Why are there IQ differences between the races

Deus ex machina

Lately, the social sciences have occasionally dabbled with genetic and psychometric data - which raises all sorts of questions: methodological as well as moral.

In the mid-1990s I spent a year at an American high school in the southern United States. Much was different there, not least the school system. I have spent significant parts of my time memorizing 50 states, 50 capitals, 40 presidents, 9 Supreme Court judges, countless historical facts, and lots of numbers, prime numbers, squares, sines.

The latter are essential in the American education system. From school graduation to doctorate, “standardized tests” in the mathematical part often require faster and better mental arithmetic skills than in Germany - or you simply have the important numbers available by heart. During the year I had ample opportunity to practice - but at first the concept of “standardized tests” was very alien to me. I was uncomfortable with drawing round blobs in pencil (and only with pencil!) On a reply card that would later be read by the computer, although sometimes I was uncomfortable with the entire approach of imparting and evaluating knowledge at high school .

A completely different dimension of performance measurement is currently being discussed in the USA - namely the content of a doctorate at Harvards Kennedy School of Government. A doctoral thesis from 2009 with the title “IQ and Immigration” has been hotly debated for a few days between the east and west coasts, and probably most of all on both coasts.

© Own picture

The then doctoral student, now an ex-research fellow at a conservative think tank, tries to show that firstly, Hispanic immigrants in the USA have, on average, a lower IQ than the white non-immigrant population; second, the reasons for this are partly genetic; and third, that is bad for the US for various reasons and the immigration policy should be adjusted accordingly.

The supervisors at that time, three well-known professors, have now taken a position and claim that the work is technically and methodologically perfectly clean, even if the conclusions and policy recommendations are premature and actually far-reaching.

The public discussion about whether this is racism, whether it is allowed to do a doctorate, and what consequences are now appropriate, begins to turn in circles after a few days - the most astonishing thing about the “Affaire Richwine”, however, is that so far hardly anyone has written the dissertation seems to have read. Understandable with 166 pages, but even a fleeting reading is enlightening.

Richwine explains in detail the concept of IQ and the related G-factor with many references to the American Psychological Association (APA), although of course he cannot completely dispel the suspicion that the IQ test measures IQ, and that is what IQ is what the IQ test measures. He uses various data sets from the American population to show that the average IQ of Hispanic immigrants is lower than that of “Caucasians”, even among the second and third generation of immigrants.

© Own picture

The second part of the dissertation is followed by a very informative list of socio-economic factors that statistically correlate with the IQ (myopia, impulsiveness, generosity, willingness to cooperate - the list is so exhaustive that almost everything seems to be positively or negatively correlated somehow). From this abundance, among other things, the “social capital” picks out - a term that has been fashionable in the social sciences for some time, but just as difficult to pin down as the IQ. A table on p. 103 in which Americans were asked is particularly interesting. how much they trust the neighbors in their neighborhood. If a high proportion of black or Hispanic immigrants lived in the neighborhood, trust was significantly lower, but for many Asian immigrants only a little lower. As soon as the author, however, took into account the influence of income, education, etc. both of the respondents and in the quarter (quasi: calculated out), the contribution to uncertainty of all three ethnic groups was similar.

From this one can conclude two things: first, that one obviously trusts one's neighbors less if they have a different cultural background. Surprise. Second, that this is of course very far from causal identification, and that the author's evidence, despite much painstaking small-scale work in collecting sources and statistical tests, is possibly incomplete.

I haven't delved into the statistical details, but presumably one could just as well use the data to argue that immigrants and their children are permanently disadvantaged in the education system and simply do worse on standardized tests, i.e. the differences are more related to cultural characteristics, the socio-economic Environment and upbringing. There are corresponding studies, by the way, and the results show that with “transracial adoptions” the influence of the environment is great.

© Own picture

Accordingly, the conclusions that Richwine draws are certainly high. One can certainly argue about whether one considers the work to be worthy of a dissertation - but one can also argue about what one should and may measure in an enlightened, opinion-free society.

Richwine is not alone with his topic, psychometrics as such is a well-established field of research, and in recent times other disciplines like to attack these topics, often with controversial results. At the forefront of this is a researcher named Lynn who, among other things, comes to the conclusion that the! Kung people in Africa have the lowest IQ of all, roughly the size of an eight-year-old child. Except that, of course, an eight-year-old child would certainly not want to be abandoned in the hostile climate of the Namibian desert, while eight! Kung children can get by there, and the! Kung are generally veritable survivors, given their hostile climatic environment.

In general, the work of that researcher - apart from methodological deficits - raises the question to what extent such results are simply the consequence of a pronounced western centrism in research. Anyone who grew up in a Western culture with Western categories and thought processes probably scores better in such tests than people from other cultures, and such cultural differences may be lasting and persist for a few generations after emigration or immigration.

© Own picture

Which brings you back to the initial question, what the IQ test actually measures. Some time ago, two other renowned researchers dealing with genetic diversity of populations and economic success struggled with a similar deficit. They actually used a physiological measure of genetic diversity, but only for a relatively small group of populations. This in turn correlates clearly with the distance to the cradle of mankind in Africa (more precisely: Addis Abbeba in Ethiopia). They then work with this “migratory distance” as an approximation for genetic diversity, whereby the question also arises here as to what these two variables actually measure. And whether there aren't other factors that were left out of the analysis, but could also explain the differences in income and development of countries. In this case at least there was a serious discussion, not only about the moral limits of social science research, but also about the methodological weaknesses and deficits.

Both research papers illustrate very clearly the inherent limits of social science research and show how important a critical discourse is about methodological deficits and methods and credibility in general. Science should never work uncleanly. Admittedly, this is a debate in which above all the experts in the disciplines should participate - on the other hand, societies as a whole must discuss and define which questions they want to ask - taking into account the possible answers.

Keywords: everyday life, work culture, education, black box, discourse, ethics, research, society, psychology, statistics, numerical understanding
To press
Send post by email

What science is allowed to do

By Sophia Amalie Antoinette Infinitesimalia

Lately, the social sciences have occasionally dabbled with genetic and psychometric data - which raises all sorts of questions: methodological as well as moral.

An error has occurred. Please check your entries.

Send post by email

Many Thanks
The post was sent successfully.