I proposed in an earlier post that the measured high American Jewish IQ is a result of selection. This is the explanation given by Richard Lynn and it seemed to me the most probable, but at that time I did not have any way to confirm it. Recently I found a post by Razib Khan from 2012 [1]. It is made from the results of WORDSUM, which correlates well with verbal IQ. Khan used results after the year 2000. Richard Lynn had earlier evaluated the verbal IQ of American Jews as 107.5 based on a representative sample of WORDSUM results. Below is the plot from Khan. (I did not ask for permission to use the figure, but as he gave the data that I could calculate the figure myself, I guess it is OK this time, thanks Razib.)
Two issues are immediately obvious from the distribution of Jewish scores: the distribution is not normal, it is heavily skewed, and the number of Jewish samples must be small as the curve has many bumps in a homogeneous population.
As there statistical significance of a small sample is not high, I will make a very rough calculation of the Jewish mean. The distribution of the Jewish scores can be roughly approximated by a straight line growing from the frequency 2 at score 3 to the frequency 22 at the score 9. The average is integral from 3 to 9 of xp(x) where p(x)=(5/108)x-1/9 and x is the score in WORDSUM. This yields 6.833 as the Jewish average.
Estimating from the figure, the Christian mean (Protestant and Catholic) in the WORDSUM seems to be 6. About 2/3 of values higher than the mean 6 seem to be below 7.5 and about 1/3 are above 7.5. This gives a rough estimate of the standard deviation of the Christian population as 7.5-6=1.5. The Jewish verbal IQ is thus 0.56 SD over the average of Christians. Christians in the USA are not all whites but almost all Americans are Christians, we can use the USA average IQ of 98 as the Christian mean. With the standard deviation 15 this yields 106.4 as the verbal IQ of American Ashkenazi Jews. This rough estimate does not essentially differ from the value obtained by Richard Lynn.
Ashkenazi Jews have clearly stronger verbal IQ than spatial IQ. Pumpkin person made in [2] an estimate of American Jewish overall IQ starting with data giving the Jewish verbal IQ as 109. He obtained the overall American Jewish IQ as 102. If the verbal IQ is actually closer to 107.5, the overall IQ is very close to the white mean of 100. Still the IQ profile is different. Caucasians typically have stronger spatial IQ than verbal IQ, more notably in Northern Europe. I have no data from the Mediterranean countries, but they are genetically similar to Ashkenazi Jews and may have a similar IQ profile. It is known that American Blacks, and more generally black Africans, have stronger verbal IQ. It is also known that Eastern Asians have especially strong spatial IQ, while their verbal IQ does not essentially differ from that of whites.
Richard Lynn calculated from a representative sample that the Jewish verbal IQ is 107.5, but later wrote a book where he states that the IQ of American Jews is 110. These seemingly contradictory results are both fully understandable. The lower one is from his own measurement, which he considers as one of the two existing valid IQ tests of American Jewish IQ, while the higher figure is the average of many tests, which he does not consider so valid but which are as valid as the tests that he has used when estimating national IQs of other nations and ethnic groups. That is, the figure 110 is as valid as many other estimates of national IQs that Lynn has done in his long research work and it is done by using the same method: by averaging over many tests, not by performing own tests.
I trust the WORDSUM results of Khan better than averaging over many possibly invalid tests. The reason I wrote this post, however, is not to verify Lynns figure 107.5 for the verbal IQ of American Ashkenazi Jews, but to confirm the assumption that the reason for this elevated IQ is selection.
Even if the overall IQ of American Ashkenazi Jews is about 100, there still was selection because there is no other way to explain why the WORDSUM score distribution is so skewed. What I think has happened is the following: the Jews moving to the USA mostly had an IQ that was higher than 100. There were two reasons for it. One was that Czar’s Russia imposed laws that forced many Jews to leave their position or hindered them from getting a position that required higher than average intelligence. Jews, who migrated to the USA, were educated people, who had lost their jobs, unlike native Poles, where those who migrated to the New World were mainly poor and uneducated people from the mountains. The second, and I thin the main, reason was that since 1922 the USA applied IQ-tests for screening immigrants, first by a simple puzzle, then by a full test. There were quotas against immigration from Eastern Europe, but I think that Jews went around these restrictions by migrating through a third country, in the quota of that country. If so, then Jews were directly screened by IQ tests and we should expect to see a skewed distribution as in Khan’s plot.
How would the IQ distribution look like if from an original distribution with the average around 100 would be screened the part below 100 and then we let the population stay for two-three generations? In the first generation after the selection happens the regression to the mean. The average IQ after the selection is about 112, if the part below 100 is dropped. Regression to the mean can be calculated from breeder’s equation if we know heritability of IQ and the mean of the original population. Let us say, the mean was 100 and the heritability is 60%. Then the average IQ of the first generation of descendants is 107.2. in later generations this average does not any more change. As this figure fits well to the verbal IQ of American Jews, we may assume that the verbal IQ of the original population was 100. If so, and as the IQ profile of Jews is strong on verbal IQ, we can estimate that the overall IQ of the original population was 7 points lower, 93. This agrees well with the Ashkenazi mixture of slightly more European than Middle Eastern genes.
