Psychology Today
Is it true Feynman's IQ score was only 125?
Feynman was universally regarded as one of the fastest thinking and most creative theorists in his generation. Yet it has been reported-including by Feynman himself-that he only obtained a score of 125 on a school IQ test. I suspect that this test emphasized verbal, as opposed to mathematical, ability. Feynman received the highest score in the country by a large margin on the notoriously difficult Putnam mathematics competition exam, although he joined the MIT team on short notice and did not prepare for the test. He also reportedly had the highest scores on record on the math/physics graduate admission exams at Princeton. It seems quite possible to me that Feynman's cognitive abilities might have been a bit lopsided-his vocabulary and verbal ability were well above average, but perhaps not as great as his mathematical abilities. I recall looking at excerpts from a notebook Feynman kept as an undergraduate. While the notes covered very advanced topics-including general relativity and the Dirac equation-they also contained a number of misspellings and grammatical errors. I doubt Feynman cared very much about such things.
Do you think we will ever find another Einstein?
This is a very difficult question. Einstein was special for many reasons, and was the dominant figure in 20th century physics, if not all of science.
In the modern era many more people have access to advanced education-think of India and China plus all of the developed world. While I believe even an average scientist these days is quite an exceptional person, both in terms of ability and the amount of training he or she has received, it is much more difficult now to stand out in the crowd. Think of the NBA: the average player today is much better than the average player of 50 years ago. Any guard in the league is an athletic freak of nature. But when they play against each other they are relatively evenly matched. It may be a long time before we encounter another giant like Einstein who so far surpasses his contemporaries.
If you are interested in psychometrics and the far tail of cognitive ability, I recommend several of Jonathan's papers, including:
Wai, J., Lubinski, D., & Benbow, C. P. (2005). Creativity and occupational accomplishments among intellectually precocious youths: An age 13 to age 33 longitudinal study. Journal of Educational Psychology, 97, 484-492. pdf
On the highly verbal Terman test, Shockley and Alvarez both failed to meet the 135 threshold for further longitudinal tracking. According to Shockley, he initially tested at a 129 on the Terman exam.
ReplyDeletehttp://www.quora.com/Who-are-some-famous-scientists-who-scored-low-on-IQ-tests
"William Shockley's was 129 when testing for the Terman Study, then 125 when tested a year later (his 2007 bio)
Luis Alvarez's was below 135 (he failed to qualify for Terman Study)"
> Albert Einstein was a visual thinker who failed his high school language requirement and relied on visual methods of study (Holton 1971-72). ...
ReplyDelete> skewed heavily towards S and who also excelled at M, but was much lower at V, wouldn't this fit the stereotypical East Asian cognitive profile to a T?
Incorporating Einstein into the East Asian population (new Jews, indeed) is... a slight stretch. Not to stop this entertaining exercise in "race science" but you might want to get some facts in order before picking out a Chinese name for Albert.
Einstein did not fail his high school language requirement. He passed all requirements for the Matura, including languages ( http://en.wikipedia.org/wiki/File:Einstein-matura.jpg ), after almost gaining admission to ETH at age 16, without a high school diploma, shortly after arriving in Switzerland. Which is to say that Einstein caught up several years of the Swiss high school curriculum in one year, in addition to pursuing his ideas about physics. Failing the ETH exam was primarily from not knowing French well enough; he had just moved to Switzerland without prior mastery of the language, was on his own in a new country, and ETH required passing a French language exam written for a population who spoke the language fluently. Einstein was a prodigy but not a polymath, and he failed.
Einstein's verbal skills were superlative, in both German and English. He was a prolific writer, had volumes of letters and nonscientific writings published, gave speeches on subjects such as politics and pacifism, and of course had a well known way with words that led to a number of striking quotations.
Einstein was not "autistic" but quite social and happy to be a celebrity in his Berlin days. He was also socially and politically astute (and active, e.g., in anti-Nazism, Zionism, pacifism and other causes), and practically ambitious enough to organize the patenting of the refrigerator technology he developed with Szilard. He was a womanizer of sorts. The idea of Einstein as a socially maladjusted Spock or anorak is preposterous, though of course he had his eccentricities. It was the move to the USA with its relative lack of German speakers (Einstein he went to IAS primarily to talk with Goedel) and Jews, and after 1938, the destruction of the people and culture he had grown up in, that led to a lot of Einstein's incongruity, in his environment. Of course, he could also afford to indulge his eccentricities, as the world's leading scientist.
Einstein was also relatively musical. He played the violin when he had the opportunity (e.g., the famous picture with Ehrenfest). He was not an exceptional talent but certainly some form of musicality was part of his profile of abilities.
Feynman shared many of the same traits.
Musical talent is a sign of autism.
ReplyDeleteFrom Lynn
ReplyDeletehttp://books.google.com/books?id=XaiF6MeXFmQC&pg=PA95&lpg=PA95&dq=albert+einstein+poor+verbal+ability&source=bl&ots=FVRjG4gIfu&sig=gp_uGi20iHD-s438DHOhzEASJ6c&hl=en&sa=X&ei=snL5Tu6ELIyA2QXegNm8CA&ved=0CCYQ6AEwAQ#v=onepage&q=albert%20einstein%20poor%20verbal%20ability&f=false
"Albert Einstein was brilliant at physics and mathematics, and at the age of 12 he found an original proof of Pythagora's theorem. However, he was weak in language based subjects and at the age of 16 he failed the entrance examination at Zurich because of his poor performance in languages, literature, history, and art.)
"Feynman shared many of the same traits."
Like botching the English and History portions of the entrance exams to Princeton. Any thoughts about that, given that English was obviously his native language?
"had a well known way with words that led to a number of striking quotations."
The most striking quotation may be the one I posted above.
http://thinkexist.com/quotatio...
“These
thoughts did not come in any verbal formulation. I rarely think in
words at all. A thought comes, and I may try to express it in words
afterward.”
-Albert Einstein"
"Einstein was also relatively musical. He played the violin when he had the opportunity"
Well now, it seems to me that there's another ethnic group in the US who are considered relatively musical and stereotypically gravitate towards instruments such as the violin...
Given everything we know about Einstein, and given the repeated references in the relevant literature to him being primarily a visual thinker, I think I am correct in stating that Einstein in all likelihood skewed heavily towards S and M at the expense of V.
Richard "MCAT overpredicts Asian performance" Lynn is not exactly an Einstein scholar, if he is in fact a scholar of anything. Einstein did not skew toward anything *at the expense of* the verbal, except in the degree to which he allocated time to one pursuit rather than another. You should translate his list of grades from the Matura before telling us how bad he was at languages etc (as opposed to merely not having mastered the entire Swiss high school curriculum before moving to a new country). His "V" if you call it that was presumably very high, though not at the stratospheric level of his other abilities. Much higher than Feynman or Terry Tao to whom you compare his cognitive profile. I'd recommend reading, e.g., Einstein's volume of Ideas And Opinions -- much of it written very strikingly in a second language -- before judging him to be skewed away from the verbal.
ReplyDeleteIt can also be a sign of musical ability.
ReplyDeleteNote how Einstein had an editor and a translator. I doubt it's that hard to scrap a book with inputs from professional writers, and much of "Ideas and Opinions" was originally written in German. Since it was just an assortment of excerpts, I find it hardly convincing the book points to his literary genius. It wasn't like he was a prolific writer such as Bertrand Russell. It isn't uncommon for an expert in any area to have many ideas; I can think of Kasparov as one example. Most people wouldn't even be reading this book if it didn't have Einstein's name on it.
ReplyDeleteVerbal IQ of 125 still probably correlates to around 650 in SAT-V. Hence most of the eminent scientists who are considered weaker are probably around the levels of 700 SAT V scorers, not weak by any means even compared to typical Ivy League students.
Einstein had an editor because the work was collected from a larger set of unpublished material, and a translator because much was written in German but the book was published in English. If the translation is in any way faithful (or other books containing a translation of his letters, or scientific work) then he wrote beautifully in German, and if the material originally in English is any indication, he did very well in a second language that he did not necessarily study at school and used relatively little until late in his life.
ReplyDelete
ReplyDeletehttp://news.bbc.co.uk/2/hi/health/2988647.stm
>I find it interesting that when some well known giant of science
ReplyDeletesupposedly scored low on an IQ test, it was generally speaking a
verbally loaded exam.
Who are these giants of science who scored low? Two people tested just below exceptional as children when IQ tests were a novelty, multiple choice testing of any kind was uncommon, before the Flynn effect showed its effects, and before the tendency to educationally accelerate potential giants of science (increasing their vocabulary, mathematical skills and other exposure that would raise IQ scores). "V" measures would have been in some measure a function of social class, such as the degree to which people were socialized to speak, argue, expound, or read when it was not strictly necessary. It is entirely plausible that the 12 year old Alvarez and Shockley (or other overlooked child scientists) did not spend their time all that academically as children or didn't completely understand what some of the questions were asking.
As a matter of probability, since the exams are not perfect, given enough future Nobel prize winners tested, some of them have to score 1-2 SD "low" compared to the truth. That it happened twice in one sample may say that there was something wrong about the test or the way it was administered and not some overarching fact about hazy concepts of V, M, and S.
There is also no shortage of genius scientists of the V and not S type. Grothendieck, or any number of logicians come to mind. In physics it is less common since the training begins with mechanics. Leaning relatively strongly toward V or toward S as you call them, is pretty common in the population and, in addition to being generally very smart, it could be that either preference can be fine (depending on the field) and it is the presence of unusual levels of some third or fourth attributes (eg., "M", creativity, nonconformity, drive, ganas, testosterone, serum uric acid, yadda yadda) that make the difference in science.
http://news.bbc.co.uk/2/hi/health/2988647.stm
ReplyDelete"Einstein and Newton 'had autism'
Albert Einstein and Isaac Newton may have suffered from a type of autism, according to experts."
Interesting. Didn't Matthew Carnegie point out earlier the link between autism and spatial ability?
It's probably a boring cliche, but I think that Einstein's role and his intellectual abilities have been overestimated by the general public. He was simply in the right place, at the right time, with the right collection of knowledge and skills. In terms of IQ scores, I'm quite confident that his IQ was considerably lower than IQs of his contemporaries Hilbert and Poincare (and those, in turn, were below Riemann.)
ReplyDeleteI don't fully buy the bit about "access to advanced education" and it being more difficult to stand out in the crowd. Even in 1900 in Europe, if you were two sigma above the mean, you'd end up getting some kind of advanced education. Total population of Western Europe and the United States in 1900 was somewhere around 400 million. Today it is 900 million. There was exponential growth in Africa and India, which had virtually no effect on the scientific world because of low means in those parts. That leaves China - lots of people, supposedly decent education and high mean IQ, but, so far, that has not translated into any apparent kind of science dominance: to this day, the number of Chinese Nobel laureates in physics or chemistry or Fields medalists stands at exactly zero. (There were, I think, two or three who were born in China and then emigrated to the West in their teens or twenties.) We'll see if that changes.
Recall, though, that Terrance Tao could score in the top 1% of 12-13 years old in verbal ability before he was 10. In the SMPY sample, the average SAT-V score at age 13 of STEM professors was 534. 430 is the top 1/200 and 630 is the top 1/10,000. Also, I think it's very clear that Einstein had extremely high spatial ability and he mentions thinking visually. His neurology is consistent with that too, but he was also purported not to make stand-out grades in art and geography--other areas where high spatial ability is correlated with success.
ReplyDeleteSpatial (or verbal) ability is not correlated with success when you don't care, invest no time, or are busy with other things. Academic records can tell us positive things about the abilities of brilliant and unconventional scientists but are less accurate as negative indicators.
ReplyDeleteGood point. Here's a measure that is less noisy, but afflicted with a slight degree of time-lag:
ReplyDeletehttp://www.sciencewatch.com/dr/sci/misc/Top100Chemists2000-10/
If we assume that there are 2.5 billion people living in countries with "good" national IQs and healthy education systems, and these were all equal, I'd expect that the list would consist of 40 Chinese, 30 Europeans, 10 Americans, 6 Russians, 5 Japanese, 2 South Koreans, and a few people from elsewhere in the world.
By my approximate count, Russians and South Koreans are right where I expect them (6 and 3), Japanese are underrepresented (I only see 2 names), and Chinese are badly underrepresented (I count 10, not including one guy with Chinese last name who was really born in Ohio.)
Thomson Reuters also has an equivalent list for Materials Science.
ReplyDeletehttp://sciencewatch.com/dr/sci/misc/Top100MatSci2000-10/
What's interesting about this list is that while East Asians only make up 17/100 of the names listed, the top 6 most heavily cited materials scientists of the past decade based upon per capita paper citations are East Asian.
"It seems like Murray uses +1 SD as the threshold for those considered the "brightest""
ReplyDeleteIf we're talking about being difficult to "stand out in the crowd" for the person of Einstein's intelligence, +1 SD is way too low a threshold. +1 SD is 16% of population. In the article above, Steve uses the word "only" to refer to Feynman's alleged score of 125 - the implication is that 125 is too low to become a good physicist.
What really happened is that the pool for exceptional scientists increased roughly by a factor of 3 between 1900 and 2000 (we'll call it 6 when mainland China completely catches up), and, at the same time, the bar that determined how good you had to be in order to be admitted into scientist-world was sharply lowered.
On top of that, science got much, much harder (all the prerequisites needed to derive Special Relativity can be condensed into 50 pages and taught to a bright 12-year-old. All the prerequisites needed to derive Minimal Supersymmetric Standard Model would fill 5000 pages and it takes a typical scientist with this specialization about 8 years of postsecondary study to get sufficiently up to date.) And the number of sub-fields exploded.
So maybe it's not that it's harder for an Einstein to stand out in the crowd, rather, people are still the same, but it's gotten incredibly hard for an Einstein to discover something equally groundbreaking and significant.
Was it "less sociopathic" or "more autistic" that you were proposing as a description of E. Asians? I had forgotten. But more to the point I was wondering whether you had forgotten to answer the Putnam calculations that you had specially requested in the other discussion, because they appear to indicate (on your own assumptions) numerical overadmission of US E.Asians to the top universities under present policies, relative to the natural math ability levels, given the Asian nationalist variety of "g" and HBD hypotheses. I understand if you don't like the conclusion but was curious as to your opinion on the correctness of the calculations. For example, did I mis-state the HBD/g assumptions about "shifted normal distributions with similar SD", that are at the heart of that particular religion, so often promulgated on this blog?
ReplyDeleteDidn't Einstein learn to talk late? Late talking seems like a major sign of either stupidity, autism or both. And yet Einstein is believed to have been unbelievably brilliant. The obvious explanation for why someone so brilliant would be retarded in their speech development is autism.
ReplyDeleteIn addition, einstein's verbal problems extending into adulthood seem especially rare for someone so brilliant, especially a brilliant Ashkenazi Jew where verbal skills are usually a strength. It might take a powerful neurological condition like autism to override this ethnic cognitive profile.
Also, the fact that Einstein was so obsessed with certain numerical concepts is further evidence of autism.
Autistics have trouble recognizing sarcasm from others, but that doesn't mean they can't be sarcastic and funny themselves. I also think that if Einstein was an autistic, a lot of the autistic traits would be mitigated by his presumably superlative IQ.
ReplyDeleteesmith, in terms of actual numbers, what IQ range do you believe Einstein was in?
ReplyDeleteMental illness in math and physics families: (few examples plus research)Bertrand Russell: mathematician and philosopher, daughter, schizophrenia, suicideAlbert Einstein, physicist, son, schizophrenic, spent life in institutionKurt Godel, paranoid, starved himselfJohn Nash, mathematician, schizophrenic, and his son, also schizophrenicJ. Robert Oppenheimer, physicist, daughter, suicide"study, conducted by Jon Karlsson, M.D., Ph.D., director of the Institute of Genetics in Reykjavik, Iceland, has found an intriguing relationship between math talent and psychosis susceptibility in the Icelandic population. Results were published in the April British Journal of Psychiatry"reference: http://www.schizophrenia.com/sznews/archives/000572.htmlhttp://www.ncbi.nlm.nih.gov/pubmed/15056577 quote from article:method:"the two top graduates from the single school were selected for those born 1851–1910, and for those born 1911–1940 the top male graduate from each of the two schools was chosen. Thus, there were in this survey 180 index scholars, approximately 0.1% of the total population of 171 000; all were male. These individuals and their relatives were then evaluated for rates of ever having been admitted to the mental hospital, comparing their risk of hospital-treated psychosis with the figure of just under 1% already established for the general population."Results:"In this group there were 12 people with schizophrenia and 15 with an affective disorder. Using either approach, the rates are more than doubled in both index cases and close relatives."
ReplyDelete"Didn't Einstein learn to talk late?"
ReplyDeleteYup. Apparently Feynman was afflicted with same condition.
http://en.wikipedia.org/wiki/Language_delay
"Neuroscientist Steven Pinker
postulates that a certain form of language delay may be associated with
exceptional and innate analytical prowess in some individuals, such as Albert Einstein, Richard Feynman and Edward Teller.[1]"
Interesting. I wonder what evidence pinker has for his theory. I think a simpler explanation for the exceptional STEM achievements of eminent late talkers is that they're just really high functioning autistics and the autistic single minded obsession with math related concepts allowed for superlative STEM achievements, especially if these individuals were also blessed with high g, math/spatial talent and/or savant like abilities.
ReplyDeleteI don't see what this has to do with autism necessarily, but with respect to the speculations about Asians, is there a difference in the time at which babies begin to talk? (And are you going to answer the Putnam calculations? Just asking.)
ReplyDelete> Conservatory students who are native speakers of tonal languages
ReplyDelete(languages like Mandarin and Vietnamese
> in which pitch conveys meaning)
display perfect pitch more frequently than do their English-speaking
counterparts.
Was it ever resolved whether this is genetic, or from learning a tonal first language?
"Was it ever resolved whether this is genetic, or from learning a tonal first language?"
ReplyDeleteI'm not sure, but the second half of what I pasted addresses your question.
Alternately, prevalence among Asians may have a genetic basis.
Another study, which did not consider which language subjects spoke,
found that 32 percent of Asian-American music students had perfect pitch
compared with 7 percent of non-Asian-American music students."
Yan, East Asians may resemble autistics but there's also a popular theory that Ashkenazi Jews have aspergers (which is a variant of autism where verbal skill is much greater than spatial). The theory argues that aspergers (which impairs social comprehension) is the reason Jewish populations have been persecuted and had trouble getting along with other ethnicities.
ReplyDeleteI'd say somewhere around 160.
ReplyDelete'91-07 USA IMO Team Members
ReplyDeleteEthnicity/Name/Putnam Fellow?
*W-white E=East Asian I=Indian
1. W Rosenberg, Joel-No
2. W Kedlaya, Kiran-Yes
3. W Kleinberg, Robert-Yes
4. E Ng, Lenhard-Yes
5. W Sunitsky, Michail-No
6. W Breydo, Ruby-No
7. E Huang, Wei-Hwa-Yes
8. W Levin, Sergey-Yes
9. W Schultz, Andrew-No
10. W Dittmer, Andrew-No
11. E Wang, Stephen-Yes
12. W Bem, Jeremy-Yes
13. W Chklovski, Tim-No
14. W Khazanov, Aleksandr-No
15. W Lurie, Jacob-No
16. W Shazeer, Noam-No
17. W Weinstein, Jonathan-No
18. W Nichols-Barrer, Josh-No
19. E Chang, Christopher C.-No
20. E Chyung, Jay H.-No
21. W Gnepp, Andrei C.-No
22. W Bosley, Carl-No
23. W Korn, Michael-No
24. W Saltman, Alexander-No
25. W Curtis, Nathan-Yes
26. W Miller, Carl-No
27. E Chen, Li-Chung-No
28. W Clyde, John-No
29. W Stronger, Daniel-No
30. W Barton, Reid-Yes
31. W Carroll, Gabriel-Yes
32. W Schwartz, Sasha-Yes
33. W Lacker, Kevin-Yes
34. W Valiant, Paul-No
35. W Wood, Melanie-Yes
36. E Loh, Po Shen-No
37. W Detlor, Lawrence-No
38. E Lee, George Jr.-Yes
39. I Nir, Oaz-No
40. E Liu, Ricky I.-Yes
41. E Loh, Po-Ru-Yes
42. E Le, Ian T.-No
43. E Liu, Tiankai-Yes
44. E Shin, Dong (David) H.-No
45. W Kane, Daniel-Yes
46. W Kaseorg, Anders-No
47. E Xue, Alex-No
48. W Pixton, Aaron-Yes
49. W Lipson, Mark-No
50. E Zhang, Yan-No
51. W Golberg, Oleg-Yes
52. W Miller, Alison-No
53. E Zhang, Tony-No
54. W Ince, Matt-Yes
55. W Cordwell, Robert-No
56. E Gong, Sherry-No
57. E Kim, Hyun Soo-No
58. W Lawrence, Brian-Yes
59. W Mildorf, Thomas-No
60. W Price, Eric-No
61. W Abel, Zachary-No
62. W Brady, Zarathustra-No
63. E Ko, Taehyeon-No
64. E Sun, Yi-No
65. I Tripathy, Arnav-Yes
66. E Zhai, Alex-Yes
67. W Larson, Eric-No
68. E Leung, Tedrick-No
69. W Christiano, Paul-No
70. I Kishore, Shaunak-No
Gand Total-70 (100%)
Yes-24 (34.3%)
No-46 (65.7%)
W Total-45 (64.2%)
Yes-15 (33.3%)
No-30 (66.7%)
E Total-22 (31.4%)
Yes-8 (36.4%)
No-14 (63.7%)
I Total-3 (4.3%)
Yes-1 (33.3%)
No-2 (66.7%)
Your debunking of the false claim made by "Sineruse" doesn't surprise me one bit. It's been decades since I paid much attention to the Olympiad or Putnam results, so when I was skimming the previous thread, I found his complex and detailed analysis alleging a clear pattern of East Asian under-performance at higher levels not necessarily implausible.
ReplyDeleteHowever, I then decided to look at the actual data, and discovered that in recent years, East Asians had constituted nearly 60% of the Olympiad winners and *also* nearly 60% of the Putnam winners, which hardly seems like evidence of "under-performance" to me. This irritated me, since I don't like falling for dishonest propaganda.
He then claimed that many of the East Asian Putnam winners were not American residents; this is certainly possible, but he hasn't provided any actual evidence in support. He also began slicing and dicing the Putnam lists in all sorts of complex ways to prove that the 60% figure my eyes had seen was just some sort of optical illusion. Needless to say, this is a favorite trick used by dishonest propagandists.
It's clear he's devoted a great deal of time and effort to this extremely obscure subject and is also a zealous propagandist. I almost half-suspect he might be some sort of PR consultant retained by the universities who have faced harsh criticism for their obvious system of racial quotas against East Asian applicants.
