Express skepticism and internet karma answers back

Just three days ago I expressed my skepticism about international IQ comparisons in a post, Complex Comparisons.

I believe IQ measurements are demonstrably and reliably useful predictors of specified outcomes at a population level. I become much more leery about inter-country comparisons, not because of a distrust in the utility of IQ but a distrust in the reliability and meaningfulness of measurement on a global scale where you are dealing with all the nuance of language and culture.

This morning I see out of Oxford, Cognitive Abilities in Young Lives: An Overview of Results from Ethiopia, India, Peru and Vietnam by Richard Lynn, Helen Cheng, and Gerhard Meisenberg. From the Abstract.

This paper compiles cognitive test results for children in Ethiopia, Andhra Pradesh (India), Peru and Vietnam from multiple rounds of the Young Lives study. In this international project, the same cognitive tests were administered to children of the same age under standardized conditions, allowing comparisons between countries and between social, ethnic, linguistic and religious groups within countries. Comparisons between countries on non-verbal tests show differences that closely resemble those that have been seen in earlier assessments of scholastic achievement and intelligence. Within each of the four countries there are significant differences between social, ethnic, linguistic and in some cases religious groups that are related to socioeconomic conditions. These results have implications for the management of inequalities that have either been present for a long time or that arise in developing countries during the process of modernization.

Much more structured, much more disciplined in their administration of tests. It looks like a sophisticated group being very conscientious about trying to answer a pressing set of development questions.

The key finding is that IQ tests administered in much more consistent and disciplined ways for Ethiopia, India, Peru, and Vietnam return similar IQ results and variances as the tests about which I have expressed skepticism.

I see many fewer obvious potential sources of error in this study than I do in most so I have to take this on board. But I remain cautious. Culture and personal behavior are big mediators in life outcomes and which are much harder to measure consistently over time and in a causal fashion. IQ is an important contributor to life outcomes but by no means the only one.