The distribution was originally truncated, but it cannot stay as such. For every pair of parents, the genetic IQ of the children is normally distributed around the combined IQ determined by the total parental DNA, both the expressed and unexpressed genes. Because of this, the IQ distribution will smoothen and tend to a similar skewed distribution we seen in the WORDSUM scores.
One conclusion from the confirmation of the selection is that the explanation of Jewish academic achievements by higher IQ cannot be correct. Selection by screening the part below the average IQ increases the frequency of very high IQ scores only by a factor of two. This increase has no relevance what-to-ever in jobs, which people with below average IQ cannot get – these jobs have a screening for IQ surpassing the immigration screening for IQ. The Jewish high achievement can possibly be partially explained by the IQ profile, but higher verbal IQ is actually not preferred in many high IQ tasks. The most natural explanation is being in the correct place in the correct time. This does not mean nepotism, it means that high achievements are done in some places and in some times.
As IQ is a so loaded concept, let us look at sport. For all the time when NHL players could not play in Olympics, Finland usually was the fourth in Olympic hockey. When NHL players were allowed, Finland often got some medal. Genetically the players did not change, but in NHL Finnish top players learned to play better and this is similar with IQ. Every nation, if its population is not really small, has quite enough very high IQ people to get even all Nobel prizes. If the average IQ of a Nobelist is, say 154, then Finland has 800 such people at any time. Even after admitting that only old men get Nobels, there still are 80 old smart men in Finland. But it is not the raw IQ. These people should do groundbreaking scientific work and creating a workplace where one can do such work takes time. There are needed many smart people, funding, scientific traditions and so on, just like in sport there are needed trainers, doctors, funding and so on. One has to be in the correct place in the correct time. In 400 BC that place was Athens, and it mattered not that the average IQ of the Greeks was probably the same as now, below 100.
Short is short, I stop here.
References:
[1] Razib Khan, Verbal intelligence by demographic, Discover, 2012, http://blogs.discovermagazine.com/gnxp/2012/04/verbal-intelligence-by-demographic/#.WoyEwHxG3IV
[2] Pumpkin person, Excellent data on Jewish IQ raises more questions than it answers
Excellent data on Jewish IQ raises more questions than it answers
2 Comments
Jews do not have high IQ’s most Jews have to cheat to get into good colleges. Jews brag that all the Nobel prizes they have won, But that’s in America. You might have a Jew that’s the leader at the University of a group, but all the people doing the most research are White’s and Yellow’s not Jews. Just look at Israel they haven’t won that many Nobel prizes there have they and their all Ashkenazi Jews.
American Jews have in average longer education and higher income. These environmental factors correlate positively with the IQ score, so we can expect that according to IQ tests American Jews are some points above the US average IQ (some points, not 15 points), so are White Americans, and the Asian minority, and many small minorities, like e.g. Nigerians in the USA (a selected group mainly students). This advantage of Jews seem to be environmental. If it were genetic, the IQ scores in Israel would be higher. Jews have got many of Nobel Prizes, but as you point out: he work is usually group work. Israel still has many Nobel Prizes for its size, but the results were often done in the USA. Where there is more money to research, results are better and there seems to be also some bias favoring research done in some countries. The work of some of the early Jewish geniuses (Marx, Freud and lately also Einstein) has later turned out to be incorrect, which may suggest that Jews support each others and in this way their results are evaluated higher than they should be.
If we look at actual IQ tests of Jews, the most recent tests are the ones Richard Lynn
made. His papers date to around 2004 and data is from 1990-96 wordsum test (the American General Social Surveys carried out by the National Opinion Research Centre in the years 1990–1996). It is a vocabulary test and Lynn found the difference 7.5 points. This test
was sent to people, so they could have used a dictionary or asked somebody. In any case, vocabulary correlates strongly with education and reading books, and is therefore strongly
environmental. It is a good measure in a large group, but a poor measure of differences
between groups as a major reason is the environment. Yet, Lynn’s tests are 30-30 years
old. There is no reason to suspect that the earlier tests yielding 115 were wrong. They
were correct at the time when made, but the score has gone down. It went down to 106 in
1990s (as the verbal score of Jews is higher than the total score). We may well expect that
it has gone down since 1990s and today is somewhere around 102-103, which is typical for
whites with a bit better education, as Jews have.
There is no need to assume Jews are cheating in college admissions. Their parents pay the
ridiculous fees and most white Americans do not send their children to these expensive
universities. Only those with scholarship go, but they must be more intelligent getting admission for free.
American Jews were a selected group as they were Russian Jews who lost their positions in 1890s. They very probably had the average IQ around 110, but that was three generations ago and now the advantage is smaller, and will be still smaller in the future. They are a typical white population with a typical white IQ, but they support each others.