help us evaluate that hypothesis.