白鬼子们像是热锅上的蚂蚁，天天跑到这里来证明中国人不行。狗皮膏药商人的日子。
ReplyDeleteThis seems pretty plausible to me. I'd suspect that a sub-clinical degree of mental illness is often a very useful contributor to ultra-high intellectual achievement.
ReplyDeleteIn fact, I'd think that this would be one of the likeliest hinderances to ultra-high achievement by East Asians, coupled with their generally lower levels of aggressiveness, fanaticism, and anti-social tendencies. Obviously, all of these traits are mostly negative ones, but under the right circumstances---and when combined with very high intelligence---can produce positive results of a revolutionary nature.
Consider just the "anti-social" factor. A few years back, Bruce Lahn and his team were producing absolutely stunning results in analyzing the evolutionary sweeps of brain-related genes, and Lahn seemed on a possible trajectory to become one of the leading scientific figures of our era. However, some feared that these research results might have highly un-PC interpretations, and supposedly Lahn was warned of the dangerous implications for social harmony in our multi-cultural society. Eventually Lahn, a Chinese immigrant, apparently backed down and shifted his work to areas less likely to conflict with America's reigning social taboos and thereby provoke controversy.
By contrast, consider the case of William Shockley, by all accounts a brilliant but exceptionally anti-social individual. He was enormously eager to violate the deepest social taboos and to do publicly to the widest possible audience. As a result his name was blackened and his reputation totally destroyed within the reach of the mainstream media.
Obviously, the Lahn and Shockley examples are extreme cases, and neither is exactly an ideal example to emulate. But I do think they illustrate the considerable skew between certain European and East Asian personality traits.
But, there are also logical challenges to this line of thought.
ReplyDeleteEast Asians do well both at the center and the means of quantitative ability and mathematical achievement. Meta-analysis does not always find the same to be true of autistic individuals.
http://aut.sagepub.com/content/11/6/547.short
Ethnic incidence of autism does not find a clear, consistent tendency for Asians to be autistic more often than other groups. Some data implies white people are more often than Hispanics. Factors like sampling could limit conclusions and so forth.
http://www.uth.tmc.edu/clinicalneuro/institute/2005/Croen,%20Grether,%20Hoogstrate,%20et%20al.pdf
>Your claim about East Asian over prediction regarding IMO-Putnam seems
ReplyDeleteto be false.
No, you got the data wrong. Ron's celebrations are, shall we say, "premature".
After correcting for errors that you introduced into the analysis, even a
data set of this tiny size supports the I > W > EA claim about
over- and under- performance. It would get
much stronger (and more statistically significant) if you use the 60-80
data points per year coming from the whole Putnam high score list, and
account for multiple Putnam results per individual and the total results
in high school. At a minimum the top 15 rather than top 5 should be
considered, to reduce the noise in the top 5 results due to small sample
size, foreign student effects, and an excessive impact from the
multiple winners (nearly all of whom are not E. Asian Americans, as enumerated in the other thread).
>I excluded '08-'11
You also excluded a third of the white Putnam winners by (mis)taking the
US IMO team for the Putnam selection pool. It is a lot broader than
that, and the only reasonably proxy for which data is easily available
is the USAMO Top 24 (winners + HM) or, more credibly, the list of USAMO
qualifiers. There are students who in high school go to the physics or
computer olympiads or various summer programs, instead of IMO, or just
are not well-trained or interested enough to win the math olympiad, but
later do much better on the Putnam. Most of the ones who later win
Putnam at least qualified for USAMO when there were 150-200 who took the
test, and all of them do so in recent years when the number of
qualifiers increased to 400-500. Using a more realistic selection pool
and set of Putnam winners would greatly increase the number of E.Asians
in the initial (pre-Putnam) pool and the number of whites in the list of winners.
The upshot, once again, is that the demographics of the Putnam winners,
top 15, or top 60-80 (HM+) is depleted in E.Asians when compared to
the USAMO lists; and that when looking backward from the Putnam list to the
ranks of the same students on the USAMO, after college admission the ranking of whites has
generally increased relative to that of E. Asians. I explained in the
other thread what this calculation looks like, i.e., several-to-1 odds
ratios disfavoring US E.Asians, when controlling for time period by
following one cohort of students, such as those eligible to take the 2009 Putnam contest.
Another problem is the choice of 1991-present as the time period. The
USAMO and the IMO team selection became much more difficult and with a
greatly enlarged selection pool during that time. The exams became
longer (e.g., the USAMO became 9 hours long in imitation of the IMO
format) and the number and difficulty of problems increased. The Putnam
may have also become a more difficult exam, but not nearly as much.
This means that the USAMO results and IMO qualification are more
predictive of Putnam results than they were in the past, and a sample
that has most of its Asian students in the recent years will make Asian
credentials seem more predictive than they are, relative to non-Asians
where the predictivity is being assessed more for the early years.
This message is long so the calculations will be in the following reply.
Lurie is presently a full prof at Harvard. Of interest, Jacob Rasmussen an olympiad alternate, is an associate prof at Cambridge.
ReplyDeleteFun data point.2010 Putnam fellows:
ReplyDeleteYU DENGMassachusetts Institute of TechnologyBRIAN R. LAWRENCECalifornia Institute of TechnologySEOK HYEONG LEEStanford UniversityCOLIN P. SANDONMassachusetts Institute of TechnologyALEX (LIN) ZHAIHarvard University
"I almost half-suspect he might be some sort of PR consultant retained by the universities who have faced harsh criticism for their obvious system of racial quotas against East Asian applicants."
ReplyDeleteWhat do you think about this? See:
http://infoproc.blogspot.com/2011/12/asian-admissions-statistical-prediction.html#comment-393766127
"It's clear he's devoted a great deal of time and effort to this extremely obscure subject.."
ReplyDeleteFor those who do not know, Sineruse has been obsessing over East Asians for years. Years ago, he used to post extensive on the forum College Confidential.
I have never seen someone so obsessed with another ethnic group since Adolf Hitler's obsession with the Jews back in the day. I don't think this kind of behavior is normal or healthy.
Interesting take. I would add a handful of comments.
ReplyDeleteThe first is that mathematical ability is a narrow, albeit important, avenue in the neighbourhood of physical sciences, and thus it being the case that virtually all of the examples given are mathematicians, one is a bit at risk of wandering onto slippery rocks by projecting from "mathematicians are more prone to mental illness" onto the space of "scientists are more prone to mental illness." It may be so, but I'd like to see some empirical evidence that top biologists, chemists, genetics, or other less quantitative specialities are equally affected.Second, RKU raises some interesting points, though I do not agree that East Asians have generally lower levels of fanatacism, aggressiveness, or anti-social tendencies. One would need to define "anti-social" for a start. And as for fanaticism, some of the most egregious examples of cult-of-personality-based savagery have taken place in China (the Boxer Rebellion comes to mind, as does the notorious Great Proletarian Cultural Revolution, the likes of which has never been seen in the multi-cultural U.S.) Third, I've seen arguments in the sphere of neurology that certain aspects of, say Asperger's Syndrome (extreme focus, ability to perform what most consider boring, detailed tasks for prolonged periods with ease) correlate strongly with mathematical ability. I've not seen similar research correlating Asperger's with, say, people like Watson and Crick. Could be just that the link has not been uncovered; don't know.Finally, I wonder if it's helpful to conflate ethnic "personality traits" - which are almost surely heavily influenced by culture - with intellectual ability, which i would suspect has a much higher genetic component to it? It's obvious, I think, that the variances among so-called "European personality traits" if you compare say, Germans, Swedes, and Americans (mutts, really), are similar to the variances among Chinese, Japanese, Filipinos, and Southeast Asians. How would one even compare a multi-ethnic country like the US (is it "European?") to Germany or Japan for that matter?But you're absolutely right that research in this area tends to get tamped down because the possible conclusions could be catastrophic. Shockley, the Nobel Laureate, ended his career in shame and ignominy. I have never heard of Bruce Lahn, but is his shying away from violating social taboos that different from, say Galileo or others who, when confronted with the unfortunate social consequences of their work, made similar decisions? I find thus the likelihood that putative lower levels of "anti-social" tendencies are likely to be a significant hindrance to ultra-high East Asian scientific achievement, and my own anecdotal observation of some highly dysfunctional geniuses I encountered in graduate school give me faith that in this arena, Asians are likely to compete pretty successfully with their white peers...
Well, I don't have any expertise in psychological research, so my suggestions were pretty speculative. But I do think it's very likely that East Asians are innately much less likely to be aggressive or anti-social than most of the various European peoples, with Nordics and Germanics certainly falling closest to the East Asian side of the European distribution curve. I think there's a great deal of circumstantial evidence supporting this sort of idea, based on international data on crime, delinquency, and personality responses, and with the clear pattern of these traits generally being rather time- and location-independent and relatively unaffected by language or local culture. To some extent the focus on the heritability of intelligence/IQ has masked the somewhat similar profile of lots of other major personality traits, partly since these tend to be much more difficult to quantitatively and objectively measure.
ReplyDeleteProbably my use of the term "fanaticism" wasn't really correct, since I really meant fanaticism in the face of counter-vailing social pressures. And while Chinese history does have a great deal of large-scale violence, the patterns tend to be much lower than that of most European countries relative to influences such as desperate poverty or political disorder.
As for Lahn, a few years back his fascinating discoveries were regularly making the front-page of the New York Times and all the major media outlets, and opened up huge new areas of genetic research. So it was quite unfortunate that he apparently decided to abandon his brain research in the face of contrary social pressure. Interestly enough, I remember an interview in which he explicitly suggested that the reason Chinese and various other East Asian peoples tended to be so much more socially-conformist than Westerners was probably that for thousands of years, the government had continually executed the more anti-social individuals, and sometimes even exterminated the entire extended families of the most severe malfactors. This had always been my own hypothesis as well.
The Galileo example is a perfectly valid one, but bear in mind that he faced the threat of arrest and possible torture by the political authorities, while at worst Lahn's scientific work would have merely generated a little opprobrium in certain small but vocal sections of the community.
Okay, here's a very simple test of your "foreign Asian" hypothesis.
ReplyDeleteI had looked at the Putnam winner lists for the years 2006-2010, which contain 26 entries, of which 15 or 60% had East Asian names.
Your claim is that something like 2/3rds of East Asian Putnam winners have generally been foreigners, i.e. that they graduated high school outside the U.S. and afterward "parachuted in" to become college Putnam winners.
Given your vast energy and the huge amount of data you've accumulated on this subject, just track down the biographical details of those 26 names, and let us know what fraction of the Asian (and non-Asian) individuals are actually "foreign" in that objective sense. A dataset of 26 isn't huge, but given your remarkable claim that 2/3rds of the Asian winners are generally foreigners, we should see some nice evidence in this particular simple instance.
I've perfectly willing to accept aspects of your hypothesis if you can actually provide some hard evidence for it.
Brief follow-up from Baron-Cohen's 2001 paper:
ReplyDelete"scientists (including mathematicians) scored significantly higher than both humanities and social sciences students, confirming an earlier study that autistic conditions are associated with scientific skills. Within the sciences, mathematicians scored highest. This was replicated in Group 4, the Mathematics Olympiad winners scoring significantly higher than the male Cambridge humanities students."
reference: http://www.ncbi.nlm.nih.gov/pubmed?term=baron-cohen%20mathematicians
I have to confess, this is the first time I've ever heard "Germanics" described as "less likely to be aggressive." I do not want to offend Godwin's Law, so I'll just say your views are somewhat novel on that one.
ReplyDeleteIt may be possible that East Asians are innately less likely to be aggressive or anti-social, but I disagree that it's "very likely," given the lack of empirical evidence I've seen. I suppose if we defined "anti-social" more explicitly, that might help. Measurement, as you point out, may make such research impossible (how do you quantify "aggression" or other personality traits? I suspect one could construct a means, but the approach would be bedeviled by subjectivity.
Your correction, or, clarification, of "fanaticism" is a good and valuable one. On the other hand, most of what I think you, with the clarification, describe as upheaval or gross violence in the face of "counter-vailing social pressures" occurs in the toxic admixture of fear, poverty, and political disorder.
I confess, I had not heard of Professor Lahn, so I looked him up quickly, and indeed, what you say fits the narrative I have found. His research in human genetics drew a quick and unequivocal rebuke, largely because of the fear of how it might be used in the wrong hands.
My reading of Professor Lahn's story (from the Wall Street Journal) actually moves me more in the thinking that he is very much like Galileo. Dr Lahn, the article said, moved away from this sort of research because of rumblings about his impending tenure and his place in the scientific community. Being black-balled in your profession is a lot more than mere "opprobrium," and for an academic, while clearly not as severe as being burnt at the stake, is tantamount to a professional execution. After all, Professor Lahn plainly did not fear upsetting the social order in protesting the corruption of the Chinese Communist Party in 1989, a move that ultimately contributed to his move to the U.S.
Interesting topic, really.
I find it funny how you have no qualms about taking foreign "paratroopers"
ReplyDeleteto inflate the number of "domestic" Asian USAMO qualifiers to draw
USAMO-IMO overrepresentation but conveniently exclude Alex Zhai, who has
competed in US HS math competitions. The fact is immigration is still a large part of Asian American culture, and you'll be surprised to find the % of East Asian US residents (or even citizens) who aren't native born.
If you're going to nit pick at the
control group, you need to define your control group first and also
exclude those of East Asian ancestry that hasn't had their full trials in Putnam. The fact is that you're just slicing & dicing the data in your favor to further your propaganda that seems forced at best. Even accepting your "unjust" slicing & dicing of the data, white and East Asian difference seems marginal (not statistically significant given relatively low # of samples). I'm even wiling to hear you out on some of the points if you actually track down all USA IMO team members instead of tweaking it to one side.
Even accepting your propaganda, it seems that East Asians are overrepresented at talented group (3 SD) and still vastly overrepresented - but less so - at the fat tails (4 SD). Since I don't think human intelligence goes far beyond Putnam Fellows, the premise "white has higher per capita representation at the fat tails than East Asians" will not be satisfied if that was your aim. It also seems that your effort is vastly helped by the presence of Ashkenazi Jews, whose number in US is almost equal to the entire East Asian American populace. Excepting Ashkenazi Jewish people, the disparity in performance between Asian Americans and white gentiles, the group some Asians actually claim IQ advantage over, will become more nuanced.
I remember Dr Baron-Cohen's paper. He's one of the world's leading researchers in so-called autism spectrum disorder. The hypotheses make intuitive sense, though as you point out, the data are pretty thin and in need of further research. This work was later featured in one of Malcolm Gladwell's popular books.
ReplyDeleteOn a personal note, my wife was horrified to discover that I sometimes dream about systems of linear equations. Well, 'horrified' is probably too strong a word for it.
That's interesting about Lahn's tenure concerns, which I hadn't realized. I'd assumed that since his work was generating so much international coverage and he was described as the head of a laboratory (or maybe a large group), he'd already gotten tenure years earlier. Certainly that issue would tend to explain his caution. I'd never bothered looking up Lahn's career and was just going by the memory of what I'd read at the time six or seven years ago.
ReplyDeleteAs I said, one difficulty of measuring personality traits is that the data tends to be far more subjective and non-quantitive than something like intelligence, and also probably more impacted by cultural influences. But I really do think there's a great deal of circumstantial evidence that East Asians are much less aggressive than Westerners, and certainly international crime/delinquency statistics can provide some hard evidence for "anti-social" tendencies.
The same sort of data shows the same patterns for Nordics and Germanics, although to a lesser extent, and this certainly accords with evidence of the last couple of centuries, including the behavioral tendencies of the very large German and Scandinavian communities in the U.S. and the German ones in Latin America. Also "individual aggressiveness" is obviously a very different thing from "national military aggressiveness," with the latter tendency often greatly enabled by the low-aggressiveness of the individual member of the society.
But even if we just focus on the latter question, it's important to carefully distinguish widely absorbed propaganda from actual historical reality. For example, during the couple of centuries prior to 1914, the most militarily aggressive European nation was most certainly France, which had fought endless wars against just about all the other European nations and had spent hundreds of years trying to dominate or even conquer the entire continent. By contrast, Germany was nearly the only leading world power which had fought not a single major war for the previous two generations, in contrast to the numerous wars fought by Britain, America, Russia, and Japan. The very widespread notion of "German militarism" is almost as much of a canard as the similarly widespread belief in traditional German/Russian hatred, absurd since those two peoples had been on the same side in most of the wars for hundreds of years. And while Britain and France had been enemies in nearly one thousand straight years of bitter warfare, Britain and Germany (i.e. Prussia) were traditionally the closest of allies.
I'm not much interested in aggressiveness or personality of individual European races nor much knowledgeable about European history, but here is my take:
ReplyDelete1. Germany has only been a major power in modern history after Napoleonic War or perhaps even Franco-Prussian War in 1870.
After the division of the Holy Roman Empire, Germanic states have been largely divided and never had power to instigate aggression on its neighbors. Having read about Napoleonic Wars, I get the impression that Prussia was quite a bit weaker than its neighbors U.K., Austria, or Russia and had lesser role in the downfall of French Empire. I think traditional English-Germany alliance hinged around their common enemy France. Germany had the land presence (army) and English had finances (after they became naval power). It would be interesting to investigate the track record of Holy Roman Empire and Austria's before those periods, but I'm afraid I'm not much knowledgeable about it.
2. Germany had major involvement in every major conflict in Europe after the Crimean War.
Austro-Prussian War, Franco-Prussian War, World War I, World War II
They certain weren't shy of displaying aggression when they had the power.
Thought-provoking response. The circumstantial (and that I think includes a whiff of the anecdotal) evidence is familiar - data were published a couple of years ago by Ron Unz comparing several cities including San Jose, California, Los Angeles, and El Paso, Tx. (I apologise if I am mis-remembering it; it's been a few years) The question raised was why was violent crime so much lower in San Jose than in, say, Seattle, and part of the explanation was the demographics - that in effect, the more Latin a city was (e.g., LA) the higher the rates of violent crime, and the more Asian it was (e.g., San Jose) the lower, even controlling for income. The study, like your point, if you can pardon a mild rebuke, grossly simplify the case by lumping "Asians" together (in fact, rates of crime are vastly different among ethnic Chinese, Japanese, Vietnamese, or Hmong people), but it certainly fits the "model minority" meme pretty well.
ReplyDeleteI don't mean to say that I necessarily disagree with your point; not at all. It's utterly plausible. I just question the statement, absent rigorous analysis, that it is "highly likely." I think it's possible, maybe even probable. I don't know. Any such conclusions are subject to, e.g., sampling and recall bias (ask any Japanese person on the streets of Tokyo who the most criminally inclined sorts of people are, and aside from black people - a side effect of too much American television - you're likely to hear complaining about Chinese immigrants, in much the same way people in California complain about Mexicans).
You're right about letting certain stereotypes about the Germans cloud historical reality; you've no doubt heard the old saw that when the legend and the truth conflict, print the legend. But Ju Hyung Ahn raises some valid points, that "Germany" is a relatively new country, and that the Prussians or other mini-kingdoms frequently were too weak or disjointed to cause the sorts of trouble that the French or the British could. The Romans surely did not consider the Germanic peoples to be too civilly-minded.
Anyhow, thanks for the discussion; it's been thought-provoking.
Yes, I'd certainly grant that there's a fuzzy and subjective line between "plausible" or "probable" and "highly likely."
ReplyDeleteIncidentally, the central and rather surprising conclusion of that Unz paper you mention was that Hispanics actually had the same general crime rates as whites of the same age, or perhaps even a bit lower. It's already quite widely known in sociological circles that Asians have crime rates much lower than either of these groups, though just as you say, there's a good deal of variation among different Asian groups just as there is among different white groups. For example, San Jose and Seattle are both significantly Asian, but the biggest single demographic difference is that Seattle is America's whitest large city, while San Jose is one-third Hispanic.
Yes, I'd say that's factually correct, but here's my own perspective:
ReplyDelete(1) The Thirty Years War of the early 1600s broke the power of Spain and Austria, which had previously been the dominant powers, and permanently reduced them to second-tier status. Instead, France became the dominant power, and over the 200+ years which followed, nearly all the major wars revolved around France's attempt to gain complete military dominance over Europe. In these wars, France usually faced a coalition of most or all of the other major powers, invariably led by Britain, and with (small but militarily efficient) Prussia usually being Britain's closest ally.
(2) Then in the 1860s, Prussia, led by Bismarck, fought a series of three short and successful wars against Denmark, Austria, and France, thereby creating a unified Germany and automatically becoming Europe's greatest power (the defeat of mighty France under Emperor Napoleon III by small Prussia was an enormous surprise to all observers). At that point, Bismarck declared that rather than behaving like an aggressive, expansionist parvenu in the French tradition, Germany would instead become a status-quo member of the Concert of Europe, and that's exactly what happened during the 40-odd years which followed, since it fought no further wars and instead focused exclusively on internal social and economic development. Meanwhile, nearly all the other major world powers fought numerous expansionist wars against each other and various other countries, during the process conquering nearly all of Africa and the rest of the world (although Germany did grab a few random bits near the end).
(3) Obviously Germany did play a very leading role in WWI and WWII, but my analysis was for the centuries prior to 1914. Also the widespread notion that Kaiser Wilhelm was some sort of crazed warmonger is an obvious legacy of WWI propaganda, and actually rather hilarious, considering that he didn't fight a single war during the first twenty-five years of his reign, while virtually all of his contemporary world leaders did. A fair analogy might be that the leaders of today's America routinely accuse the leaders of today's China as being "warmongers"...
2006-2010 top 5 results per Ron's request:
ReplyDeleteTiankai Liu (Harvard) 2006 [also 2005]
[China] Hansheng Diao (MIT) 2006
Po-Ru Loh (Caltech) 2006
[Canada] Yufei Zhao (MIT) 2006, 2008, 2009
Jason C. Bland (Caltech) 2007
Brian R. Lawrence (Caltech) 2007, 2008, 2010
[China] Qingchun Ren (MIT) 2007, 2009
[China] Xuancheng Shao (MIT) 2007
Arnav Tripathy (Harvard) 2007, 2008, 2009
[Korea] Seok Hyeong Lee (Stanford) 2008, 2010
Bohua Zhan (MIT) 2008
William Johnson (U of Washington) 2009
[China] Xiaosheng Mu (Yale) 2009
[China] Yu Deng (MIT) 2010
Colin P. Sandon (MIT) 2010
Alex (Lin) Zhai (Harvard) 2010
group counts
US E.Asians: 4 individuals, 4 qualifications [all 4 from US IMO]
foreign E.A: 6 individuals, 11 qualifications [all had IMO or IPhO gold]
US whites : 4 individuals, 6 qualifications [2 from US IMO, 2 non-IMO]
Statistical observations:
- the foreign students outperform everybody (more qualification per individual) because they were selected from the highest IMO scorers from all over the world. The US population includes silver medalists and non-IMO students, but nothing less than at least one gold is tolerated by the Yale/MIT/Stanford/etc admissions office when filling out their Putnam teams. The sports recruitment model has been carried over to math competitions.
