National IQs were collected by one psychologist, Richard Lynn, sitting in his study. He said he found them more interesting than collecting stamps. Early in his career he had collected intelligence test results when working in Ireland, but thought that the results would be unpalatable, (they seemed to show that brighter Irish people had emigrated leaving duller ones behind) and so he sat on them for some years. Eventually he began collecting any papers which mentioned intelligence test results in different nations, and thus built up a picture of national intelligence test results. He also began encouraging researchers across the world to collect intelligence results in their countries.

Eventually in 2002 he published the results he had obtained, necessarily based on papers with variable sample sizes, somewhat different tests, administered at different ages. He was open about the fact that some nations had much better data than others, and that for many nations no results were available. Additionally, some nations participated in international scholastic tests, and others did not. The often-forgotten point was that for the first time he had gathered together data not otherwise easily available. He then showed that the results on these paper-and-pencil puzzles correlated substantially with national economic and social variables.

The results were interesting, and began to be incorporated into international studies, though they were infrequently mentioned in international economic studies. His work also drew criticism which cleverly drew attention to the weaker studies, thus seeking to invalidate the overall conclusions, which was that national intelligence levels were closely related to important economic variables. Unfair, but effective. Given the criticisms, I argued that it was necessary to re-work the entire database, listing every study in such a way that each paper could be evaluated in terms of a data quality. David Becker rose to the challenge, and the whole dataset has been worked over again and is available for researchers as an open resource.

One line of criticism of the original work was that the samples were too small and unrepresentative to serve as proper national results. However, since most psychology papers use very small and unrepresentative samples to make claims about all human nature, this would have invalidated all psychology. In fact, psychometric samples tend to be bigger than experimental psychology samples, so have a better claim on our attention.

Despite the re-working of the dataset by David Becker, in the current over-heated climate of denunciation, researchers have been heavily criticised for basing any of their work on national IQs. The usual procedure in academic research is to improve the data by further research, and by suggesting methodological refinements. Oddly, the dominant voice in academia has been those who denounce the work entirely, in a quasi-religious renunciation and condemnation. As the British say, this is somewhat Over The Top. All national measures can be improved, and we should not reject out of hand potential explanatory variables.

Not all national statistics on literacy, income, wealth and health are themselves reliable. A friend who worked at the World Health Organisation headquarters tried to improve their statistics by interviewing every national health minister when they came for a visit. He found that most of the African ones just gave him the figures without linking them to a source he could check, so he felt they represented national pride, not actual verifiable findings.

There are of course other refinements possible, such as comparing cities with provincial locations, so as to determine whether countries have strengths in depth.

Into this fist-fight about national IQs steps Russell Warne, to give his own evaluation of the controversy.

Warne, R.T. National Mean IQ Estimates: Validity, Data Quality, and Recommendations. Evolutionary Psychological Science (2022). https://doi.org/10.1007/s40806-022-00351-y

In a calm and wide-ranging paper, Warne had followed a steady path in his evaluations. He shows that the general picture of national differences has been confirmed as broadly right. He also shows that the low African figures on intelligence tests are probably right when one looks at the similarly low achievements in very simple mathematics shown on scholastic exams. He explains why the low scores do not mean that Africans are mentally defective, as might be the case for low scores in European populations. (Jensen had explained this in 1980).

He also makes the case that Raven’s Matrices may present particular difficulties for African test-takers. I find this a worrying suggestion, because it may imply that Africans are different in kind, not simply in general intellectual power. If this is true, it has severe consequences. Far from just saying “Matrices are not a good measure of African intelligence” it could equally be argued that “Africans cannot follow logical sequences that all other races can solve”. This is not a pleasant proposal, though of course we should be open to testing it. Raven showed that his test items had similar response curves, with only three or four showing deviations from the normal pattern, suggestive of cultural artefacts. Personally, I think that power differences are more likely than type differences. Brains are similar, though they differ in power.

Personally, have listened to David Becker talking about low scores in Africa, I think that on occasion the tests are badly administrated. (As Warne and Rindermann and others have pointed out, many African teachers are of low ability, poorly paid, and consequently poor attenders at school). However, such defective administrations are usually detectable from looking at the results carefully.

As a consequence of his own researches, Warne identifies a matter which must be improved. The Lynn database was put together on the previously acceptable basis of gathering as many studies as possible by searching for them, and increasingly by collaborating with researchers across the world. All this is well and good, but modern standards are more demanding. One has to show exactly how references were searched for. This is a welcome improvement, because it ensures that the database is fully representative of the available publications.

Warne himself has found studies which ought to have been included, and this is of itself a contribution to the research project. Of course, this task is endless, but it would be excellent if more scholars contributed to the joint project. (In reality, this has proved difficult. Some researchers I encouraged to contribute felt that their methodological suggestions had not been taken up, so didn’t participate further. Others simply didn’t want to be associated with national intelligence research, a refusal which damages psychology generally).

Warne has done good work, and has gathered extra data which should be added to the overall database of national intelligence research. He has made a good case for continuing and improving the international research project. His overall judgment is “The database is useful, probably broadly accurate, strongly linked to other national variables, but should be improved by a more systematic inclusion process and the addition of other relevant research”. I hope some scholars will take up the challenge.

A final note: all psychologists should take heart that a single researcher collecting publications on a particular topic can sometimes produce findings of considerable international interest, with strong predictive power, and in line with other massively funded international scholastic studies.