-when reading a list of high scoring students, it is the relative number of qualifications, not individuals, that determines the proportion of foreign Asians you are likely to be seeing. 2/3 was clearly a good estimate (we observe 65% in Ron's sample), and as I enumerated earlier for the 2009 also true for the full annual list of the top 70-80 students in recent years.
- US whites in this group had lower HS performance but higher college performance than the US East Asians. Whether you describe patterns of that type, for much larger data sets, as white overperformance or US E.Asian underperformance seems like a distinction without a difference. If you are the admissions office trying to maximize the number of very high Putnam scores at your school, and have a limited number of scholarships to offer, these group differences have game theoretic implications.
2005
"W 14:26 (53.8%)
ReplyDeleteE.A 7:14 (50.0%)
I 2:1 (66.7%)
and of course the I > W > EA pattern reasserts itself"
This is why RKU accused you of slicing and dicing the data. Imagine that hypothetically Indian Americans outnumbered East Asian Americans by 20 in the most recent cohorts. Imagine that over the past 10 years we had 100 East Asian American IMO competitors and only 2 Indian American ones. [These numbers obviously don't make sense, since there are only 6 IMO competitors each year. I'm merely using them to illustrate a hypothetical point]. Now, you might point out that on a per capita basis, East Asian Americans were over-represented by a factor of 1000 relative to East Asians on the IMO.
Now, to continue with our hypothetical example. Imagine that all 2 of the Indian IMO competitors became Putnam Fellows, but only 70 out of the 100 East Asians did. So East Asian Americans outnumber Indian Americans 35:1 amongst Putnam Fellows, leading to a per capita over-representation of 700x for East Asian Americans relative to Indian Americans.
Of course, what you would point is that hey, out of all of the Indians who made IMO, a larger percentage of them became Putnam Fellows relative to East Asian Americans. Ergo Indian Americans clearly over-perform relative to East Asian Americans despite being under-represented by a factor of 700x
You have an annoying habit of analyzing data in meaningless ways. Hence RKU accused you of dishonest propaganda.
If you're gonna nit pick at the table in question, you need to give the same benefit of doubt to every members in the above Putnam table, or at least to everyone who is "no." Unless you track down every IMO members and check # of trials each had in Putnam, it is not reasonable to fabricate the data in this way. This kind of slicing & dicing is what dishonest statisticians do. Even accepting your unjust criteria, it seems that white to East Asian disparity is marginal and statistically insignificant given the relatively small number of samples.
ReplyDeleteYour statement regarding East Asian overprediction of USAMO-IMO seems to be also false. It is a fairly recent event that East Asians comprise of 50-60% of USAMO pool (about the last 3-4 years perhaps) The percent representation was much less in 1990s and somewhat less in early 2000s. The recent East Asian representation among US IMO contestants seems to support this.
*East Asian representation among recent USA IMO contestants
2011 4 out of 6
2010 4 out of 6
2009 3 out of 6
2008 1 out of 6
2007 3 out of 6
2006 3 out of 6
50% over the last 6 years and higher representation than white population.
USAMO Qualfiers in recent years: http://ww.amc8.org/e-exams/e8-usamo/archiveusamo.shtml
I've come to think that your perception of under-performance by East Asians comes from the mistake (extrapolating unfairly) of directly comparing JHS:MATHCOUNTS-HS:USAMO&IMO-U:Putnam results all within the same calendar year. There should be about 4 years of lag effect between each successive steps. This seems very likely with the steady growth of East Asian representation (e.g. USAMO qualifiers, IMO contestants) in American Mathematics competitions recent years.
Another thing to note is that even accepting your view of overprediction-underprediction, which happens in the order of East Asians, whites, Indians, the current college admissions process still remains unfair, since Indian Americans face the similar discrimination East Asian Americans face.
If you're gonna nit pick at the table in question, you need to give the
ReplyDeletesame benefit of doubt to every members in the above Putnam table, or at
least to everyone who is "no." Unless you track down every IMO members
and check # of trials each had in Putnam, it is not reasonable to
fabricate the data in this way. This kind of slicing & dicing is
what dishonest statisticians do. Even accepting your unjust criteria,
it seems that white to East Asian disparity is marginal and
statistically insignificant given the relatively small number of
samples.
Your statement regarding East Asian overprediction of USAMO-IMO seems to
be also false. It is a fairly recent event that East Asians comprise
of 50-60% of USAMO pool (about the last 3-4 years perhaps) The percent
representation was much less in 1990s and somewhat less in early 2000s.
The recent East Asian representation among US IMO contestants seems to
support this.
*East Asian representation among recent USA IMO contestants
2011 4 out of 6
2010 4 out of 6
2009 3 out of 6
2008 1 out of 6
2007 3 out of 6
2006 3 out of 6
50% over the last 6 years and higher representation than white population.
USAMO Qualfiers in recent years: http://ww.amc8.org/e-exams/e8-usamo/archiveusamo.shtml
I've come to think that your perception of under-performance by East
Asians comes from the mistake (extrapolating unfairly) of directly
comparing JHS:MATHCOUNTS-HS:USAMO&IMO-U:Putnam results all within
the same calendar year. There should be about 4 years of lag effect between each successive steps.
This seems very likely with the steady growth of East Asian
representation (e.g. USAMO qualifiers, IMO contestants) in American
Mathematics competitions recent years.
Another thing to note is that even accepting your view of
overprediction-underprediction, which happens in the order of East
Asians, whites, Indians, the current college admissions process still
remains unfair, since Indian Americans face the similar discrimination
East Asian Americans face.
If you're gonna nit pick at the table in question, you need to give the
ReplyDeletesame benefit of doubt to every members in the above Putnam table, or at
least to everyone who is "no." Unless you track down every IMO members
and check # of trials each had in Putnam, it is not reasonable to
fabricate the data in this way. This kind of slicing & dicing is
what dishonest statisticians do. Even accepting your unjust criteria,
it seems that white to East Asian disparity is marginal and
statistically insignificant given the relatively small number of
samples.
Your statement regarding East Asian overprediction of USAMO-IMO seems to
be also false. It is a fairly recent event that East Asians comprise
of 50-60% of USAMO pool (about the last 3-4 years perhaps) The percent
representation was much less in 1990s and somewhat less in early 2000s.
The recent East Asian representation among US IMO contestants seems to
support this.
*East Asian representation among recent USA IMO contestants
2011 4 out of 6
2010 4 out of 6
2009 3 out of 6
2008 1 out of 6
2007 3 out of 6
2006 3 out of 6
50% over the last 6 years and higher representation than white population.
USAMO Qualfiers in recent years: http://ww.amc8.org/e-exams/e8-usamo/archiveusamo.shtml
I've come to think that your perception of under-performance by East
Asians comes from the mistake (extrapolating unfairly) of directly
comparing JHS:MATHCOUNTS-HS:USAMO&IMO-U:Putnam results all within
the same calendar year. There should be about 4 years of lag effect between each successive steps.
This seems very likely with the steady growth of East Asian
representation (e.g. USAMO qualifiers, IMO contestants) in American
Mathematics competitions recent years.
Another thing to note is that even accepting your view of
overprediction-underprediction, which happens in the order of East
Asians, whites, Indians, the current college admissions process still
remains unfair, since Indian Americans face the similar discrimination
East Asian Americans face.
If you're going to nitpick at the table in question, you
ReplyDeleteneed to give the same benefit of doubt to every members in the above Putnam
table, or at least to everyone who is "no." Unless you track
down every IMO members and check # of trials each had in Putnam, it is not
reasonable to fabricate the data in this way. This kind of slicing &
dicing is what dishonest statisticians do. Even accepting your unjust
criteria, it seems that white to East Asian disparity is marginal and
statistically insignificant given the relatively small number of samples.
Your statement regarding East Asian over prediction of USAMO-IMO seems to be
also false. It is a fairly recent event that East Asians comprise of
50-60% of USAMO pool (about the last 3-4 years perhaps) The percent
representation was much less in 1990s and somewhat less in early 2000s.
The recent East Asian representation among US IMO contestants seems to support
this.
*East Asian representation among recent USA IMO contestants
2011 4 out of 6
2010 4 out of 6
2009 3 out of 6
2008 1 out of 6
2007 3 out of 6
2006 3 out of 6
50% over the last 6 years and higher representation than white population.
USAMO Qualifiers in recent years:
http://ww.amc8.org/e-exams/e8-usamo/archiveusamo.shtml
I've come to think that your perception of under-performance by East Asians
comes from the mistake (extrapolating unfairly) of directly comparing
JHS:MATHCOUNTS-HS:USAMO&IMO-U:Putnam results all within the same calendar
year. There should be about 4 years of lag effect between each successive
step. This seems very likely with the steady growth of East Asian
representation (e.g. USAMO qualifiers, IMO contestants) in American Mathematics
competitions recent years.
Another thing to note is that even accepting your view of
overprediction-underprediction, which happens in the order of East Asians,
whites, Indians; the current college admissions process still remains unfair,
since Indian Americans face the similar discrimination East Asian Americans
face.
If you're going to nitpick at the table in question, you
ReplyDeleteneed to give the same benefit of doubt to every members in the above Putnam
table, or at least to everyone who is "no." Unless you track
down every IMO members and check # of trials each had in Putnam, it is not
reasonable to fabricate the data in this way. This kind of slicing &
dicing is what dishonest statisticians do. Even accepting your unjust
criteria, it seems that white to East Asian disparity is marginal and
statistically insignificant given the relatively small number of samples.
Your statement regarding East Asian over prediction of USAMO-IMO seems to be
also false. It is a fairly recent event that East Asians comprise of
50-60% of USAMO pool (about the last 3-4 years perhaps). The percent
representation was much less in 1990s and somewhat less in early 2000s.
The recent East Asian representation among US IMO contestants seems to support
this.
*East Asian representation among recent USA IMO contestants
2011 4 out of 6
2010 4 out of 6
2009 3 out of 6
2008 1 out of 6
2007 3 out of 6
2006 3 out of 6
50% over the last 6 years and higher representation than white population.
USAMO Qualifiers in recent years:
http://ww.amc8.org/e-exams/e8-usamo/archiveusamo.shtml
I've come to think that your perception of under-performance by East Asians
comes from the mistake (extrapolating unfairly) of directly comparing
JHS:MATHCOUNTS-HS:USAMO&IMO-U:Putnam results all within the same calendar
year. There should be about 4 years of lag effect between each successive
step. This seems very likely with the steady growth of East Asian
representation (e.g. USAMO qualifiers, IMO contestants) in American Mathematics
competitions recent years.
Another thing to note is that even accepting your view of
overprediction-underprediction, which happens in the order of East Asians,
whites, Indians; the current college admissions process still remains unfair,
since Indian Americans face the similar discrimination East Asian Americans
face.
If you're going to nitpick at the table in question, you
ReplyDeleteneed to give the same benefit of doubt to every members in the above Putnam
table, or at least to everyone who is "no." Unless you track
down every IMO members and check # of trials each had in Putnam, it is not
reasonable to fabricate the data in this way. This kind of slicing &
dicing is what dishonest statisticians do. Even accepting your unjust
criteria, it seems that white to East Asian disparity is marginal and
statistically insignificant given the relatively small number of samples.
Your statement regarding East Asian over prediction of USAMO-IMO seems to be
also false. It is a fairly recent event that East Asians comprise of
50-60% of USAMO pool (about the last 3-4 years perhaps). The percent
representation was much less in 1990s and somewhat less in early 2000s.
The recent East Asian representation among US IMO contestants seems to support
this.
*East Asian representation among recent USA IMO contestants
2011 4 out of 6
2010 4 out of 6
2009 3 out of 6
2008 1 out of 6
2007 3 out of 6
2006 3 out of 6
50% over the last 6 years and higher representation than white population.
USAMO Qualifiers in recent years:
http://ww.amc8.org/e-exams/e8-usamo/archiveusamo.shtml
I've come to think that your perception of under-performance by East Asians
comes from the mistake (extrapolating unfairly) of directly comparing
JHS:MATHCOUNTS-HS:USAMO&IMO-U:Putnam results all within the same calendar
year. There should be about 4 years of lag effect between each successive
step. This seems very likely with the steady growth of East Asian
representation (e.g. USAMO qualifiers, IMO contestants) in American Mathematics
competitions recent years.
Another thing to note is that even accepting your view of
overprediction-underprediction, which happens in the order of East Asians,
whites, Indians; the current college admissions process still remains unfair,
since Indian Americans face the similar discrimination East Asian Americans
face.
2010 Putnam results of former USAMO qualifiers:
ReplyDeleteTop5 (Putnam Fellows ; US top 3)
BRIAN R. LAWRENCE California Institute of Technology
COLIN P. SANDON Massachusetts Institute of Technology
ALEX (LIN) ZHAI Harvard University
The Next Nine Highest Ranking Individuals (US rank 4-8)
VLAD FIROIU Massachusetts Institute of Technology
ERIC K. LARSON Harvard University
JOHN V. PARDON Princeton University
JACOB N. STEINHARDT Massachusetts Institute of Technology
ARNAV TRIPATHY Harvard University
The Next Ten Highest Ranking Individuals (US rank 9-14)
YAKOV I. BERCHENKO-KOGAN California Institute of Technology
SERGEI S. BERNSTEIN Massachusetts Institute of Technology
JASON C. BLAND California Institute of Technology
ZARATHUSTRA E. BRADY California Institute of Technology
WILLIAM A. JOHNSON University of Washington, Seattle
EVAN M. O'DORNEY [grade 12 ; UC Berkeley]
(Do you notice anything "interesting" about the results so far? Let's continue down the list.)
Honorable Mention (who were USAMO qualifiers - 35 students):
ROHIT AGRAWAL [grade 12 ; U Minnesota, Twin Cities]
ALEXANDER G. ANDERSON, Washington University, St. Louis
JOHN D. BERMAN, Massachusetts Institute of Technology
TIMOTHY J. BLACK, California Institute of Technology
PAUL F. CHRISTIANO, Massachusetts Institute of Technology
MILES D. EDWARDS, Indiana University, Bloomington
SAM S. ELDER, California Institute of Technology
IAN A. FRANKEL, Princeton University
DAVID D. GEE, University of California, Berkeley
HIRAM B. GOLZE, Brigham Young University
SHERRY GONG [f], Harvard University
BRIAN C. HAMRICK, Massachusetts Institute of Technology
TRAVIS J. HANCE, Massachusetts Institute of Technology
MARCELLO M. HERRESHOFF, Stanford University
ADAM C. HESTERBERG, Princeton University
SAM G. KELLER, Stanford University
ERICK P. KNIGHT, Princeton University
JUSTIN KOPINSKY, University of Illinois, Urbana-Champaign
MIKHAIL P. LAVROV, Duke University
AVI W. LEVY, University of Texas, Dallas
PAUL D. LEWIS, University of Michigan, Ann Arbor
DANIEL LI, Massachusetts Institute of Technology
SHIYU LI [f], University of California, Berkeley
JEFFREY A. MANNING, California Institute of Technology
PALMER C. MEBANE, Harvey Mudd College
OFIR NACHUM, Massachusetts Institute of Technology
DANIEL E. POORE, Pomona College
DAVID S. ROLNICK, Massachusetts Institute of Technology
DAVID B. RUSH, Massachusetts Institute of Technology
KRISHANU R. SANKAR, Massachusetts Institute of Technology
NATHANIEL B. SHAR, Stanford University
MINSEON SHIN, Massachusetts Institute of Technology
ANTHONY Y. WANG, Massachusetts Institute of Technology
GEORGE L. XING, Massachusetts Institute of Technology
ALEX W. ZORN, University of Chicago
Demographics of Putnam 2010 for former USAMO qualifiers:
Winners : 1 E Asian 2 white
Top 8 : 1 E.Asian 6 white 1 Indian
Top 14 : 1 E.Asian 12 white (incl. at least 3.5 Jewish, 1 autistic) 1 Indian
Top 49 : 8 E. Asian 38 white (incl. approx. 10 Jewish + 1 autistic) 3 Indian
Maybe this was an unlucky year for E.Asians in the top 14, but being outnumbered 4-to-1 by whites in the top 50
after a huge E.Asian dominance in USAMO qualifiers from 2006-2010 is quite a reversal.
All just "propaganda", right?
可笑的是这个文章本是关于一个白人和Einstein。但是尽然也变成关于黄种人成就和智力的讨论。白鬼子的日子看样子不好过了。这一点本身证明了他们的心虚。黄种人的优秀的确使它们心虚。
ReplyDeleteIf you're going to nitpick at the
ReplyDeletetable in question, you need to give the same benefit of doubt to every members
in the above Putnam table, or at least to everyone who is "no."
Unless you track down every IMO members and check # of trials each had in Putnam,
it is not reasonable to fabricate the data in this way. This kind of
slicing & dicing is what dishonest statisticians do. Even accepting
your unjust criteria, it seems that white to East Asian disparity is marginal
and statistically insignificant given the relatively small number of samples.
Your statement regarding East Asian over prediction of USAMO-IMO seems to be
also false. It is a fairly recent event that East Asians comprise of
50-60% of USAMO pool (about the last 3-4 years perhaps). The percent
representation was much less in 1990s and somewhat less in early 2000s.
The recent East Asian representation among US IMO contestants seems to support
this.
*East Asian representation among recent USA IMO contestants
2011 4 out of 6
2010 4 out of 6
2009 3 out of 6
2008 1 out of 6
2007 3 out of 6
2006 3 out of 6
50% over the last 6 years and higher representation than white population.
USAMO Qualifiers in recent years:
http://ww.amc8.org/e-exams/e8-usamo/archiveusamo.shtml
I've come to think that your perception of under-performance by East Asians
comes from the mistake (extrapolating unfairly) of directly comparing
JHS:MATHCOUNTS-HS:USAMO&IMO-U:Putnam results all within the same calendar
year. There should be about 4 years of lag effect between each successive
step. This seems very likely with the steady growth of East Asian
representation (e.g. USAMO qualifiers, IMO contestants) in American Mathematics
competitions recent years.
Another thing to note is that even accepting your view of overprediction-underprediction,
which happens in the order of East Asians, whites, Indians; the current college
admissions process still remains unfair, since Indian Americans face the
similar discrimination East Asian Americans face.
Conclusion: From the apparent lack
of trend of under-performance among East Asians between USAMO and IMO and
between IMO and Putnam given above, it seems your speculation is generally
false. By dissecting, slicing & dicing, and even disintegrating given
data, you'll generally stumble upon some patterns accidentally going in your
direction, since you're also employing your wildest imaginations here.
Subjecting certain data points to different standards than others to tweak the
data is a fine case in point. The possibilities of claims one can make by
employing such devious ploy is endless.
Ohhhh, and out comes Hitler! I was waiting for that. I guess they bring out the heaviest artillery when actual numbers and math prove them wrong, lol.
ReplyDeleteWhy are you leaving out Yu Deng and Seok Hyeong Lee?They are East Asian,right?You seem to make distinctions
ReplyDeletewhen the East Asians just came to the US and no distinction with the Whites.What makes you think none of the whites is international?
It is pretty obvious that 3 of the 5 putnam fellows are East Asian.
ReplyDeleteI'm not sure I understand the argument here. You are comparing Asian and white USAMO contestants, and attempting to show that Asians do relatively less well at the Putnam, but you don't tell us how Asian USAMO participants do. Are they showing up and somehow doing worse than their USAMO dominance would predict? If they are, I still don't really understand the argument. So the East Asian individuals do worse on the Putnam relatively than they did on the USAMO, but how does this contribute to the East Asian disappearance thesis? Your argument seems to be saying "look, East Asian USAMO contestants did relatively worse on the Putnam vs. white contestants", but what are you alleging that to mean?
ReplyDeleteSorry for being dense, but it is Eggnog season.
Another fun data point:2011 IMO Gold Medallists,US Team
ReplyDelete1.David Yang .
2.Benjamin Gunby
3.Xiaoyu He
4.Mitchell Lee
5.Wenyu Cao
6.Evan O'Dorney
If you're
ReplyDeletegoing to nitpick at the table in question, you need to give the same benefit of
doubt to every members in the above Putnam table, or at least to everyone who
is "no." Unless you track down every IMO members and check # of
trials each had in Putnam, it is not reasonable to fabricate the data in this
way. This kind of slicing & dicing is what dishonest statisticians do. Even
accepting your unjust criteria, it seems that white to East Asian disparity is
marginal and statistically insignificant given the relatively small number of
samples.
Your statement regarding East Asian over prediction of USAMO-IMO seems to be
also false. It is a fairly recent event that East Asians comprise of 50-60% of
USAMO pool (about the last 3-4 years perhaps). The percent representation was
much less in 1990s and somewhat less in early 2000s. The recent East Asian
representation among US IMO contestants seems to support this.
*East Asian representation among recent USA IMO contestants
2011 4 out of 6
2010 4 out of 6
2009 3 out of 6
2008 1 out of 6
2007 3 out of 6
2006 3 out of 6
50% over the last 6 years and higher representation than white population.
USAMO Qualifiers in recent years: http://ww.amc8.org/e-exams/e8-...
I've come to think that your perception of under-performance by East Asians
comes from the mistake (extrapolating unfairly) of directly comparing
JHS:MATHCOUNTS-HS:USAMO&IMO-U:Putnam results all within the same calendar
year. There should be about 4 years of lag effect between each successive step.
This seems very likely with the steady growth of East Asian representation
(e.g. USAMO qualifiers, IMO contestants) in American Mathematics competitions
recent years.
Another thing to note is that even accepting your view of
overprediction-underprediction, which happens in the order of East Asians,
whites, Indians; the current college admissions process still remains unfair,
since Indian Americans face the similar discrimination East Asian Americans
face.
From the apparent lack of trend of under-performance among East Asians between
USAMO and IMO and between IMO and Putnam given above, it seems your speculation
is generally false. By dissecting, slicing & dicing, and even
disintegrating given data, you'll generally stumble upon some patterns
accidentally going in your direction, since you're also employing your wildest
imaginations here. Subjecting certain data points to different standards than
others to tweak the data is a fine case in point. The possibilities of claims
one can make by employing such devious ploy is endless.
Since you continue to limit your data to short time periods
ReplyDeleteto further your propaganda, I’ll pretend to play your game and come up with
equally outlandish claim of my own.
East Asian Representation in each category
2011 USA IMO Team Members: 4 out of 6 (66.7%)
2011 USAMO Winners: 8 out of 12 (66.7%)
2011 USAMO Honorable Mention: 10 out of 13 (76.9%)
2011 USAMO Top 25: 18 out of 25 (72%)
2010 USA IMO Team Members: 4 out of 6 (66.7%)
2010 USAMO Winners: 7 out of 12 (58.3%)
2010 USAMO Honorable Mention: 12 out of 16 (75.0%)
2010 USAMO Top 28: 19 out of 28 (67.9%)
East Asians are over performing (despite being vastly
overrepresented per capita basis in USAMO pool) since they’re not quite 2/3 of
USAMO pool
Whites must be underperforming to some degree in return
Now, if I was as idiotic and obnoxious as you are, I’d say
this is a perfectly justifiable evidence to impose higher USAMO index cut on
white population. This is just a glimpse
of what dishonest propagandizing can amount to.
From what I can tell, your aim in pushing for such a
preposterous propaganda must be either
1. Given the nonexistent trend of talent-thinning effect
East Asians experience, there are more white talent at the fat tails (SD 5+?)
2. Universities are justified in discriminating Asian Americans
in college admissions process due to the aforementioned nonexistent trend.
or a combination of both.
Besides the obvious reason that East Asian talent thinning
effect seems nonexistent,
- 1 is unlikely because the human intelligence doesn’t go much
further than Putnam Fellows, where East Asian Americans are still vastly
overrepresented per capita basis. There
isn’t enough room left for the discrepancy to converge due to the nonexistent
trend
-2 is still unjustified because Indian Americans are still
discriminated in similar way. According
to you test results tend to under predict Indian American performance.
Clearly, you will fail in asserting either of them.
Since creating “optical illusions” seems to be your forte, I
suspect your best bet is to try law or politics where your goal is easier since
you’re only trying to fool the average population. Perhaps you are a PR consultant as RKU
suspects; in this case, you’d feel right at home.
Since
ReplyDeleteSince you continue to limit your data to short time periods
ReplyDeleteto further your propaganda, I’ll pretend to play your game and come up with
equally outlandish claim of my own.
East Asian Representation in each category
2011 USA IMO Team Members: 4 out of 6 (66.7%)
2011 USAMO Winners: 8 out of 12 (66.7%)
2011 USAMO Honorable Mention: 10 out of 13 (76.9%)
2011 USAMO Top 25: 18 out of 25 (72%)
2010 USA IMO Team Members: 4 out of 6 (66.7%)
2010 USAMO Winners: 7 out of 12 (58.3%)
2010 USAMO Honorable Mention: 12 out of 16 (75.0%)
2010 USAMO Top 28: 19 out of 28 (67.9%)
East Asians are over performing (despite being vastly
overrepresented per capita basis in USAMO pool) since they’re not quite 2/3 of
USAMO pool
Whites must be underperforming to some degree in return
Now, if I was as idiotic and obnoxious as you are, I’d say
this is a perfectly justifiable evidence to impose higher USAMO index cut on
white population. This is just a glimpse
of what dishonest propagandizing can amount to.
A large number of the whites (and non-E.Asians) who win the contest are international, as I explained many times. Most of the foreign winners of the contest were not E.Asian. In the 1990's they were from the Russian and European ex-communist countries, and in the 2000's from the whole world, apparently including recruitment pipelines from Korea to Stanford, China to MIT, and other such games. UCLA has an IMO scholarship advertised on Terry Tao's blog, and I think Steve said something about Caltech using a scholarship, that he endowed, to enroll some IMO/IPhO competitors from Romania(?) and other places. Like it or not, "athletic recruitment" for university math competitions has been a reality for many years.
ReplyDeleteAs an example, in making the 2010 US Putnam list, for the second tier of winners, ranks 6 to 14, I excluded from the US list 3 gold medalists from the Moldova,China,and Korea IMO teams, and another Chinese student from the Renmin HS; and for the third tier, ranks 15-24, I excluded 4 IMO superstars from Russia, Korea, China and Moldova. The Honorable Mention list had IMO medalists from Iran, Bulgaria, Albania, Hong Kong, Canada, and Taiwan. For the top 5
To answer your other question: the underperformance discussion is not about race or country, but race *within* one country. This is in part because it stemmed from the college admissions discussion which pertains to whites and Asian (especially, Chinese and Korean generation 1, 1.5, and 2 immigrant) Americans. Speaking of which, the "admissions discrimination" theory would predict that overqualified Asian-Americans would pile up at Berkeley, Caltech and lower tier schools because they are locked out of the Ivy League. What we in the Putnam list is a bit different:
- all of the 7 high-scoring students at Caltech are white
- more than 20% of the white high scores are from lower-ranked schools: U Washington, UIUC, UT Dallas, U Indiana, BYU, Michigan, Harvey Mudd, Pomona, Duke and Chicago. Either highly qualified whites don't care as much about school prestige as Asians (see next line), or being a math wunderkind does not guarantee admission to top programs for whites.
- all of the US Asian (EA and Indian) students on the list were at Harvard and MIT, except for two at Berkeley. Of those two, one (from Sunnyvale, CA) was the #1 or #2 female math competition winner in the USA in her graduation year, which is a ticket to attend whatever college one wants. It is likely that Berkeley was her or her family's preference and not a sign of exclusion by higher-ranked schools. Which leaves 1 of 11 Asians attending a higher ranked school than 1 in 7 of the whites on the list.
>>"W 14:26 (53.8%) E.A 7:14 (50.0%) I 2:1 (66.7%) ... the I > W > EA pattern reasserts itself"
ReplyDelete>
>This
is why RKU accused you of slicing and dicing the data.
Does he make that accusation now, or are you making it up for him? Ron piped down once detailed lists and counts were posted, and has silently but obviously withdrawn whatever concrete objections he had raised. He never explained what he meant by improper "slicing and dicing", but I'm sure he is elated at the spectacle of internet trolls appropriating his name and comments for their own purposes.
> You have an annoying habit of analyzing data in meaningless ways.
>Hence RKU accused you of dishonest propaganda.
Is he still accusing, i.e., does he know of any statement I posted for which there is a disclosable reason to suppose it is dishonest, such as his earlier (and now totally discredited) suspicion that I was imagining the foreign student numbers in the Putnam? Note: disliking the implications is not a reason. I'll point out the obvious again, which is that several people who know this stuff in an up close and personal sort of way, read this blog and post in the comments, and none of them have found anything wrong in the numerous factual assertions I have made about the competitions, their recent history, training patterns, Putnam demographics, etc.
I don't think RKU ever called my analyses meaningless, either; that is you borrowing his identity again.
Nor have you explained what is meaningless in (for example) the analysis of Putnam attainment levels vs college admission rates using the most generous Asian supremacist assumptions about "shifted normal distributions". You had been touting the supposed 12x overrepresentation in the Putnam as a guide to what college admission should look like and I worked out the arithmetic for you. Will you be commenting on the specific of that analysis, because it sure seems to imply that the status quo is an over-representation of Asians relative to ability levels.
> Of
course, you have a persistent habit of analyzing data along these
lines,
> by trying to argue that out of all of the members of group X who
qualified for USAMO,
> a lower percentage of them qualified for IMO than
out of all of the members of group Y
> who qualified for USAMO.
If the data supports a multitude of such X-Y disparities, large and small, on big data sets and tiny ones, and there is no credible example of the opposite pattern, and the X-Y patterns have a simple mathematical underpinning consistent with abundant external evidence (e.g., the million and one pieces of evidence that E. Asians tend to "work harder" at a given level of ability, start earlier on math competitions, etc) -- then what exactly do you claim is wrong with using the X-Y comparisons as evidence? It is not the particular result of this selection or that year that is important on its own, even if it does always point in the same direction. It is the universality of the pattern and the total absence of falsification (sans data engineering) or statistical/mathematical arguments as to why the data could be misinterpreted, that make the case credible.
Excellent. Assuming this data is correct, this indicates that 11/15 of the East Asian names on Putnam winner lists from 2006-2011 came from foreign students who had not attended high school in the U.S., affirming your 2/3rds claim.
ReplyDeletePresumably, this very much helps to explain the remarkable 60% dominance of East Asians among Putnam winners in recent years, which had so totally astonished me when I checked the lists.
Looking at things another way, East Asians are probably less than 5% of American H.S. students while whites are around 50%. Therefore, absent foreign students, a 60% East Asian percentage would imply that they outperformed whites (including Jews) by a factor of around 20x at the ultra-high-end of mathematics, which seems rather implausible, even given zealous training. On the other hand, if we net out those foreign East Asians, the EA/white winner ratio drops to 4/6, corresponding to an East Asian outperformance factor of 6x relative to whites, which seems much more reasonable. (Obviously these datasets are absurdly small for an valid statistical inferences.) After all, East Asians are certainly a high-ability group in American society, but they're not the *only* high ability group.
Also, in glancing over the Putnam lists, there seems to have been a recent disappearance of the Eastern European names, which had been common on the lists for a decade or so following the fall of Communism, and this makes sense since the influx of those top-ability immigrant families was probably a one-time occurance. Also, Jewish names, once quite common, seem to have recently disappeared, perhaps because American Jews have gotten much lazier and less math/science oriented. So these two factors, plus the strong influx of foreign students from Asian backgrounds, probably helps explain why East Asians have suddenly jumped from their traditional 10-20% presence to the currently astonishing 60%.
Jews, based on names, appear to be about 20% of the top 50 US students in the most recent Putnam list (2010), including 3 of the top 14. The Mertz/Gallian/Kane study of ethnicity of math contest winners listed an additional student in the top 3 as having one Jewish parent.
ReplyDeleteFor analysis of competition data I think it is much better to stick to name-based ethnicity rather than trying to dig beneath the data to determine genetic ethnicity. The underperformance phenomenon is cultural, so that a white adopted into an East Asian family would be just as likely as an Asian of the same ability in the same family, to overperform his natural talents and thus underperform his credentials at difficult more selections. If having an East Asian last name is a negative predictor of performance, that is a slightly different statement than the same one about genetic Asians, but it still meaningful as a statistical and predictive phenomenon, the name identification is relatively consistent across samples, and it does not involve the contentious "who is an X" sort of analyses for mixed ancestry.
I do agree that Yan Shen was wrong to play the Hitler Card. It was at minimum excessive.
ReplyDeleteSo it appears that all 6 US IMO members were gold medalists in 2011. I notice that 4/6 of the US gold medalists were of East Asian extraction.
ReplyDeleteNow the percentage of USAMO competitors in 2011 who were East Asian was around 55%+.
55% to 66 2/3%. Seems like an increase in East Asian representation the further along the pipeline we go. ;)
Eh, that just seems to be a mathematical truth.
ReplyDeleteGiven one randomly posts Encyclopaedia Britannica entries over and over (allowing for repetition), the probability that you'll find the desired entry grows (approaches 1) as the number of entries grow.
I'm not sure we can really talk about plural "fascinating discoveries" from Lahn that were publicised. Really the ASPM and microcephalin stories - were guessed that ASPM and microcephalin are related to increases in brain size in Eurasian populations - I think, were really the only things that I can remember that came from him aside from speculation that were in the public eye. And those ideas were sort of shown to be false.
ReplyDeleteI'm sure he's a very intelligent man and very productive in his field, but I don't know if he is a good case study of science being backed away from.
And really it would've been surprising if those hypotheses were true. Brain size has been decreasing in all regions during the period he speculated about, even net of expectation based on body size (current Europeans have smaller brains than earlier Europeans, current Asians have smaller brains than earlier Asians, &c.).
His backing away from the field might have been more due to the fact that there was a perception that "Bruce is in a hurry to be famous" (as a colleague of his said) and realisation that the result had publicised in a way that suggested that he was seeking fame more than doing proper science.
In context, the content regarding Lahn's tenure in the article is "The accuracy of Dr. Lahn's work and his views on race came up in his tenure review last fall, says a person familiar with it. After debate, his department voted unanimously in his favor, according to another faculty member. A more senior committee agreed and awarded Dr. Lahn the post of full professor, although it wasn't unanimous, this person says."
...
Northern Europeans being less anti-social in the sense of committing crimes seems bizarre in the face of the fact that homicide and property crimes either all increase in the north of the continent and within German regions relative to the south or are identical (the region with the most elevated murder rate is the Balto-Finnic/Russian area - the Finns themselves don't get up to Taiwanese and South Korean levels of murder, but the Latvians, Estonians and Lithuanians exceed them by some margin).
Actually, you still get around 20x representation.
ReplyDeleteNon Hispanic-white 64.7% (2010)
Jewish American (over 80% Ashkenazi) 1.7-2.2%
Chinese (ancestry-including Taiwan) American 1.2%
Korean American .5%
Japanese American .3%
East Asian American 2.0%
Given the notorious reputation of East Asian households tending to have only 1 or 2 kids, East Asian over representation in grade school population is probably minimal, even accounting for some foreign students studying in US.
[Regarding US demographics: from Wikipedia] "In either case, such growth is unlike most European countries, especially Germany, Russia, and Greece, or Asian countries such as Japan or South Korea, whose populations are slowly declining, and whose fertility rates are below replacement."
Many Chinese Americans I've known also tended to be an only child due to one-child policy in mainland China.
Allowing to count Brian Lawrence 3 times,
(64.7/2)X(2/3)=21.57
Just head count,
64.7/2=32.35
I think East Asian American group probably got some boost (less than Indian, SE Asians) from selective immigration, but the # still seems fairly high.
Regardless it was a pointless and truly offensive ad hominem from Yan Shen. And Yan Shen, you really should be the literally last person in the world to accuse someone of being obsessed with this subject.
ReplyDeleteActually, you still get about X20 representation.
ReplyDeleteNon-Hispanic White: 64.7% (2010)
Jewish Americans (over 80% Ashkenazi): 1.7-2.2%
Chinese (ancestry-including Taiwan) Americans: 1.2%
Korean Americans: .5%
Japanese Americans: .3%
East Asian Americans: 2.0%
Since East Asian American households are notorious for
having 1 or 2 kids, I don’t think their representation in grade school deviates
much from above even accounting for the foreign students studying in US.
Many Chinese Americans I’ve known were an only child due to one
child policy in mainland China.
[Regarding US Demographics: From Wikipedia] “In either case,
such growth is unlike most European countries, especially Germany, Russia, and Greece, or Asian countries such as Japan or South Korea,
whose populations are slowly declining, and whose
fertility rates are below replacement.”
Counting Brian Lawrence 3 times,
(64.7/2)x(2/3)=21.57
By head count,
64.7/2=32.35
*Note whether his name suggests or not Brian Lawrence,
superstar among white Putnam Fellows, is Jewish/White (Jewish impact of above
number for whites is clearly apparent whether sineruse likes it or not)
Obviously, East Asian #s are helped by selective immigration
patterns (less so than Indians or SE Asians) from East Asia to US. However, the above ratio seems a bit high.
I am merely pointing out the indisputable fact that Sineruse has literally been obsessing over East Asians for years, as evidenced by his years of activity on forums like College Confidential. He probably knows the autobiographical details of every single East Asian American who has ever qualified for the USAMO over the past 20 years.
ReplyDeleteThe last time someone was this obsessed with members of another ethnic group was when Adolf Hitler obsessed over the Jews, although perhaps that is not entirely fair since I've noticed that Steve Sailer also has a bizarre Jewish obsession.
By way, you do not seem to understand the definition of ad hominem. An ad hominem is when you try to discredit a person's argument by attacking their character. I did not try to discredit any of Sineruse's arguments by attacking his character. I merely pointed out that he is truly and bizarrely obsessed with East Asians.
http://en.wikipedia.org/wiki/Ad_hominem
An ad hominem (Latin for "to the man" or "to the person"), short for argumentum ad hominem, is an attempt to negate the truth of a claim by pointing out a negative characteristic or belief of the person supporting it.[1]
I've lost track of all the IMO/Putnam/ethnicity census analysis on this thread. While you guys are at it, can you verify for me that Caltech has historically produced more Putnam Fellows per undergraduate than any other school? (I'm guessing 2x lead over MIT and even bigger vs Harvard :-)
ReplyDeleteIf we just count US HS educated competitors, a roughly 6x overrepresentation of E. Asians would result from, e.g.:
+.4 SD math advantage; Putnam winner threshold +4 SD; tail area ratio: (+3.6 SD / +4 SD) = (16/100k) / (3/100k) = 5-6
Sineruse: I used similar SDs for the two groups -- I hope that doesn't bother you :-)
But perhaps the correct number is more than 6x? There's a high performing W subpopulation that is hard to keep track of ...
BTW, when examining small numbers of extreme cases there are many factors to take into account. I know a lot of people from the population under discussion. A fair number of HS math competitors end up in CS or physics in college and don't really want to spend energy on Putnam (e.g., learning theoretical CS or quantum mechanics or general relativity are sufficiently taxing that they don't have spare cycles for competition math...). Are you sure factors like this aren't dominating the effect you are looking for? I can imagine a lot of pushy Asian parents telling their kids to do more practical stuff than math. Some IMO winners even end up as doctors or lawyers!
You are pretty obviously trying to discredit him--'dishonest propaganda', etc on grounds that he is biased. I have three problems with that
ReplyDelete1) You haven't adequately addressed his argument. Your hypothetical "Indians vs East Asians" scenario doesn't do the logical work at hand. If East Asians were 10x represented in IMO vis. Indians but all the Indians became Putnam fellows vs only half the East Asians, that shows the kind of "striving" process Sineruse is describing. East Asians are still over represented, but that isn't the point. The main point is that at each level of selection their credentials overpredict. I guess the second point would be that there then may be some threshold in which they are not over represented. That will make you go wild I'm sure, but the observation that East Asian American IMO dominance gets dramatically reduced at the Putnam tells us that a lot more than G goes into IMO results, and presumably (though to a lesser extent) Putnam results.*
2) You accusing other people of bias on this issue is silly. You are obviously a partisan. At the very least you should try to be polite. One of the things I like about this blog is that the posts are insightful and the discussions civilized. Lets leave the Hitler comparisons for the rest of the internet.
*If Sineruse is correct in his interpretation of the data. I haven't and am not going to check his accuracy. But I welcome you to.
"If East Asians were 10x represented in IMO vis. Indians but all the
ReplyDeleteIndians became Putnam fellows vs only half the East Asians, that shows
the kind of "striving" process Sineruse is describing"
The problem is, does it? Let's say in the future we finally have an African American IMO competitor and he goes on to become a Putnam Fellow. Will Sineruse then argue that since historically all African American IMO competitors became Putnam Fellows, that clearly proves that African Americans excel at math relative to East Asian Americans and that clearly higher levels of East Asians at the earlier stages can be attributed to higher conscientiousness?
The fact that so few members of an ethnic group qualify for the earlier stages may indicate that there is simply less math talent available within that ethnic group on a per capita basis. That out of those who make it so far, a larger percentage make it further, may be nothing more an an interesting statistical anomaly.
All of that hard work by Sineruse to prove that at every stage of the way, East Asian Americans are significantly over-represented relative to their population numbers.
ReplyDelete"But perhaps the correct number is more than 6x? There's a high
performing W subpopulation that is hard to keep track of ... Also, what
is the actual W to E. Asian population ratio?"
Sure, the correct way should be to separate out Jews from whites when comparing the relative sizes of the most recent East Asian American to white American cohorts. RKU estimated that East Asian Americans made up 5% or so of recent high school cohorts while non-Jewish white Americans made up more than 50%. I think it's likely that 60% is a fairer estimate. So assuming a 12x over-representation at the population level and from RKU's post...
" On the other hand, if we net out those foreign East Asians, the EA/white winner ratio drops to 4/6[Are any of the 6 whites actually Jews?].." for Putnam Fellows from 2006-2011, we obtain 8x as the correct degree of over-representation relative to whites, assuming that all 6 of the whites were actually non-Jewish.
"I guess the second point would be that there then may be some threshold in which they are not over represented."
ReplyDelete20+ times over representation according to above at Putnam Fellow level isn't overrepresentation? I addressed striving issue in USAMO qualifier to IMO members and IMO members to Putnam in tables in this thread. Even after tampering with data (give "anecdotal evidence" to certain data points), no statistical significance difference is found. He also committed logical fallacies in his claims such as predicting IMO Team members from the '90s by looking at recent USAMO representation by ethnicity when the Asian representation is obviously growing. He also drew interrelations between JHS - HS - College Math competitions within the same calendar year. He then tries to point to specific data sets (spanning 1 or 2 years) to further his outlandish claims. I've shown I can do the same using recent USAMO qualifier to USAMO winners representation. He also conveniently excuses himself from explaining discriminations Indian Americans are facing, or apparently "huge" under performance by white population. His intentions are clear. He will continue to conjure up ludicrous ideas by forcing some limited data sets. Perhaps using "Hitler" analogy was overboard, but the defamation was certainly well deserved.
I don't know how the 5% representation of East Asians among US grade schools is drawn up. They represent 2% of US Population according to census. The number of foreign students (immigrant US residents would be reflected in the US Census) who study by themselves in pre-University level is pretty small.
ReplyDelete2% to 5% obviously creates huge (X2.5) ratio differential
If you want to compare Asians (about 5%) to whites, you need to include Indian American representation in the figure.
Since I don't understand the justification of counting same person several times in finding discrete number of Putnam Fellows, there's over 30 times relative differential between East Asians and whites in the limited sample below.
*Can't divide one person (e.g. Brian Lawrence) 3 times to get 3 highly talented individuals
**I only used this limited sample because sineruse himself selected it. I hope he doesn't have any problem with the pool.
I don't know how the 5% representation of East Asians among US grade
ReplyDeleteschools is drawn up. They represent 2% of US Population according to
census. The number of foreign students (immigrant US residents would be
reflected in the US Census) who study by themselves in pre-University
level is pretty small.
2% to 5% obviously creates huge (X2.5) ratio differential
If you want to compare Asians (about 5%) to whites, you need to include Indian American representation in the figure.
Since
I don't understand the justification of counting same person several
times in finding discrete number of Putnam Fellows, there's over 30
times relative differential between East Asians and whites in the
limited sample below.
*Can't divide one person (e.g. Brian Lawrence) 3 times to get 3 highly talented individuals
**I only used this limited sample because sineruse himself selected it. I hope he doesn't have any problem with the pool.
You are conflating Sineruse's argument with a sample size problem. In the Putnam case, the sample size is maybe 50. It could be a statistical anomaly, but its unlikely. Your 1 amazing Indian vs many more talented East Asian argument attempts to characterize Sineruse's claim as a kind of logical fallacy where he is unaware of the disrepresentative potential of small samples. If it were really just 1 amazing white kid vs 50 talented East Asians it would be a valid point. But its more like 50, so you are really criticizing the statistical validity of a sample size of 50.
ReplyDeleteWe can run simple Bayesianism on that to see that it is very unlikely that the kind of pattern Sineruse has drawn occurred by chance. Far more likely, an Asian "striver" effect (studying for the IMO to get into college admissions, studying less for the Putnam because they now have other academic responsibilities) is artificially boosting the Asian IMO scores per G. Which would fit with Sineruse's main contention that striving causes credentials to overpredict Asian performance.
Well I'm not sure what Sineruse can conclude from an Indian American sample size of 3.
ReplyDeleteMore information would always be nice, but distinguishing the jews from gentiles would probably be beyond our means. At best we could estimate the Jewish proportion and subtract it from the general white representation. But obviously there are a bunch of other data which would be relevant. How many Asians per population go to Stem majors? Before investigating further, that might cause higher Asian Putnam participation and success. But then we need to try to disentangle Stem majorship from putative G advantages, etc.
ReplyDeleteDo you think Sineruse deliberately aggregates gentiles and Jews into an overall white category? I believe that he does this deliberately.
ReplyDeleteI think he would argue that the relative Jewish proportion of white Putnam achievers is not at issue, because the relative proportion of Asian vs White Putnam achievers is not at issue. What is at issue is the reduction in the Asian dominance seen at the highschool level. To me, Sineruse's arguments strongly suggest that something of a "striver" or effort effect underlies part of East Asian USAMO/IMO dominance. Which connects with the earlier arguments about East Asian admission overprediction.
ReplyDelete"Do you think Sineruse deliberately aggregates Jews and gentiles into an overall white category?"
ReplyDeleteSteve Hsu and Yan Shen give themselves away here. In attempting to shore up their ethnic pride vis-a-vis white "gentiles" they've failed to think through their demand that (in this particular instance only!) whites and Jews be disaggregated.
I'll give them a hint: sineruse estimates 25% of white Putnam winners are Jewish. Based on numbers from Hillel, 50% or more of "whites" at Harvard are Jewish.
Steve (or even more laughably Yan Shen), do you continue to maintain that you're motivated purely by a desire for race-blind college admissions?
In case you seem to forget, it is not Steve Hsu or I who have decided to turn this into an HBD debate. Recently it has become impossible to debate the actual issue at hand without hordes of white nationalist types flooding the comments section with their East Asian bashing. They are the ones steering the debate in this direction. If that's how the conversation must be, then I am more than willing to play ball.
ReplyDeleteQuite interesting that NE Asians and Ashkenazim outperform Europeans yet mathematics and physics and natural science is the invention of Europeans.
ReplyDeleteAs one questioner asked Pinker after his talk with Gopnik, "What accounts for the poor intellectual performance of Jews prior to the 19th century." Pinker had no idea what the questioner was talking about. Read the midrashim or the Talmud. The Summa Theoligica, the work of one man, is a MUCH greater intellectual accomplishment than the Talmud which was produced over a span of 800 years by hundreds of people.
Perhaps the industrial revolution in China will result 200 y from now in a population inferior to Europeans. How many contemporary politicians could have written the federalist papers? How many contemporary playwrightscan write like Shakespeare? The industrial revolution was dysgenic?
Perhaps I was mistaken but it seemed sineruse com
ReplyDelete"So why haven't we heard Sineruse demand that less Jewish Americans be let in to elite colleges?"
ReplyDeletePossibly because the debate was framed by Steve Hsu as being about alleged discrimination in elite college admissions against East Asians with respect to "an overall white category". Sineruse didn't need to further break down the "white" group in order to dispute Steve's claims. On the other hand, someone truly driven by a desire for "fair" apportionment of elite college places certainly would have reason to note the degree of Jewish overrepresention. As for why Steve did not, and had no desire to consider Jews separately until he thought he could score a few points for Asian pride, I have some guesses, but I'd like to see Steve's response first.
"By his own numbers, East Asian Americans are over-represented at elite colleges by a factor less than their degree of over-representation amongst Putnam Fellows."
Didn't sineruse just get through explaining to you to you why that _should_ be the case (namely, the intelligence threshold for Putnam Fellows is higher than for Ivy League admits)? In fact, I believe that using some of your own numbers he showed that East Asian Americans are overrepresented at elite colleges relative to their representation among Putnam fellows.
Do you fancy yourself highly quantitatively gifted? Because I'm getting an entirely different impression of your abilities.
"Assuming what you've stated is correct, what this suggests is that Sineruse is an ethnic anti-East Asian activist who obsesses over East Asian Americans when he should instead argue for a decline in Jewish American numbers on elite campuses."
Again, I'd have similar questions about Steve Hsu. But nothing I've said undermines any of sineruse's arguments, which you are still going to have to address on their own merit if you want anyone to take you seriously.
Yan Shen, you do your arguments a disservice by indulging in such hyperbole.
ReplyDeleteFor the sake of being taken seriously you should refrain from such preposterous talk.
After reading most of these comments, I feel a lot less self-conscious about my occasional, half-baked comment.
ReplyDelete"Perhaps I was mistaken but it seemed sineruse complained about the +.5 SD I used in an earlier post. I thought it funny that all of his digging leads to a similar result."
ReplyDeleteThat's interesting, considering sineruse already did that calculation using some of your own assumptions in an earlier thread: http://infoproc.blogspot.com/2011/12/asian-admissions-statistical-prediction.html#comment-394696899
He used used a 12x EA to white ratio for Putnam Fellows, rather than 6x, and came up with a difference of 0.64 SD. The upshot? Even using a ratio twice the actual one, "it seems that East Asians are already enrolled beyond the levels implied by the wildest fantasies of China Pride internet trolls, in which the math-only performance distribution and "HBD"/Asian-nationalist/"g" arguments are combined to infer what a nondiscriminatory, talent-based Asian enrollment would look like."
And this debate _was_ supposedly about "fairness" in college admissions right?
"The W subpopulation affects the estimate of the .5 SD."
So what? That "W subpopulation" is more heavily represented among elite college students than among Putnam Fellows. Explain why you thought it would be a good idea to break them out here, but not when complaining about anti-Asian admissions discrimination in favor of "whites"?
"Didn't sineruse just get through explaining to you to you why that
ReplyDelete_should_ be the case (namely, the intelligence threshold for Putnam
Fellows is higher than for Ivy League admits)? In fact, I believe that
using some of your own numbers he showed that East Asian Americans are
overrepresented at elite colleges relative to their representation among
Putnam fellows."
Perhaps you don't understand. Sineruse's entire argument is that East Asian over-representation decreases the further up the cognitive pipeline you go. In theory, it should increase, according to what he dubs the SND representation. According to him, East Asian conscientiousness explains their higher degree of representation at the earlier levels and it is this high representation at the earlier levels that is particularly objectionable in his view. However, I pointed out to him that East Asian over-representation at elite colleges is less than their representation amongst Putnam Fellows, which is how it should be.
"In fact, I believe that using some of your own numbers he showed that
East Asian Americans are overrepresented at elite colleges relative to
their representation among Putnam fellows. "
Uh no. According to his own numbers East Asian Americans are over-represented by a factor of at least 6x amongst Putnam Fellows. I pointed out that the corresponding degree of over-representation amongst Ivy League undergrads is only 3.5-4x. Please read the comments more carefully.
Well, I'm not a specialist in evolutionary genetics, but Lahn's discoveries did seem absolutely fascinating to me and they did as well to those specialists whom I know, and also to the top science journalists at the New York Times and the rest of the international media.
ReplyDeleteWhether the particular genes in question actually impacted the evolution of brain size itself is completely irrelevent. The key points are that (1) they play a major role in brain structure (e.g. mutations cause severe brain abnormalities) and (2) they apparently underwent massive genetic sweeps over the last few tens of thousands of years. The speed of the sweeps indicate that these particular genes were subject to huge selective pressure, almost certainly because they positively impacted cognitive functioning in some manner. Furthermore, these sweeps occurred right around the same time that archaeological evidence indicates revolutionary cultural change among the peoples experiencing those sweeps, which raises fascinating additional speculations.
Personally, I'd say this is clearly a Nobel Prize level finding. And it's really a great shame that Lahn was pressured into abandoning this line of research, though I suppose at some point he can always move home to China, where intellectual freedom for scientific inquiry is much more widely respected these days.
And as for your Northern European crime data, don't forget I specifically pointed to "Nordics and Germanics" rather than Northern Europeans in general. In fact, Russians and most other Slavs have historically had a much higher rate of "anti-social" behavior, often involving violent drunkenness. Furthermore, the Balkans is heavily Slavic, and notorious within Europe for its endless violent feuds and political disorder. And the disastrous social legacy of Communism and post-Communist Oligarchic-Capitalist has been devastating to the Soviet successor states, including Russia and the Baltic Republics.
I don't know, some of those lists provided by sineruse contained either zero Asians or so few Asians that it hardly seems to matter what percentage of the whites were Jews. What is striking - no, shocking (well, not to me, at least not anymore) is the near total absence of Asians at the very highest levels. And many of the white names are unambiguously non-Jewish. OK, now I have to go and salute the giant Hitler poster hanging in my room - because, you know, I noticed that facts contradict the Asian narrative of supremacy, which is something clearly only a disciple of hitler would have the temerity to do.
ReplyDeleteYeah, but it is pretty standard MO for people losing an argument. As long as pure facts are enough, people tend to stick to them. When the facts are not enough, one suddenly feel compelled to "merely point out" that the other person is a moral scumbag whose arguments are motivated by Nazi-like sympathies.
ReplyDeleteWhy the knee-jerk reaction, Dr. Hsu? This reaction is pretty absurd given that you rarely invoke a cultural argument for anything. I've read almost every post and comment of yours (since the inception of this blog), and don't see that healthy agnosticism reflected in many of them, but that's of course a personal reading of your posts. I don't intend to post here again (I don't think I'm adding any value to the discussion anyway), but I stand by what I see as subtle ethnic chauvinism on this blog.
ReplyDelete"I disagree with both Ju and you about the choice of metric for
ReplyDeletemeasuring math talent. Where you focus on the Putnam/IMO/USAMO/AIME
others, including me, may focus on high-end research output. "
Note how sineruse instigated whole thing by using MATHCOUNTS-IMO-Putnam correlation to cook up seemingly bizarre "striving" factor. I never claimed math competition results should be a sole metric nor Research output isn't important. It was only a natural response to use data he mentioned to refute his whole argument.
In addition, East Asian representation in sciences continue to grow. It would be interesting to see how these trends will go in 20-30 years to come. If the research atmosphere remain more or less the same (dominated by the Europeans), it would be hard not to admit your point.
My conclusion was that the 50th coin toss in a row, not significant in and of itself, has just come up Heads again. The chance of W > A in a small sample with a third of the W's excluded (i.e., the W/A ratio suppressed by about 1.5 as an artificial precondition of the analysis) and further shrinkage from disregarding the changes in the contests and their demographics over time, was not a priori all that good. But the experiment succeeds anyway, and our friend is now fighting tooth and nail to "re-engineer" the data set to include more Asians (but not whites) after a calculation to his specifications didn't work out when done properly.
ReplyDeleteAs I have emphasized many times, the point is not the individual experiments, though they are often interesting. It is that any experiment that can be run on publically available data sets always seems to point up the same in direction, unless transparently rigged such as examining data from one year and six students, while ignoring a nearby year with 1/6 Asians on the team, or the very statistically significant 3:1 or 6:1 decimation of the E.Asian/white ratio going from the 400+ person USAMO qualifier lists to the US Putnam top 50.
"Sineruse's data is meant to show (I think, although again I could be wrong because I don't follow all the comments) that Asian outperformance is mainly due to "striving" and not due to any cognitive advantage, and hence should fall off as you go further into the tail. To analyze this properly you need to take the W subpopulation into account."
ReplyDeleteYour revised explanation makes no sense in light of your original comment and your previous explanation for it, and you've avoided addressing my actual points. You also happen to be wrong about what "Sineruse's data is meant to show", but this has nothing to do with the issue at hand.
Sineruse's data does suggest striving plays some role in Asian academic success (which you say you agree with), and sineruse has argued that the data fail to show signals of anti-Asian discrimination that would be expected were such actually occurring. I haven't seen him argue that there are no cognitive differences. But if that's the argument you thought you were responding to, you certainly didn't bother to engage it much. Sineruse's data might be taken to imply striverish behavior accounts for some proportion of Asian overrepresentation even at the Putnam level. According to what you're now claiming is your interpretation of sineruse, sineruse would say _all_ Asian overrepresentation is due to Asian striving. You didn't address that argument at all in your comment. You simply assumed Asian overrepresentation at the Putnam level is entirely a result of greater average Asian math ability. And if you're going to assume that, there's no need to calculate a difference in standard deviations. The fact that Asians are overrepresented is enough to falsify a claim for no cognitive differences (again, if you're assuming "shifted normal distributions" and that ethnic differences in representation at the Putnam level are a function of ethnic differences in math ability). Nor once you've already "proved" Asian superiority over "whites" would there be any need to prove Asians are even more superior to "gentile" whites.
But none of this matters, because, again: "sineruse already did that calculation using some of your own assumptions in an earlier thread: http://infoproc.blogspot.com/2...
He used used a 12x EA to white ratio for Putnam Fellows, rather than 6x, and came up with a difference of 0.64 SD. The upshot? Even using a ratio twice the actual one, "it seems that East Asians are already enrolled beyond the levels implied by the wildest fantasies of China Pride internet trolls, in which the math-only performance distribution and "HBD"/Asian-nationalist/"g" arguments are combined to infer what a nondiscriminatory, talent-based Asian enrollment would look like.""
And, again: "That "W subpopulation" is more heavily represented among [the "white" proportion of] elite college students than among ["white"] Putnam Fellows."
"I'm for race blind admissions, as I stated very specifically in the earlier post."
So, I'm curious. Under your ideal race-blind admissions regime, we know you take it as axiomatic that Asian representation at Harvard would increase. Do you believe the non-Jewish white proportion of Harvard undergraduates would rise _less_ than the Asian proportion (or stay the same, or decline)?
"Sineruse's data is meant to show (I think, although again I could be wrong because I don't follow all the comments) that Asian outperformance is mainly due to "striving" and not due to any cognitive advantage, and hence should fall off as you go further into the tail. To analyze this properly you need to take the W subpopulation into account."
ReplyDeleteYour revised explanation makes no sense in light of your original comment and your previous explanation for it, and you've avoided addressing my actual points. You also happen to be wrong about what "Sineruse's data is meant to show", but this has nothing to do with the issue at hand.
Sineruse's data does suggest striving plays some role in Asian academic success (which you say you agree with), and sineruse has argued that the data fail to show signals of anti-Asian discrimination that would be expected were such actually occurring. I haven't seen him argue that there are no cognitive differences. But if that's the argument you thought you were responding to, you certainly didn't bother to engage it much. Sineruse's data might be taken to imply striverish behavior accounts for some proportion of Asian overrepresentation even at the Putnam level. According to what you're now claiming is your interpretation of sineruse, sineruse would say _all_ Asian overrepresentation is due to Asian striving. You didn't address that argument at all in your comment. You simply assumed Asian overrepresentation at the Putnam level is entirely a result of greater average Asian math ability. And if you're going to assume that, there's no need to calculate a difference in standard deviations. The fact that Asians are overrepresented is enough to falsify a claim for no cognitive differences (again, if you're assuming "shifted normal distributions" and that ethnic differences in representation at the Putnam level are a function of ethnic differences in math ability). Nor once you've already "proved" Asian superiority over "whites" would there be any need to prove Asians are even more superior to "gentile" whites.
But none of this matters, because, again: "sineruse already did that calculation using some of your own assumptions in an earlier thread: http://infoproc.blogspot.com/2...
He used used a 12x EA to white ratio for Putnam Fellows, rather than 6x, and came up with a difference of 0.64 SD. The upshot? Even using a ratio twice the actual one, "it seems that East Asians are already enrolled beyond the levels implied by the wildest fantasies of China Pride internet trolls, in which the math-only performance distribution and "HBD"/Asian-nationalist/"g" arguments are combined to infer what a nondiscriminatory, talent-based Asian enrollment would look like.""
And, again: "That "W subpopulation" is more heavily represented among [the "white" proportion of] elite college students than among ["white"] Putnam Fellows."
"I'm for race blind admissions, as I stated very specifically in the earlier post."
So, I'm curious. Under your ideal race-blind admissions regime, we know you take it as axiomatic that Asian representation at Harvard would increase. Do you believe the non-Jewish white proportion of Harvard undergraduates would rise _less_ than the Asian proportion (or stay the same, or decline)?
test
ReplyDelete> Indian
ReplyDeleteI think you missed the point.
Controlling for "I" strengthens the statistical experiments about W > A regardless of sample size and whether or not you are also comparing I and W.
Given the apparent empirical validity of the Indian greater-or-equal performance pattern in all the numerical and anecdotal data considered, and the dissimilarity in academic behavior and outcomes for I and EA (as well as I vs W), one has to separate the Indian data, because:
-if lumped together with whites, then especially for the small data sets like Putnam winners there would be an accusation of fudging the white data by adding a higher-performing group;
-if lumped together with E.Asians, then in small or noisy data sets, the greater-or-equal Indian performance would mask the E.Asian underperformance
There is also a lack of sociological data and anecdotes as to whether Indians strive anywhere near as hard as some E.Asian populations, and since that is at the heart of the underperformance phenomenon then it is not clear why one should follow the US Census in lumping South Asians with East Asians.
So few Indians qualify at these higher thresholds that is it meaningless to infer anything.
ReplyDelete"Sineruse's data is meant to show (I think, although again I could be wrong because I don't follow all the comments) that Asian outperformance is mainly due to "striving" and not due to any cognitive advantage, and hence should fall off as you go further into the tail. To analyze this properly you need to take the W subpopulation into account."
ReplyDeleteYour revised explanation makes no sense in light of your original comment and your previous explanation for it, and you've avoided addressing my actual points. You also happen to be wrong about what "Sineruse's data is meant to show", but this has nothing to do with the issue at hand.
Sineruse's data does suggest striving plays some role in Asian academic success (which you say you agree with), and sineruse has argued that the data fail to show signals of anti-Asian discrimination that would be expected were such actually occurring. I haven't seen him argue that there are no cognitive differences. But if that's the argument you thought you were responding to, you certainly didn't bother to engage it much. Sineruse's data might be taken to imply striverish behavior accounts for some proportion of Asian overrepresentation even at the Putnam level. According to what you're now claiming is your interpretation of sineruse, sineruse would say _all_ Asian overrepresentation is due to Asian striving. You didn't address that argument at all in your comment. You simply assumed Asian overrepresentation at the Putnam level is entirely a result of greater average Asian math ability. And if you're going to assume that, there's no need to calculate a difference in standard deviations. The fact that Asians are overrepresented is enough to falsify a claim for no cognitive differences (again, if you're assuming "shifted normal distributions" and that ethnic differences in representation at the Putnam level are a function of ethnic differences in math ability). Nor once you've already "proved" Asian superiority over "whites" would there be any need to prove Asians are even more superior to "gentile" whites.
(cont.)
ReplyDeleteBut none of this matters, because, again: "sineruse already did that calculation using some of your own assumptions in an earlier thread: http://infoproc.blogspot.com/2...He used used a 12x EA to white ratio for Putnam Fellows, rather than 6x, and came up with a difference of 0.64 SD. The upshot? Even using a ratio twice the actual one, "it seems that East Asians are already enrolled beyond the levels implied by the wildest fantasies of China Pride internet trolls, in which the math-only performance distribution and "HBD"/Asian-nationalist/"g" arguments are combined to infer what a nondiscriminatory, talent-based Asian enrollment would look like.""And, again: "That "W subpopulation" is more heavily represented among [the "white" proportion of] elite college students than among ["white"] Putnam Fellows.""I'm for race blind admissions, as I stated very specifically in the earlier post."So, I'm curious. Under your ideal race-blind admissions regime, we know you take it as axiomatic that Asian representation at Harvard would increase. Do you believe the non-Jewish white proportion of Harvard undergraduates would rise _less_ than the Asian proportion (or stay the same, or decline)?
But none of this matters, because, again: "sineruse already did that calculation using some of your own assumptions in an earlier thread:
ReplyDeleteHe used used a 12x EA to white ratio for Putnam Fellows, rather than 6x, and came up with a difference of 0.64 SD. The upshot? Even using a ratio twice the actual one, "it seems that East Asians are already enrolled beyond the levels implied by the wildest fantasies of China Pride internet trolls, in which the math-only performance distribution and "HBD"/Asian-nationalist/"g" arguments are combined to infer what a nondiscriminatory, talent-based Asian enrollment would look like.""
And, again: "That "W subpopulation" is more heavily represented among [the "white" proportion of] elite college students than among ["white"] Putnam Fellows."
"I'm for race blind admissions, as I stated very specifically in the earlier post."
So, I'm curious. Under your ideal race-blind admissions regime, we know you take it as axiomatic that Asian representation at Harvard would increase. Do you believe the non-Jewish white proportion of Harvard undergraduates would rise _less_ than the Asian proportion (or stay the same, or decline)?
Again, perhaps I am mistaken, but I recall Sineruse complaining about a SND ("Shifted Normal Distribution") model in some of his posts. He adopts it in your excerpt above to argue that *even within* a SND model Asians are overrepresented relative to ability (g) at elite universities. That doesn't mean he actually endorses the SND model -- he is just using it there to make a point. I get the impression he doesn't believe in the model, but I'd welcome his clarification on this.
ReplyDeleteIIRC, in at least some of Espenshade's admissions models the elimination of preferences would lead to an increase in the Asian population and no change in the W population.
"Again, perhaps I am mistaken, but I recall Sineruse complaining about a SND ("Shifted Normal Distribution") model in some of his posts. He adopts it in your excerpt above to argue that *even within* a SND model Asians are overrepresented relative to ability (g) at elite universities. That doesn't mean he actually endorses the SND model"
ReplyDeleteI'm not here to argue about what sineruse believes. You _are_ a proponent of the "SND model". So please address his point "that *even within* a SND model Asians are overrepresented relative to ability (g) at elite universities".
"IIRC, in at least some of Espenshade's admissions models the elimination of preferences would lead to an increase in the Asian population and no change in the W population."
I asked specifically about whites who are not members of the group you style "a high performing W subpopulation".
"So please address his point "that *even within* a SND model Asians are
ReplyDeleteoverrepresented relative to ability (g) at elite universities". "
One major fallacy that both you and Sineruse commit is to ignore the fact that Asian Americans are already significantly over-represented in the applicant pool to elite universities. It is possible that high IQ white Americans are less likely to apply to elite universities relative to high IQ East Asian Americans. But this is entirely irrelevant to the debate.
Sineruse, I'm going to give you the best advice that anyone ever has or ever will give you throughout your life. Please seek some psychiatric help.
ReplyDeleteI have never seen someone so singularly obsessed with another ethnic group before. It's obvious to me that the East Asian Question consumes you day and night. You are not normal. There is something profoundly wrong mentally with you. Anyone who has ever read your comments before on College Confidential would be amazed that you're still going at it.
As long as you have reason to suspect some relevant statistical differences, then no matter how few or how many Indians are in the data, it removes some noise from the analysis, and certainly removes some noise from the internet discussion, if one controls for the "I" variable. That there might be additional theories in play, such as your adding Indians to (or rather, below) Koreans on the list of races you deem mathematically inferior to Chinese, is irrelevant.
ReplyDeleteI think any answer to the Asian g question has to look deeply into the verbal vs spatial vs Math question. "G" loses most of its importance if for any or most given cognitive tasks verbal or spatial is more important. Jewish performance would be the best cursory way to investigate this. There are at least three important and reasonably tractable questions:
ReplyDelete1) Is there really such a thing as "M" once verbal, spatial, and study have been factored in. While it is easy to see how verbal or visuo-spatial abilities per se might have evolved it is harder to see how a discreet math module could have come to be. Plus there are some studies which suggest gender gaps in M disappear once S is held constant. M may not actually be important by itself.
2) What is the verbal, spatial, 'M' and 'g' of Ashkenazi populations?
3) Is Ashkenazi intellectual achievement representation best explained by G, pure V, M, or S? Because Ashkenazi V is supposed to be highish (108+) and S lowish, do they perform like the simple 108 or the lower combined V and S g score. Or, if M is real, their 108+ M.
> He specifically asked for an example of an increase in representation
ReplyDeletefrom current percentage of college undergraduates to some intellectual
accomplishment post-admission.
I asked for an example with increase of representation from the percentage *in the pool from which the selection is effectively made* after admission (i.e., the smallest reasonable pool that will include all those who pass the second selection).
The selection pool for Putnam fellows does not have the demographics of college students, but that of the USAMO qualifiers/winners, or the US IMO team training camp, or the same for the IPhO, or some similar elite level of the math and science selection pyramid. These pools are almost 100 percent Asian and white, and the E.A/W ratio is much higher than for college undergraduates. As it is, in the 2010 Putnam data the A/W ratio in the top 50 and 14 was far lower than in the Harvard student body (and it would get worse when excluding athletes), so even the meaningless comparison to college undergraduates does not always work the way you thought.
As another example, the selection pool for Harvard valedictorians does not have the same white and Asian percentages as the student body; it has the demographics of the top hundred or two students at the time of admission, based on academic credentials such as SAT, AP, USAMO, GPA, Intel, etc. This pool could easily be majority E. Asian by now, it was 38% Asian in Espenshade's data in fall 1997 when applicants were ranked by SAT. Maybe one in ten of the Harvard US valedictorians are Asian, but 40+ percent are Jewish.
Exactly! And he was furiously attacked from the right, not the left, by the influential journalists of Germany for his reluctance to go to war in the Moroccan crises.
ReplyDeleteEarlier, there were many entries pointing to high S ability among eminent physicists. I believe this is very likely, since a lot of problems in physics involve setting up hypothetical situations and emphasize interactions between body and energy and such. I wonder if this high S (if it is more or less distinct from M) trend carries over to pure mathematicians though. Solving mathematical problems tend to involve thinking in numbers instead of pictures in my experience. Certainly, picturing dimensional figures in number theory sounds odd. Obviously, there's geometry and more applied math'ish areas involving differential equations and transformations, but situations are more idealized in these cases and usually one isn't surprised by some unaccounted factors as in physics. I think these areas still involve high S but perhaps less than theoretical physics to some degree (with some applied math areas being close to physics)? I'm just speculating here and would love to hear from people more knowledgeable about this matter.
ReplyDeleteEarlier, there were many entries pointing to high S ability among
ReplyDeleteeminent physicists. I believe this is very likely, since a lot of
problems in physics involve setting up hypothetical situations and
emphasize interactions between body and energy and such. I wonder if
this high S (if it is more or less distinct from M) trend carries over
to pure mathematicians though. Solving mathematical problems tend to
involve thinking in numbers instead of pictures in my experience.
Certainly, picturing dimensional figures in number theory sounds odd.
Obviously, there's geometry and more applied math'ish areas involving
differential equations and transformations, but situations are more
idealized in these cases and usually one isn't surprised by some
unaccounted factors as in physics. I think these areas still involve
high S but perhaps less than theoretical physics to some degree (with
some applied math areas being close to physics)? I'm just speculating
here and would love to hear from people more knowledgeable about this
matter.
Earlier, there were many entries pointing to high S ability among
ReplyDeleteeminent physicists. I believe this is very likely, since a lot of
problems in physics involve setting up hypothetical situations and
emphasize interactions between body and energy and such. I wonder if
this high S (if it is more or less distinct from M) trend carries over
to pure mathematicians though. Solving mathematical problems tend to
involve thinking in numbers instead of pictures in my experience.
Certainly, picturing dimensional figures in number theory sounds odd.
Obviously, there's geometry and more applied math'ish areas involving
differential equations and transformations, but situations are more
idealized in these cases and usually one isn't surprised by some
unaccounted factors as in physics. I think these areas still involve
high S but perhaps less than theoretical physics to some degree (with
some applied math areas being close to physics)? I'm just speculating
here and would love to hear from people more knowledgeable about this
matter.
Earlier, there were many entries pointing to high S ability among
ReplyDeleteeminent physicists. I believe this is very likely, since a lot of
problems in physics involve setting up hypothetical situations and
emphasize interactions between body and energy and such. I wonder if
this high S (if it is more or less distinct from M) trend carries over
to pure mathematicians though. Solving mathematical problems tend to
involve thinking in numbers instead of pictures in my experience.
Certainly, picturing dimensional figures in number theory sounds odd.
Obviously, there's geometry and more applied math'ish areas involving
differential equations and transformations, but situations are more
idealized (also stationary v. dynamic) in these cases and usually one isn't surprised by some
unaccounted factors as in physics. I think these areas still involve
high S but perhaps less than theoretical physics to some degree (with
some applied math areas being close to physics)? I'm just speculating
here and would love to hear from people more knowledgeable about this
matter.
Earlier, there were many entries pointing to high S ability among
ReplyDeleteeminent physicists. I believe this is very likely, since a lot of
problems in physics involve setting up hypothetical situations and
emphasize interactions between body and energy and such. I wonder if
this high S (if it is more or less distinct from M) trend carries over
to pure mathematicians though. Solving mathematical problems tend to
involve thinking in numbers instead of pictures in my experience.
Certainly, picturing dimensional figures in number theory sounds odd.
Obviously, there's geometry and more applied math'ish areas involving
differential equations and transformations, but situations are more
idealized (also stationary v. dynamic) in these cases and usually one isn't surprised by some
unaccounted factors as in physics. I think these areas still involve
high S but perhaps less than theoretical physics to some degree (with
some applied math areas being close to physics)? I'm just speculating
here and would love to hear from people more knowledgeable about this
matter.
Another possibility is that you are obsessed with the question of whether I am obsessed with Asians. Steve just threatened another poster with banishment for implying that he is a racist. I see no reason to not apply the same standard to commenters.
ReplyDeleteI'm not an expert on population genetics either. Really just reporting my perception of the media. If you know people in that field and they've let you in on another perception, that would probably be better to pay attention to I guess.
ReplyDeleteFrom google it seems that the role of MCPH and ASPM in brain development was discovered earlier by Geoff Woods working at Leeds University and his collaborators (apparently by studying the somewhat inbred British Pakistani community around Leeds), including the National Human Genome Research Institute. Not sure if you meant to credit that to Bruce Lahn and his team or not.
I wouldn't say running a test of selection on genes related to brain development with functional mutations found by other people and then finding that although selected, you cannot find any functional phenotypical significance is Nobel worthy - as much as it's definitely a smart and worthwhile thing to have done.
....
I think I may have assumed you were using Nordic to refer to all Northern Europeans, or to the Nordic countries plus Finland and Estonia, rather than the Scandiwegian countries plus Germany and the Netherlands.
Still, even if we discard the broad sense of Northern European by removing the British Islands and the North Eastern Europeans, and you just compare an average of Germany, Austria, Norway, Sweden, Denmark and the Netherlands to an average of Italy, Spain and Portugal you get a murder rate for 2009 (when wiki last has complete data) of 0.81 versus 1.02 (per 100,000 population) (if we weight more to bigger countries they're about equal, but there's no real reason to do that).
That doesn't seem like a big difference - there are larger differences between the countries.
I've felt Southern Europeans actually seem more like Asians to me than Northern Europeans, in that Northern Europeans are known when compared to Southern European countries, in terms of crime, for violent anti-social behaviour when drunk and committing property crime and in terms of their society for relatively low collectivism and low corruption and high rates of tax observance. That seems like the pattern for at least North American Whites vs Asian societies generally.
http://www.telegraph.co.uk/news/uknews/1541699/Britain-tops-European-crime-league.html - for more European crime statistics, which don't really seem to have a Germanic-Southern pattern (although I personally would be more skeptical towards anything that isn't homicide).
Well, I'd say that Sinerase raises some perfectly valid points, but I think his seeming obsessiveness on this subject leads him to miss the forest for the trees...
ReplyDeleteI'd grant that Olympiad and Putnam results are not a perfect measure of ultra-high-end math ability, but I do believe they are the best and most objective such measures we have. And on those measures, American East Asians seem over-represented relative to whites (including Jews) by a factor of at least six, or perhaps even closer to a factor of twenty as Ju Hyung Ahn suggests, depending upon the correct age-cohort ratios, which I don't have at hand. Obviously, lots of different factors may be responsible for this, but I'd say that the most parsimonious explanation would certainly include the assumption that East Asians are just quite a lot better at math than most whites.
Consider the close analogy of basketball. Offhand, I'd guess that the over-representation of blacks in American professional basketball is roughly in the same range as the East Asian over-representation among Olympiad/Putnam winners. No doubt if S.J. Gould were around, he'd produce voluminous statistical charts and graphs and sociological studies proving that this was all just some sort of optical illusion and that blacks weren't really any better at basketball than other groups. I'm sure many of his arguments would be perfectly valid ones: young blacks probably practice more, black families and culture probably focus much more on success in that sport, and there's probably a certain amount of racial bias against non-blacks among many of the coaches and recruiters. But I still think the biggest factor is probably that blacks are just much better at basketball.
In science, it's usually pretty difficult to absolutely prove anything. But most of the time, the simplest explanation is probably the best one to follow.
I'm getting a bit impatient with your stupidity. I already posited (and I think Sineruse realizes this, but you are apparently too thick), that Asian strength of application is enhanced relative to cognitive ability. That's what the striving is all about. It's very possible that if elite schools took the rank ordered top N students (how to define "top"? see below) in the US the Asian fraction would be less than the 20% it is currently. This is an artificial scenario though, because Asians and Jews are much more interested in attending these elite schools than even the very smart fraction of W's. That is, they apply to such schools with much higher probability. When this is factored in I don't actually know what the population ratios "should" be. I advocate a race-blind market solution -- let people apply, and let schools admit, but in a race blind way.
ReplyDeleteI've never advocated that university admission should be based on pure ability (g), or some linear combination of g with Conscientiousness (striving), or plus leadership and athletic qualifications, or even based on a projection of future success (however that is defined). Each school can make up its own mind as to how it wants to admit applicants. But the criteria should be applied uniformly to all ethnicities. There should not be penalties and bonuses based crudely on race.
Regarding your last question, I don't know the answer, but if we are only talking about racial preferences, I don't think dropping them would affect gentiles / Jews differently. Legacy/athlete preference might be a wash but I am not sure.
He not only called me a racist but suggested I might treat research students in my group differently, based on ethnicity.
ReplyDeleteBTW, I don't see why it's necessary to protect the reputation of an anonymous commenter.
One day passed and no response from you. I can you're still going on about cooking up more unfounded theories.
ReplyDeleteBelow I calculated (taking your last 5 year performance as a sample) that East Asian Americans have 30+ times higher per capita representation than white population in distinct Putnam Fellows (3 time Putnam Fellow doesn't turn into 3 talented 1 time Putnam Fellow individuals; notice how he is also jewish - the group you'd like to artificially lower representation on highest echelon math).
Notice foreign Putnam Fellows you emphasized so much on were removed from consideration, so the calculation is more or less accurate. All this was with about equal # of Jewish Americans (mostly Ashkenazim) and East Asian Americans (Chinese, Japanese, and Korean group) represented in the population pool.
"This is an artificial scenario though, because Asians and Jews are much more interested in attending these elite schools than even the very smart fraction of W's. That is, they apply to such schools with much higher probability. When this is factored in I don't actually know what the population ratios "should" be."
ReplyDeleteSee, this is where I'm getting confused. You were quite eager to try to calculate absolute differences in ability between Asians and whites in the context of discussions about elite college admissions. Now, when you remember they don't really support what you were going for, they are irrelevant.
Kind of reminds me of your special pleading at http://infoproc.blogspot.com/2011/12/asian-admissions-statistical-prediction.html, in which (in the wake of sineruse's appearance) you take your principled stand on race-blind admissions while adding a footnote that Harvard better remember China will dominate in the coming century.
Since you're getting impatient with my stupidity, might I suggest you turn your attention to sineruse's points about the absence of discrimination-based Asian overperformance.
"I advocate a race-blind market solution"
I'm still wondering how you square this deeply-held commitment with your support for Barack Obama.
> ...absolute differences in ability between Asians and whites in the context of discussions about elite college admissions
ReplyDeleteI'm interested in this kind of thing (HBD) independent of college admissions. (But note I don't know to what extent these differences are genetic.) I don't think I can be any more clear about what policies I favor for college admissions.
As to why I voted for Obama, there are many factor that go into choosing one candidate over another. I can support Obama over McCain even if I disagree with him on certain issues. There's no big contradiction here, just as there wasn't with my earlier calculation.
You obviously aren't adding anything to this discussion. Why don't you just admit you are in over your head? Being dim is OK, but being aggressively dim on my blog is not.
"You wanted to know why I did the little calculation (somehow accusing me of something nefarious), and now (finally, after repeated clarification) I have explained it to you."
ReplyDeleteYour explanations made no sense, and any objective reader can see who lost that argument. You either have very poor reasoning ability or your mind was clouded by other considerations. I accused you of being motivated by ethnic pride and your nonsensical "clarification" failed to overturn that impression.
Let's review:
1. Steve posts: "+.4 SD math advantage; Putnam winner threshold +4 SD; tail area ratio: (+3.6 SD / +4 SD) = (16/100k) / (3/100k) = 5-6" "But perhaps the correct number is more than 6x? There's a high performing W subpopulation that is hard to keep track of ... Also, what is the actual W to E. Asian population ratio?"
2. Steve's explanation #1: "Perhaps I was mistaken but it seemed sineruse complained about the +.5 SD I used in an earlier post. I thought it funny that all of his digging leads to a similar result." "The W subpopulation affects the estimate of the .5 SD."
3. Steve's explanation #2: "Sineruse's data is meant to show (I think, although again I could be wrong because I don't follow all the comments) that Asian outperformance is mainly due to "striving" and not due to any cognitive advantage, and hence should fall off as you go further into the tail. To analyze this properly you need to take the W subpopulation into account."
Note that the explanations don't match and the second makes even less sense than the first, which itself makes no sense for reasons I explained.
4. Earlier post, in which Steve makes no distinction between whites and Jews and showing the context for his calculation is Ivy League admissions: "But if, say, Asians have a .5 SD advantage in g and 1 SD advantage in conscientiousness or work ethic, that might lead to a "fair" Ivy population representation which is less than 20% if by "fair" we mean: apportion slots based on future success odds."
But since I'm "obviously" not adding anything, please don't let me keep you from responding to sineruse's arguments directly.
I guess I'll let other commenters decide whether what I wrote makes sense. You do realize that +.5 SD (or whatever) is *relative* to a reference population, and that you have to separate performance of that population from others in making a calculation? There are related groups of models here, some have a "striving" effect, other just have g, etc. But it seems plausible (and I thought this was also sineruse's point) that striving has more impact on lower level competition than at the tippy top. But when you finally get to Putnam Fellows I would suggest that the effect of striving has decreased a lot. If this doesn't make sense to you, too bad. I doubt other readers are having trouble with it.
ReplyDelete"You do realize that +.5 SD (or whatever) is *relative* to a reference population, and that you have to separate performance of that population from others in making a calculation?"
ReplyDeleteSo where exactly does a non-Jewish white reference population come in? Because only non-Jewish whites apply to Harvard? But then didn't you just say that the general population doesn't matter, only the Ivy League applicant pool, and non-Jewish whites are _less_ likely to apply? And yet you're happy to suppose for the sake of your calculations that intelligent whites and Asians are comparably likely to devote themselves to competitive math (at least up to the college level, where you're ready to speculate Asians may be _less_ attracted to math). You've certainly engaged in some sloppy thinking. Possibly this was careless. It appears to me more likely that it's been helped along by conscious or unconscious ethnic chauvinism.
"But it seems plausible (and I thought this was also sineruse's point) that striving has more impact on lower level competition than at the tippy top."
Yes that's one of sineruse's major points. No an intelligent person would not try to address this line of argument by doing the calculation you did. An intelligent person would directly compare Asian/white ratios at different levels of various pipelines, as sineruse has done. If for some reason he did feel like estimating mean differences from ratios at the tail (introducing unnecessary assumptions), I'd hope he'd realize he'd need to convert all the data points, not just one.
"But when you finally get to Putnam Fellows I would suggest that the effect of striving has decreased a lot."
Decreased relative to lower level math competitions, certainly. Decreased to zero? You've done nothing to show that and made no attempt to address the issue at all when doing your "little calculation".
2010 Putnam (International)
ReplyDeleteThe Putnam Fellows--The Five Highest Ranking Individuals
YU DENG [China] Gx1
SEOK HYEONG LEE [South Korea] Gx1, Sx1
The Next Nine Highest Ranking Individuals
IURIE BOREICO [Moldova] Gx3, Sx2
ZEHAO LI
TENGYU MA [China] Sx1
GYUJIN OH [South Korea] Gx1
The Next Ten Highest Ranking Individuals
WHAN GHANG [South Korea] Gx2, Sx1
KONSTANTIN MATVEEV [Soviet Union] Gx1
XIAOSHENG MU [China] Gx1
ALEXANDR ZAMORZAEV [Moldova] Gx1, Sx3
Honorable Mention: Individuals
FRANK D. BAN
MOHAMMAD BAVARIAN [Iran] Sx1
CRAIG M. BURKHART
YI-WEI CHAN [Taiwan] Gx1, Sx1
YU CHENG
NATHAN R. HARMAN
MATTHEW A. HARRISON-TRAINOR
IN-JEE JEONG
STEVEN N. KARP [Canada] Bx1
EMIL T. LALOV
PAK HIN LEE [Hong Kong] Sx1
SUNKYU LIM [South Korea] Sx2
KEVIN LIN
ZHI KIN LOKE [Malaysia] Sx2, Bx2
NIKITA LVOV
LYUBOSLAV N. PANCHEV [Bulgaria] Gx2, Sx1
YUE PU
ALEXANDER REMOROV [Canada] Sx1, Bx1
ELINA M. ROBEVA [Bulgaria] Sx2
JONATHAN SCHNEIDER [Canada] Sx1, Bx1, participation x1
TENGYAO WANG
YINGHUI WANG
CHIU WAI WONG [Hong Kong] Sx2, Bx1, participation x1
GJERGJI ZAIMI [Albania] Sx2, Bx1, HMx1
ZEYIN ZHANG
2010 Putnam (International)
ReplyDeleteThe Putnam Fellows--The Five Highest Ranking Individuals
YU DENG [China] Gx1
SEOK HYEONG LEE [South Korea] Gx1, Sx1
The Next Nine Highest Ranking Individuals
IURIE BOREICO [Moldova] Gx3, Sx2
ZEHAO LI
TENGYU MA [China] Sx1
GYUJIN OH [South Korea] Gx1
The Next Ten Highest Ranking Individuals
WHAN GHANG [South Korea] Gx2, Sx1
KONSTANTIN MATVEEV [Soviet Union] Gx1
XIAOSHENG MU [China] Gx1
ALEXANDR ZAMORZAEV [Moldova] Gx1, Sx3
Honorable Mention: Individuals
FRANK D. BAN
MOHAMMAD BAVARIAN [Iran] Sx1
CRAIG M. BURKHART
YI-WEI CHAN [Taiwan] Gx1, Sx1
YU CHENG
NATHAN R. HARMAN
MATTHEW A. HARRISON-TRAINOR
IN-JEE JEONG
STEVEN N. KARP [Canada] Bx1
EMIL T. LALOV
PAK HIN LEE [Hong Kong] Sx1
SUNKYU LIM [South Korea] Sx2
KEVIN LIN
ZHI KIN LOKE [Malaysia] Sx2, Bx2
NIKITA LVOV
LYUBOSLAV N. PANCHEV [Bulgaria] Gx2, Sx1
YUE PU
ALEXANDER REMOROV [Canada] Sx1, Bx1
ELINA M. ROBEVA [Bulgaria] Sx2
JONATHAN SCHNEIDER [Canada] Sx1, Bx1, participation x1
TENGYAO WANG
YINGHUI WANG
CHIU WAI WONG [Hong Kong] Sx2, Bx1, participation x1
GJERGJI ZAIMI [Albania] Sx2, Bx1, HMx1
ZEYIN ZHANG
2010 Putnam (International)
ReplyDeleteThe Putnam Fellows--The Five Highest Ranking Individuals
YU DENG [China] Gx1
SEOK HYEONG LEE [South Korea] Gx1, Sx1
The Next Nine Highest Ranking Individuals
IURIE BOREICO [Moldova] Gx3, Sx2
ZEHAO LI
TENGYU MA [China] Sx1
GYUJIN OH [South Korea] Gx1
The Next Ten Highest Ranking Individuals
WHAN GHANG [South Korea] Gx2, Sx1
KONSTANTIN MATVEEV [Russia] Gx1
XIAOSHENG MU [China] Gx1
ALEXANDR ZAMORZAEV [Moldova] Gx1, Sx3
Honorable Mention: Individuals
FRANK D. BAN
MOHAMMAD BAVARIAN [Iran] Sx1
CRAIG M. BURKHART
YI-WEI CHAN [Taiwan] Gx1, Sx1
YU CHENG
NATHAN R. HARMAN
MATTHEW A. HARRISON-TRAINOR
IN-JEE JEONG
STEVEN N. KARP [Canada] Bx1
EMIL T. LALOV
PAK HIN LEE [Hong Kong] Sx1
SUNKYU LIM [South Korea] Sx2
KEVIN LIN
ZHI KIN LOKE [Malaysia] Sx2, Bx2
NIKITA LVOV
LYUBOSLAV N. PANCHEV [Bulgaria] Gx2, Sx1
YUE PU
ALEXANDER REMOROV [Canada] Sx1, Bx1
ELINA M. ROBEVA [Bulgaria] Sx2
JONATHAN SCHNEIDER [Canada] Sx1, Bx1, participation (no
reward) x1
TENGYAO WANG
YINGHUI WANG
CHIU WAI WONG [Hong Kong] Sx2, Bx1, participation (no
reward) x1
GJERGJI ZAIMI [Albania] Sx2, Bx1, HMx1
ZEYIN ZHANG
Looks like European internationals are under performing compare to their East Asian counterpart. Compare performance from IMO heavy weights (Russia & Bulgaria - China & South Korea).
Those w/o distinction were neither part of International IMO teams of USAMO.
2010
ReplyDeletePutnam (International)
The
Putnam Fellows--The Five Highest Ranking Individuals
YU
DENG [China] Gx1
SEOK
HYEONG LEE [South Korea] Gx1, Sx1
The
Next Nine Highest Ranking Individuals
IURIE
BOREICO [Moldova] Gx3, Sx2
ZEHAO
LI
TENGYU
MA [China] Sx1
GYUJIN
OH [South Korea] Gx1
The
Next Ten Highest Ranking Individuals
WHAN
GHANG [South Korea] Gx2, Sx1
KONSTANTIN
MATVEEV [Russia] Gx1
XIAOSHENG
MU [China] Gx1
ALEXANDR
ZAMORZAEV [Moldova] Gx1, Sx3
Honorable
Mention: Individuals
FRANK
D. BAN
MOHAMMAD
BAVARIAN [Iran] Sx1
CRAIG
M. BURKHART
YI-WEI
CHAN [Taiwan] Gx1, Sx1
YU
CHENG
NATHAN
R. HARMAN
MATTHEW
A. HARRISON-TRAINOR
IN-JEE
JEONG
STEVEN
N. KARP [Canada] Bx1
EMIL
T. LALOV
PAK
HIN LEE [Hong Kong] Sx1
SUNKYU
LIM [South Korea] Sx2
KEVIN
LIN
ZHI
KIN LOKE [Malaysia] Sx2, Bx2
NIKITA
LVOV
LYUBOSLAV
N. PANCHEV [Bulgaria] Gx2, Sx1
YUE
PU
ALEXANDER
REMOROV [Canada] Sx1, Bx1
ELINA
M. ROBEVA [Bulgaria] Sx2
JONATHAN
SCHNEIDER [Canada] Sx1, Bx1, participation (no reward) x1
TENGYAO
WANG
YINGHUI
WANG
CHIU
WAI WONG [Hong Kong] Sx2, Bx1, participation (no reward) x1
GJERGJI
ZAIMI [Albania] Sx2, Bx1, HMx1
ZEYIN
ZHANG
Looks
like European internationals are under performing compare to their East Asian
counterpart. Compare performance from IMO heavy weights (Russia &
Bulgaria - China & South Korea).
Those
w/o distinction were neither part of International IMO teams of USAMO.
2010
ReplyDeletePutnam (International)
The
Putnam Fellows--The Five Highest Ranking Individuals
YU
DENG [China] Gx1
SEOK
HYEONG LEE [South Korea] Gx1, Sx1
The
Next Nine Highest Ranking Individuals
IURIE
BOREICO [Moldova] Gx3, Sx2
ZEHAO
LI
TENGYU
MA [China] Sx1
GYUJIN
OH [South Korea] Gx1
The
Next Ten Highest Ranking Individuals
WHAN
GHANG [South Korea] Gx2, Sx1
KONSTANTIN
MATVEEV [Russia] Gx1
XIAOSHENG
MU [China] Gx1
ALEXANDR
ZAMORZAEV [Moldova] Gx1, Sx3
Honorable
Mention: Individuals
FRANK
D. BAN
MOHAMMAD
BAVARIAN [Iran] Sx1
CRAIG
M. BURKHART
YI-WEI
CHAN [Taiwan] Gx1, Sx1
YU
CHENG
NATHAN
R. HARMAN
MATTHEW
A. HARRISON-TRAINOR
IN-JEE
JEONG
STEVEN
N. KARP [Canada] Bx1
EMIL
T. LALOV
PAK
HIN LEE [Hong Kong] Sx1
SUNKYU
LIM [South Korea] Sx2
KEVIN
LIN
ZHI
KIN LOKE [Malaysia] Sx2, Bx2
NIKITA
LVOV
LYUBOSLAV
N. PANCHEV [Bulgaria] Gx2, Sx1
YUE
PU
ALEXANDER
REMOROV [Canada] Sx1, Bx1
ELINA
M. ROBEVA [Bulgaria] Sx2
JONATHAN
SCHNEIDER [Canada] Sx1, Bx1, participation (no reward) x1
TENGYAO
WANG
YINGHUI
WANG
CHIU
WAI WONG [Hong Kong] Sx2, Bx1, participation (no reward) x1
GJERGJI
ZAIMI [Albania] Sx2, Bx1, HMx1
ZEYIN
ZHANG
Looks
like European internationals are under performing compare to their East Asian
counterpart. Compare performance from IMO heavy weights (Russia &
Bulgaria - China & South Korea).
Those
w/o distinction were neither part of International IMO teams nor USAMO.
Since
ReplyDelete;;
ReplyDeleteSince
ReplyDelete;;
ReplyDeleteLooks
like European IMO credentials tend to over predict their performance in Putnam.
2010
Putnam (International)
The
Putnam Fellows--The Five Highest Ranking Individuals
YU
DENG [China] Gx1
SEOK
HYEONG LEE [South Korea] Gx1, Sx1
The
Next Nine Highest Ranking Individuals
IURIE
BOREICO [Moldova] Gx3, Sx2
ZEHAO
LI
TENGYU
MA [China] Sx1
GYUJIN
OH [South Korea] Gx1
The
Next Ten Highest Ranking Individuals
WHAN
GHANG [South Korea] Gx2, Sx1
KONSTANTIN
MATVEEV [Russia] Gx1
XIAOSHENG
MU [China] Gx1
ALEXANDR
ZAMORZAEV [Moldova] Gx1, Sx3
Honorable
Mention: Individuals
FRANK
D. BAN
MOHAMMAD
BAVARIAN [Iran] Sx1
CRAIG
M. BURKHART
YI-WEI
CHAN [Taiwan] Gx1, Sx1
YU
CHENG
NATHAN
R. HARMAN
MATTHEW
A. HARRISON-TRAINOR
IN-JEE
JEONG
STEVEN
N. KARP [Canada] Bx1
EMIL
T. LALOV
PAK
HIN LEE [Hong Kong] Sx1
SUNKYU
LIM [South Korea] Sx2
KEVIN
LIN
ZHI
KIN LOKE [Malaysia] Sx2, Bx2
NIKITA
LVOV
LYUBOSLAV
N. PANCHEV [Bulgaria] Gx2, Sx1
YUE
PU
ALEXANDER
REMOROV [Canada] Sx1, Bx1
ELINA
M. ROBEVA [Bulgaria] Sx2
JONATHAN
SCHNEIDER [Canada] Sx1, Bx1, participation (no reward) x1
TENGYAO
WANG
YINGHUI
WANG
CHIU
WAI WONG [Hong Kong] Sx2, Bx1, participation (no reward) x1
GJERGJI
ZAIMI [Albania] Sx2, Bx1, HMx1
ZEYIN
ZHANG
European
internationals are under performing compare to their East Asian
counterpart. Compare performance from IMO heavy weights (Russia &
Bulgaria - China & South Korea).
Those
w/o distinction were neither part of International IMO teams nor USAMO.
::
ReplyDeleteLooks
like European IMO credentials tend to over predict their performance in Putnam.
2010
Putnam (International)
The
Putnam Fellows--The Five Highest Ranking Individuals
YU
DENG [China] Gx1
SEOK
HYEONG LEE [South Korea] Gx1, Sx1
The
Next Nine Highest Ranking Individuals
IURIE
BOREICO [Moldova] Gx3, Sx2
ZEHAO
LI
TENGYU
MA [China] Sx1
GYUJIN
OH [South Korea] Gx1
The
Next Ten Highest Ranking Individuals
WHAN
GHANG [South Korea] Gx2, Sx1
KONSTANTIN
MATVEEV [Russia] Gx1
XIAOSHENG
MU [China] Gx1
ALEXANDR
ZAMORZAEV [Moldova] Gx1, Sx3
Honorable
Mention: Individuals
FRANK
D. BAN
MOHAMMAD
BAVARIAN [Iran] Sx1
CRAIG
M. BURKHART
YI-WEI
CHAN [Taiwan] Gx1, Sx1
YU
CHENG
NATHAN
R. HARMAN
MATTHEW
A. HARRISON-TRAINOR
IN-JEE
JEONG
STEVEN
N. KARP [Canada] Bx1
EMIL
T. LALOV
PAK
HIN LEE [Hong Kong] Sx1
SUNKYU
LIM [South Korea] Sx2
KEVIN
LIN
ZHI
KIN LOKE [Malaysia] Sx2, Bx2
NIKITA
LVOV
LYUBOSLAV
N. PANCHEV [Bulgaria] Gx2, Sx1
YUE
PU
ALEXANDER
REMOROV [Canada] Sx1, Bx1
ELINA
M. ROBEVA [Bulgaria] Sx2
JONATHAN
SCHNEIDER [Canada] Sx1, Bx1, participation (no reward) x1
TENGYAO
WANG
YINGHUI
WANG
CHIU
WAI WONG [Hong Kong] Sx2, Bx1, participation (no reward) x1
GJERGJI
ZAIMI [Albania] Sx2, Bx1, HMx1
ZEYIN
ZHANG
European
internationals are under performing compare to their East Asian
counterpart. Compare performance from IMO heavy weights (Russia &
Bulgaria - China & South Korea).
Those
w/o distinction were neither part of International IMO teams nor USAMO.
Looks
ReplyDeletelike European IMO credentials tend to over predict their performance in Putnam.
2010
Putnam (International)
The
Putnam Fellows--The Five Highest Ranking Individuals
YU
DENG [China] Gx1
SEOK
HYEONG LEE [South Korea] Gx1, Sx1
The
Next Nine Highest Ranking Individuals
IURIE
BOREICO [Moldova] Gx3, Sx2
ZEHAO
LI
TENGYU
MA [China] Sx1
GYUJIN
OH [South Korea] Gx1
The
Next Ten Highest Ranking Individuals
WHAN
GHANG [South Korea] Gx2, Sx1
KONSTANTIN
MATVEEV [Russia] Gx1
XIAOSHENG
MU [China] Gx1
ALEXANDR
ZAMORZAEV [Moldova] Gx1, Sx3
Honorable
Mention: Individuals
FRANK
D. BAN
MOHAMMAD
BAVARIAN [Iran] Sx1
CRAIG
M. BURKHART
YI-WEI
CHAN [Taiwan] Gx1, Sx1
YU
CHENG
NATHAN
R. HARMAN
MATTHEW
A. HARRISON-TRAINOR
IN-JEE
JEONG
STEVEN
N. KARP [Canada] Bx1
EMIL
T. LALOV
PAK
HIN LEE [Hong Kong] Sx1
SUNKYU
LIM [South Korea] Sx2
KEVIN
LIN
ZHI
KIN LOKE [Malaysia] Sx2, Bx2
NIKITA
LVOV
LYUBOSLAV
N. PANCHEV [Bulgaria] Gx2, Sx1
YUE
PU
ALEXANDER
REMOROV [Canada] Sx1, Bx1
ELINA
M. ROBEVA [Bulgaria] Sx2
JONATHAN
SCHNEIDER [Canada] Sx1, Bx1, participation (no reward) x1
TENGYAO
WANG
YINGHUI
WANG
CHIU
WAI WONG [Hong Kong] Sx2, Bx1, participation (no reward) x1
GJERGJI
ZAIMI [Albania] Sx2, Bx1, HMx1
ZEYIN
ZHANG
European
internationals are under performing compare to their East Asian
counterpart. Compare performance from IMO heavy weights (Russia &
Bulgaria - China & South Korea).
Those
w/o distinction were neither part of International IMO teams nor USAMO.
Looks
ReplyDeletelike European IMO credentials tend to over predict their performance in Putnam.
2010
Putnam (International)
The
Putnam Fellows--The Five Highest Ranking Individuals
YU
DENG [China] Gx1
SEOK
HYEONG LEE [South Korea] Gx1, Sx1
The
Next Nine Highest Ranking Individuals
IURIE
BOREICO [Moldova] Gx3, Sx2
ZEHAO
LI
TENGYU
MA [China] Sx1
GYUJIN
OH [South Korea] Gx1
The
Next Ten Highest Ranking Individuals
WHAN
GHANG [South Korea] Gx2, Sx1
KONSTANTIN
MATVEEV [Russia] Gx1
XIAOSHENG
MU [China] Gx1
ALEXANDR
ZAMORZAEV [Moldova] Gx1, Sx3
Honorable
Mention: Individuals
FRANK
D. BAN
MOHAMMAD
BAVARIAN [Iran] Sx1
CRAIG
M. BURKHART
YI-WEI
CHAN [Taiwan] Gx1, Sx1
YU
CHENG
NATHAN
R. HARMAN
MATTHEW
A. HARRISON-TRAINOR
IN-JEE
JEONG
STEVEN
N. KARP [Canada] Bx1
EMIL
T. LALOV
PAK
HIN LEE [Hong Kong] Sx1
SUNKYU
LIM [South Korea] Sx2
KEVIN
LIN
ZHI
KIN LOKE [Malaysia] Sx2, Bx2
NIKITA
LVOV
LYUBOSLAV
N. PANCHEV [Bulgaria] Gx2, Sx1
YUE
PU
ALEXANDER
REMOROV [Canada] Sx1, Bx1
ELINA
M. ROBEVA [Bulgaria] Sx2
JONATHAN
SCHNEIDER [Canada] Sx1, Bx1, participation (no reward) x1
TENGYAO
WANG
YINGHUI
WANG
CHIU
WAI WONG [Hong Kong] Sx2, Bx1, participation (no reward) x1
GJERGJI
ZAIMI [Albania] Sx2, Bx1, HMx1
ZEYIN
ZHANG
European
internationals are under performing compare to their East Asian
counterpart. Compare performance from IMO heavy weights (Russia &
Bulgaria - China & South Korea).
Those
w/o distinction were neither part of International IMO teams nor USAMO.
Looks like European IMO credentials tend to over predict their performance in Putnam.
ReplyDelete2010 Putnam (International)
The Putnam Fellows--The Five Highest Ranking Individuals
YU DENG [China] Gx1
SEOK HYEONG LEE [South Korea] Gx1, Sx1
The Next Nine Highest Ranking Individuals
IURIE BOREICO [Moldova] Gx3, Sx2
ZEHAO LI
TENGYU MA [China] Sx1
GYUJIN OH [South Korea] Gx1
The Next Ten Highest Ranking Individuals
WHAN GHANG [South Korea] Gx2, Sx1
KONSTANTIN MATVEEV [Russia] Gx1
XIAOSHENG MU [China] Gx1
ALEXANDR ZAMORZAEV [Moldova] Gx1, Sx3
Honorable Mention: Individuals
FRANK D. BAN
MOHAMMAD BAVARIAN [Iran] Sx1
CRAIG M. BURKHART
YI-WEI CHAN [Taiwan] Gx1, Sx1
YU CHENG
NATHAN R. HARMAN
MATTHEW A. HARRISON-TRAINOR
IN-JEE JEONG
STEVEN N. KARP [Canada] Bx1
EMIL T. LALOV
PAK HIN LEE [Hong Kong] Sx1
SUNKYU LIM [South Korea] Sx2
KEVIN LIN
ZHI KIN LOKE [Malaysia] Sx2, Bx2
NIKITA LVOV
LYUBOSLAV N. PANCHEV [Bulgaria] Gx2, Sx1
YUE PU
ALEXANDER REMOROV [Canada] Sx1, Bx1
ELINA M. ROBEVA [Bulgaria] Sx2
JONATHAN SCHNEIDER [Canada] Sx1, Bx1, participation (no reward) x1
TENGYAO WANG
YINGHUI WANG
CHIU WAI WONG [Hong Kong] Sx2, Bx1, participation (no reward) x1
GJERGJI ZAIMI [Albania] Sx2, Bx1, HMx1
ZEYIN ZHANG
European internationals are under performing compare to their East Asian counterpart. Compare performance from IMO heavy weights (Russia & Bulgaria - China & South Korea).
Those w/o distinction were neither part of International IMO teams nor USAMO.
::
ReplyDeleteLooks like European IMO credentials tend to over predict their performance in Putnam.
2010 Putnam (International)
The Putnam Fellows--The Five Highest Ranking Individuals
YU DENG [China] Gx1
SEOK HYEONG LEE [South Korea] Gx1, Sx1
The Next Nine Highest Ranking Individuals
IURIE BOREICO [Moldova] Gx3, Sx2
ZEHAO LI
TENGYU MA [China] Sx1
GYUJIN OH [South Korea] Gx1
The Next Ten Highest Ranking Individuals
WHAN GHANG [South Korea] Gx2, Sx1
KONSTANTIN MATVEEV [Russia] Gx1
XIAOSHENG MU [China] Gx1
ALEXANDR ZAMORZAEV [Moldova] Gx1, Sx3
Honorable Mention: Individuals
FRANK D. BAN
MOHAMMAD BAVARIAN [Iran] Sx1
CRAIG M. BURKHART
YI-WEI CHAN [Taiwan] Gx1, Sx1
YU CHENG
NATHAN R. HARMAN
MATTHEW A. HARRISON-TRAINOR
IN-JEE JEONG
STEVEN N. KARP [Canada] Bx1
EMIL T. LALOV
PAK HIN LEE [Hong Kong] Sx1
SUNKYU LIM [South Korea] Sx2
KEVIN LIN
ZHI KIN LOKE [Malaysia] Sx2, Bx2
NIKITA LVOV
LYUBOSLAV N. PANCHEV [Bulgaria] Gx2, Sx1
YUE PU
ALEXANDER REMOROV [Canada] Sx1, Bx1
ELINA M. ROBEVA [Bulgaria] Sx2
JONATHAN SCHNEIDER [Canada] Sx1, Bx1, participation (no reward) x1
TENGYAO WANG
YINGHUI WANG
CHIU WAI WONG [Hong Kong] Sx2, Bx1, participation (no reward) x1
GJERGJI ZAIMI [Albania] Sx2, Bx1, HMx1
ZEYIN ZHANG
European internationals are under performing compare to their East Asian counterpart. Compare performance from IMO heavy weights (Russia & Bulgaria - China & South Korea).
Those w/o distinction were neither part of International IMO teams nor USAMO.
::
ReplyDeleteLooks like European IMO credentials tend to over predict their performance in Putnam.
2010 Putnam (International)
The Putnam Fellows--The Five Highest Ranking Individuals
YU DENG [China] Gx1
SEOK HYEONG LEE [South Korea] Gx1, Sx1
The Next Nine Highest Ranking Individuals
IURIE BOREICO [Moldova] Gx3, Sx2
ZEHAO LI
TENGYU MA [China] Sx1
GYUJIN OH [South Korea] Gx1
The Next Ten Highest Ranking Individuals
WHAN GHANG [South Korea] Gx2, Sx1
KONSTANTIN MATVEEV [Russia] Gx1
XIAOSHENG MU [China] Gx1
ALEXANDR ZAMORZAEV [Moldova] Gx1, Sx3
Honorable Mention: Individuals
FRANK D. BAN
MOHAMMAD BAVARIAN [Iran] Sx1
CRAIG M. BURKHART
YI-WEI CHAN [Taiwan] Gx1, Sx1
YU CHENG
NATHAN R. HARMAN
MATTHEW A. HARRISON-TRAINOR
IN-JEE JEONG
STEVEN N. KARP [Canada] Bx1
EMIL T. LALOV
PAK HIN LEE [Hong Kong] Sx1
SUNKYU LIM [South Korea] Sx2
KEVIN LIN
ZHI KIN LOKE [Malaysia] Sx2, Bx2
NIKITA LVOV
LYUBOSLAV N. PANCHEV [Bulgaria] Gx2, Sx1
YUE PU
ALEXANDER REMOROV [Canada] Sx1, Bx1
ELINA M. ROBEVA [Bulgaria] Sx2
JONATHAN SCHNEIDER [Canada] Sx1, Bx1, participation (no reward) x1
TENGYAO WANG
YINGHUI WANG
CHIU WAI WONG [Hong Kong] Sx2, Bx1, participation (no reward) x1
GJERGJI ZAIMI [Albania] Sx2, Bx1, HMx1
ZEYIN ZHANG
European internationals are under performing compare to their East Asian counterpart. Compare performance from IMO heavy weights (Russia & Bulgaria - China & South Korea).
Those w/o distinction were neither part of International IMO teams nor USAMO.
Looks like European IMO credentials tend to over predict their performance in Putnam.
ReplyDelete2010 Putnam (International)
The Putnam Fellows--The Five Highest Ranking Individuals
YU DENG [China] Gx1
SEOK HYEONG LEE [South Korea] Gx1, Sx1
The Next Nine Highest Ranking Individuals
IURIE BOREICO [Moldova] Gx3, Sx2
ZEHAO LI
TENGYU MA [China] Sx1
GYUJIN OH [South Korea] Gx1
The Next Ten Highest Ranking Individuals
WHAN GHANG [South Korea] Gx2, Sx1
KONSTANTIN MATVEEV [Russia] Gx1
XIAOSHENG MU [China] Gx1
ALEXANDR ZAMORZAEV [Moldova] Gx1, Sx3
Honorable Mention: Individuals
FRANK D. BAN
MOHAMMAD BAVARIAN [Iran] Sx1
CRAIG M. BURKHART
YI-WEI CHAN [Taiwan] Gx1, Sx1
YU CHENG
NATHAN R. HARMAN
MATTHEW A. HARRISON-TRAINOR
IN-JEE JEONG
STEVEN N. KARP [Canada] Bx1
EMIL T. LALOV
PAK HIN LEE [Hong Kong] Sx1
SUNKYU LIM [South Korea] Sx2
KEVIN LIN
ZHI KIN LOKE [Malaysia] Sx2, Bx2
NIKITA LVOV
LYUBOSLAV N. PANCHEV [Bulgaria] Gx2, Sx1
YUE PU
ALEXANDER REMOROV [Canada] Sx1, Bx1
ELINA M. ROBEVA [Bulgaria] Sx2
JONATHAN SCHNEIDER [Canada] Sx1, Bx1, participation (no reward) x1
TENGYAO WANG
YINGHUI WANG
CHIU WAI WONG [Hong Kong] Sx2, Bx1, participation (no reward) x1
GJERGJI ZAIMI [Albania] Sx2, Bx1, HMx1
ZEYIN ZHANG
European internationals are under performing compare to their East Asian counterpart. Compare performance from IMO heavy weights (Russia & Bulgaria - China & South Korea).
Those w/o distinction were neither part of International IMO teams nor USAMO.
Looks like European IMO credentials tend to over predict their performance in Putnam.
ReplyDelete::
ReplyDelete2010 Putnam (International)
ReplyDeleteThe Putnam Fellows--The Five Highest Ranking Individuals
YU DENG [China] Gx1
SEOK HYEONG LEE [South Korea] Gx1, Sx1
The Next Nine Highest Ranking Individuals
IURIE BOREICO [Moldova] Gx3, Sx2
ZEHAO LI
TENGYU MA [China] Sx1
GYUJIN OH [South Korea] Gx1
The Next Ten Highest Ranking Individuals
WHAN GHANG [South Korea] Gx2, Sx1
KONSTANTIN MATVEEV [Russia] Gx1
XIAOSHENG MU [China] Gx1
ALEXANDR ZAMORZAEV [Moldova] Gx1, Sx3
Honorable Mention: Individuals
FRANK D. BAN
MOHAMMAD BAVARIAN [Iran] Sx1
CRAIG M. BURKHART
YI-WEI CHAN [Taiwan] Gx1, Sx1
YU CHENG
NATHAN R. HARMAN
MATTHEW A. HARRISON-TRAINOR
IN-JEE JEONG
STEVEN N. KARP [Canada] Bx1
EMIL T. LALOV
PAK HIN LEE [Hong Kong] Sx1
SUNKYU LIM [South Korea] Sx2
KEVIN LIN
ZHI KIN LOKE [Malaysia] Sx2, Bx2
NIKITA LVOV
LYUBOSLAV N. PANCHEV [Bulgaria] Gx2, Sx1
YUE PU
ALEXANDER REMOROV [Canada] Sx1, Bx1
ELINA M. ROBEVA [Bulgaria] Sx2
JONATHAN SCHNEIDER [Canada] Sx1, Bx1, participation (no reward) x1
TENGYAO WANG
YINGHUI WANG
CHIU WAI WONG [Hong Kong] Sx2, Bx1, participation (no reward) x1
GJERGJI ZAIMI [Albania] Sx2, Bx1, HMx1
ZEYIN ZHANG
European internationals are under performing compare to their East Asian counterpart. Compare performance from IMO heavy weights (Russia & Bulgaria - China & South Korea).
Those w/o distinction were neither part of International IMO teams nor USAMO.
Looks like European IMO credentials tend to over predict their performance in Putnam.
ReplyDeleteForeign Influence Among Putnam Fellows
ReplyDelete2010 2 EA (1 China, 1 Korea), 0 white
2009 3 EA (2 China, 1 Canada), 0 white
2008 2 EA (1 Korea, 1 Canada), 0 white
2007 2 EA (2 China), 0 white
2006 2 EA (1 China, 1 Canada), 0 white
There isn't big enough differential in foreign IMO recruits between East Asians and white population to warrant such a lopsided difference.
European IMO recruits are significantly under performing here.
So after losing an argument, Steve's response is to ban me and delete my final comment. Repost:
ReplyDelete"You do realize that +.5 SD (or whatever) is *relative* to a reference population, and that you have to separate performance of that population from others in making a calculation?"
So where exactly does a non-Jewish white reference population come in? Because only non-Jewish whites apply to Harvard? But then didn't you just say that the general population doesn't matter, only the Ivy League applicant pool, and non-Jewish whites are _less_ likely to apply? And yet you're happy to suppose for the sake of your calculations that intelligent whites and Asians are comparably likely to devote themselves to competitive math (at least up to the college level, where you're ready to speculate Asians may be _less_ attracted to math). You've certainly engaged in some sloppy thinking. Possibly this was careless. It appears to me more likely that it's been helped along by conscious or unconscious ethnic chauvinism.
"But it seems plausible (and I thought this was also sineruse's point) that striving has more impact on lower level competition than at the tippy top."
Yes that's one of sineruse's major points. No an intelligent person would not try to address this line of argument by doing the calculation you did. An intelligent person would directly compare Asian/white ratios at different levels of various pipelines, as sineruse has done. If for some reason he did feel like estimating mean differences from ratios at the tail (introducing unnecessary assumptions), I'd hope he'd realize he'd need to convert all the data points, not just one.
"But when you finally get to Putnam Fellows I would suggest that the effect of striving has decreased a lot."
Decreased relative to lower level math competitions, certainly. Decreased to zero? You've done nothing to show that and made no attempt to address the issue at all when doing your "little calculation".
So after losing an argument, Steve's response is to ban me and delete my final comment. Repost:
ReplyDelete"You do realize that +.5 SD (or whatever) is *relative* to a reference population, and that you have to separate performance of that population from others in making a calculation?"
So where exactly does a non-Jewish white reference population come in? Because only non-Jewish whites apply to Harvard? But then didn't you just say that the general population doesn't matter, only the Ivy League applicant pool, and non-Jewish whites are _less_ likely to apply? And yet you're happy to suppose for the sake of your calculations that intelligent whites and Asians are comparably likely to devote themselves to competitive math (at least up to the college level, where you're ready to speculate Asians may be _less_ attracted to math). You've certainly engaged in some sloppy thinking. Possibly this was careless. It appears to me more likely that it's been helped along by conscious or unconscious ethnic chauvinism.
"But it seems plausible (and I thought this was also sineruse's point) that striving has more impact on lower level competition than at the tippy top."
Yes that's one of sineruse's major points. No an intelligent person would not try to address this line of argument by doing the calculation you did. An intelligent person would directly compare Asian/white ratios at different levels of various pipelines, as sineruse has done. If for some reason he did feel like estimating mean differences from ratios at the tail (introducing unnecessary assumptions), I'd hope he'd realize he'd need to convert all the data points, not just one.
"But when you finally get to Putnam Fellows I would suggest that the effect of striving has decreased a lot."
Decreased relative to lower level math competitions, certainly. Decreased to zero? You've done nothing to show that and made no attempt to address the issue at all when doing your "little calculation".
After losing an argument, Steve's response is to ban me and delete my final comment. Repost:
ReplyDelete"You do realize that +.5 SD (or whatever) is *relative* to a reference population, and that you have to separate performance of that population from others in making a calculation?"
So where exactly does a non-Jewish white reference population come in? Because only non-Jewish whites apply to Harvard? But then didn't you just say that the general population doesn't matter, only the Ivy League applicant pool, and non-Jewish whites are _less_ likely to apply? And yet you're happy to suppose for the sake of your calculations that intelligent whites and Asians are comparably likely to devote themselves to competitive math (at least up to the college level, where you're ready to speculate Asians may be _less_ attracted to math). You've certainly engaged in some sloppy thinking. Possibly this was careless. It appears to me more likely that it's been helped along by conscious or unconscious ethnic chauvinism.
"But it seems plausible (and I thought this was also sineruse's point) that striving has more impact on lower level competition than at the tippy top."
Yes that's one of sineruse's major points. No an intelligent person would not try to address this line of argument by doing the calculation you did. An intelligent person would directly compare Asian/white ratios at different levels of various pipelines, as sineruse has done. If for some reason he did feel like estimating mean differences from ratios at the tail (introducing unnecessary assumptions), I'd hope he'd realize he'd need to convert all the data points, not just one.
"But when you finally get to Putnam Fellows I would suggest that the effect of striving has decreased a lot."
Decreased relative to lower level math competitions, certainly. Decreased to zero? You've done nothing to show that and made no attempt to address the issue at all when doing your "little calculation".
Refer to bottom:
ReplyDeleteFor some reason, foreign East Asians are vastly outperforming foreign Europeans & Canadians with similar IMO credentials.
Given that we get more international IMO talent than domestic ones (limited to only 6 each year) in Putnam, the statistics seem at least as relevant as the US ones.
Even assuming there are more foreign East Asian IMO talents (probably not by much esp. with fixed # from Canada each year), the difference is so large that if I try to get a ratio of Putnam representation between the two groups, I get none other than infinity.
I hate being blunt but people like sineruse and you're asking for it.
My view of Lahn's results was that he was the first person in history to discover clear evidence for some of the likely candidate genes whose appearance created the necessary mental basis for the rise of human civilization. If so, I'd say that's certainly a Nobel Prize worthy research result.
ReplyDeleteOtherwise, I was using "Nordic" in the more modern sense, when it's roughly equivalent to "Scandinavian" plus maybe the Finns and Estonians. And in considering crime and other anti-social behavior, one perhaps needs to somewhat adjust for wealth and historical circumstances. In general, I'd say that both East Asian and Nordic+Germanic societies tend to have much higher levels of intrisic collectivism that other groups you can look at.
From a genetic/historical perspective, the Anglo-Saxons also obviously cluster with the Nordics and Germanics, and I think it's more than pure coincidence that every single one of the 15-20 least corrupt countries on earth according to the international tables is either East Asian or derives from one of those three European groups. Given its current poverty and legacy of Communism, China is certainly quite corrupt these days. But I wouldn't be surprised if in another thirty or forty years, it's around where Singapore and Hong Kong are today.
One somewhat distortive fact is that since Northern Europeans (certainly including Russians) have been civilized for far fewer generations than most of the other groups we're considering, their weakness for alcoholism hasn't yet been bred out, and drunkenness is a huge factor in elevating their local violent crime rates relative to peoples who have gradually become immune from that problem.
Anyway, that's my speculative perspective on these sorts of comparisons which may possibly have a significant racial component.
Courtesy of Siserune (Sineruse?)
ReplyDeletehttp://talk.collegeconfidential.com/massachusetts-institute-technology/904583-data-putnam-competition-performance-us-domestic-students-mit-all-schools.html
*removed foreign non-IMO
2009 Putnam (IMO contestants International)
Putnam Fellows
[China] XIAOSHENG MU Yale Gx1
[China] QINGCHUN REN MIT Gx2
[Canada]YUFEI ZHAO MIT Gx1, Sx1, Bx1
The Next Ten Highest Ranking Individuals
[Lithuania] KESTUTIS CESNAVICIUS Rice Gx1, Bx1
The Next Ten Highest Ranking Individuals
[Taiwan] YI-WEI CHAN U of Illinois, Urbana-Champaign Gx1,
Sx1
[Korea] WHAN GHANG MIT Gx2, Sx1
[Korea] SEOK HYEONG LEE Stanford Gx1, Sx1
[Thailand] PANUPONG PASUPAT MIT Gx1, Sx2
Honorable Mention
[China] YE LUO, MIT Gx1
[Moldova] IURIE BOREICO, Harvard Gx3, Sx2
[Moldova] IVAN BORSENCO, MIT Gx1, Bx3
[Brazil] HENRIQUE P. OLIVEIRA PINTO, MIT Gx1, Sx2
[Korea] DOO SUNG PARK, Caltech Gx1, Sx1
[Brazil]GABRIEL BUJOKAS, MIT Gx1, Sx1
[Canada]DONG UK (DAVID) RHEE, Uof Waterloo Sx1, Bx1, HMx1
[China] YU DENG, MIT Gx1
[Canada] XIAOLIN (DANNY) SHI, MIT Gx1, Bx1
[Korea] JEONG SOO SIM, Stanford Sx1
[Korea] YOUNG HUN JUNG, Stanford Sx1, Bx1
[Canada]STEVEN N. KARP, U of Waterloo Bx1
[Thailand] WUTTISAK TRONGSIRIWAT, U of Virginia Sx1, Bx1
[Korea] YEOIL YOON, Caltech Gx1
[Romania] ADRIAN I. ZAHARIUC, Princeton Gx2, Bx2
[Albania] GJERGJI ZAIMI, Caltech Sx2, Bx1, HMx1
General trend happens again, 14 East Asians and 6
Europeans (including Canadians for both sides; excluded Brazilians from Europeans).
East Asians are top heavy (3 to 0 Putnam Fellows, 6 to 1 Top 25).
Courtesy of Siserune (Sineruse?)
ReplyDeletehttp://talk.collegeconfidential.com/massachusetts-institute-technology/904583-data-putnam-competition-performance-us-domestic-students-mit-all-schools.html
*removed foreign non-IMO
2009 Putnam (IMO contestants International)
Putnam Fellows
[China] XIAOSHENG MU Yale Gx1
[China] QINGCHUN REN MIT Gx2
[Canada]YUFEI ZHAO MIT Gx1, Sx1, Bx1
The Next Ten Highest Ranking Individuals
[Lithuania] KESTUTIS CESNAVICIUS Rice Gx1, Bx1
The Next Ten Highest Ranking Individuals
[Taiwan] YI-WEI CHAN U of Illinois, Urbana-Champaign Gx1,
Sx1
[Korea] WHAN GHANG MIT Gx2, Sx1
[Korea] SEOK HYEONG LEE Stanford Gx1, Sx1
[Thailand] PANUPONG PASUPAT MIT Gx1, Sx2
Honorable Mention
[China] YE LUO, MIT Gx1
[Moldova] IURIE BOREICO, Harvard Gx3, Sx2
[Moldova] IVAN BORSENCO, MIT Gx1, Bx3
[Brazil] HENRIQUE P. OLIVEIRA PINTO, MIT Gx1, Sx2
[Korea] DOO SUNG PARK, Caltech Gx1, Sx1
[Brazil]GABRIEL BUJOKAS, MIT Gx1, Sx1
[Canada]DONG UK (DAVID) RHEE, Uof Waterloo Sx1, Bx1, HMx1
[China] YU DENG, MIT Gx1
[Canada] XIAOLIN (DANNY) SHI, MIT Gx1, Bx1
[Korea] JEONG SOO SIM, Stanford Sx1
[Korea] YOUNG HUN JUNG, Stanford Sx1, Bx1
[Canada]STEVEN N. KARP, U of Waterloo Bx1
[Thailand] WUTTISAK TRONGSIRIWAT, U of Virginia Sx1, Bx1
[Korea] YEOIL YOON, Caltech Gx1
[Romania] ADRIAN I. ZAHARIUC, Princeton Gx2, Bx2
[Albania] GJERGJI ZAIMI, Caltech Sx2, Bx1, HMx1
General trend happens again, 14 East Asians and 6
Europeans (including Canadians for both sides; excluded Brazilians from
Europeans). East Asians are top heavy (3
to 0 Putnam Fellows, 6 to 1 Top 25).
"Courtesy of Siserune (Sineruse?)
ReplyDeletehttp://talk.collegeconfidentia..."
Yes. That's the same Sineruse. You managed to find his numerous posts from years ago on the forum College Confidential. He has literally been obsessing over East Asians for years. At the risk of being repetitive, let me state that I have never met anyone who has been so singularly obsessed with another ethnic group before.
I am pretty sure Sineruse is the same guy who writes the blog http://racehist.blogspot.com
ReplyDeleteHe is on relatively good behavior here, but he's a WN. Just look at his blog, esp. the comments.
I largely agree: the most parsimonious explanation is a significant Asian Math advantage. But it is worth pointing out that there is an important disanalogy between black basketball performance and Asian math performance. Whereas black representation increases dramatically along the pipeline (highschool-->college-->NBA), Sineruse argues that Asian representation shrinks. If Sineruse is right about declining representation towards the top the situation is potentially a lot more interesting than the basketball case. Either Sineruse is tricking me or the striving effect is powerful enough to create a dramatically different IMO and Putnam performances.
ReplyDeleteThat seems like a problem for any simple theory of normal distributions. The IMO gold might be more 'strivable' than the Putnam but it also looks like an extremely high G accomplishment. If the only difference between the dramatically different Asian IMO and Putnam achievement is striving then striving is really, really important and we have no way to know how much it might be altering the Putnam results.
Shockley is an interesting case study, and there are lots of other white Gentile scientists with that sort of genuine eccentricity (mental illness?): George Price sprung to mind, but it's a long list to be sure, especially if you are willing to stretch back a few centuries. Anecdotally, I do think most Asians would tend closer to the Teller/von Neumann end of the personality spectrum.
ReplyDeleteComputer programming:Top Coder Rankings(username|country)
ReplyDelete1.AC Rush-China
2.Shangjingbo-China
3.Zhuojie-China
4.a9108tc-China
5.crazyboy-China
6.liympanda-China
7.WJMZBMR-China
8.g201513-China
7.jx-wuyi-China
8Blue.Mary-China.
> For some reason, foreign East Asians are vastly outperforming foreign Europeans & Canadians with similar IMO credentials.
ReplyDeleteThanks for the continued support. Your new table, like the old one, corroborates the model that predicts underperformance for (US) East Asians.
Foreign East Asians come from larger countries with far more difficult IMO selection than the non-Asians you are comparing them against, from places like Moldova, Albania and Canada. Nobody has ever made it to the IMO team three times from China, and only 11 people have done it twice (one was a Putnam winner recruited to MIT). Anyone who can make the China IMO team once would have gone to the IMO 3 times or more if they lived in Moldova, compiling a chest of medals like GGSB. The guys who made the China team twice would have had records more like GGGS. Instead of noticing this obvious fact, you announced that a Chinese with one gold medal who attains the same result as a Moldovan with GGSB on the Putnam contest is "vastly outperforming" him based on "IMO credentials".
Moldova has a population of 3.5 million, 375x smaller than China. In producing IMO high scorers Moldova is doing *much better* per capita than China, which is another self-evident conclusion you did not draw. Bulgaria with its population of 8 million (180x smaller than China) has always outperformed China at the IMO on any reasonable per capita measure, and in 2003 actually beat China and the rest of the world in absolute terms to win the competition. This is astonishing based on population size and can only be explained by levels of training that are high and relatively specific to the IMO. Training implies underperformance: one expects the Bulgarian IMO medalists to not do as well on the Putnam contest as medalists from most other countries with the same scores, because if there is higher training, there is a lower estimated level of ability at any given IMO score. The same principle applies to the analysis of US E.Asian over- and under-performance. If you want to reject that principle then of course it would be fun to hear your or Yan's advocacy for a "Bulgarian g supermen" theory.
As we have seen, many people are "pretty sure" of things that are totally false. I actually had never heard of that blog, nor do I operate in circles where acronyms like WN are common. The term "white nationalist" (not that I specified my race or anything) seems to be a Yan Shen-ism for "making arguments I find hard to dispute on the merits".
ReplyDelete> Well, I'd say that Sineruse raises some perfectly valid points, but I
ReplyDeletethink his seeming obsessiveness on this subject leads him to miss the
forest for the trees...
I'm willing to make specific arguments and defend them on the merits, which is not the same as "obsessiveness" by any stretch. I first mentioned, in passing, the phenomenon of Asian underperformance a couple of years ago on another board. As one might expect, the China Pride warriors went crazy and attempted to refute it with statistics, including one Chinese guy who counted and racially classified all the math olympiad competitors since 1981. Thus began the game of locating public data sets to test various hypotheses. Because the underperformance phenomenon is a mathematical one, engaging in these statistical jousts is like being the casino in a roulette game, except that the odds in favor are closer to 100 percent than 51-49 at each experiment. So it is fun to play, but to relieve boredom I gather a few new data sets each time the subject comes up, in addition to going through the more pedestrian routine of showing that the junk calculations that are always put forth are either obviously wrong, or support the underperformance effect when done in accordance with ordinary mathematics and not the fuzzy kind used to prop up ethnic pride.
> the most parsimonious explanation would certainly
include the assumption that East Asians are just quite a lot better at
math than most whites.
Could be, but it is orthogonal to (and perfectly consistent with) anything I am talking about. As I stressed many times, you can assume that E.Asian ability distribution is higher, lower, or different compared to whites, and you can assume whatever you like about what ability might be and how it works (e.g, is it a finite dimensional Gaussian distribution, are its effects additive and so on). The main impact of a "higher Asian g" assumption on the underperformance discussion is that the more favorable the E.A distribution is, the *stronger* the underperformance phenomenon has to be to account for observed depletions. The suppression of striver effects later in the pipeline has to be strong enough to cause those declines despite the assumed exponential increase in the natural (striving-free) Asian representation at those levels.
The core problem seems to be that the Chinese scientists have trouble connecting with the global scientific community unless they emigrate to Western countries. As you said, all top six heavily cited scientists in material sciences are Chinese, but five out of six hail from American universities! In contrast, all but one Japanese scientist and all Koreans on both lists publish from universities in their home countries.
ReplyDeleteObviously, I was saying how their IMO credentials (winning multiple medals) under predict performance. # of population in native country is irrelevant here.
ReplyDeleteAnd if you want to argue, likewise East Asian Americans are outnumbered by non-white Hispanic Americans by at least 30 times. Some double standards here?
Steve,
ReplyDeleteNot that this is of any importance in the grand scheme of this blog (assume a scheme that is grand), but you've really now lost me and, I would imagine, others like me.
Any physicist whose research delves into population genetics would sooner or later have to discuss Einstein. But how could a blog on the topic de-generate so quickly into a debate about Asian academic/intellectual superiority? I haven't read enough to answer this question, but I hope the reason is that it's been hijacked by dim-witted, euro-american supremacists, as some of your posts seem to imply. Even still, I've read enough of the posts to believe, rightly or wrongly, that this topic holds such emotional significance for you that you're willing to go out on an intellectual limb to defend some of your shakier positions.
As no one in particular (other than a huge fan of your blog since 2004), I urge you to stop.
Oh, and, full disclosure, I am an American-born physicist of jewish decent who knows that Einstein is an outlier in every conceivable sense of the word.
Thanks.
Anonymous.
Sineruse / Siserune: There are many, many comments on this blog that I completely disagree with, on both sides of the argument that you are involved in. Every blog owner has to decide how much effort they are willing to devote to dealing with comments, and I have mainly adopted a laissez faire attitude. (Also, I think it is interesting to see opinions of the sort that are normally suppressed.) The only commenters I have ever banned are people who have directly insulted me (in this case, suggested I am a racist of some sort).
ReplyDeletePS I read through some of your comments on College Confidential and found I agreed with most of them. Is there an email address that I can reach you at?
Hmm... good advice!
ReplyDeleteMy handle here or at CC, at (GOOG mail entity) dot com. I don't check daily but will definitely see it.
ReplyDelete@MtMoRu:
ReplyDeleteBabylonians and Egyptians were the first to ruminate over mathematical abstractions (e.g. fractions, geometry, etcetera). In addition, the Chinese and Indians (just to name a few) have developed a concept of mathematics on their own. In fact, any modern textbook on the history of mathematics can contradict your viewpoint.
(As an example, please research the Rhind mathematical papyrus.)
The third paragraph of your post is perhaps the most telling: "China... population inferior to Europeans."
Will having more smart Europeans make you smarter?
I am a mathematician.
It's interesting the way scientists do science. They have a theory and when they get data, instead of trying to adopt the theory to data, they try to adopt data to the theory. It makes me sick and I've lost all respect to most of the scientists and I'm seriously.
ReplyDeleteFeynman was one of the greatest scientists of all time, one of the sharpest minds and his IQ is "only 125" . Now we need to find all kinds of explanations how come his IQ is only 125? (of course the reverse is also possible, after we found out that his IQ was "so low" we can state "oh... maybe wasn't that smart after all") What is the cutoff point for Feynman's IQ that you would accept as possible so your theories about IQ==intelligence stays preserved?
If it was 145, I think that wouldn't sound good as well, after all he was one of the most brilliant minds. If we could only make it 180, then we could be satisfied.
If we take Einstein for example, his IQ was about 150, we should have millions of Einsteins if IQ is the dominant trait for Einstein's. Most scientists with very high IQs do terrible science. Most of the people with high IQs are complete idiots like the rest of the population. You can find lot of high IQ people with average life accomplishments and with average abilities in 99% of the tasks that are not IQ tests, that is the data and all you do is talk about high IQ as an trait that defines genius! Sure, most geniuses have high IQ as well, as most best attacking basketball players are also good defenders (speed and athletic ability are helpful on both ends of a floor) but's it's not the way you'll find next Einstein. If it was that simple Usain Bolt would be a NBA all star, and for the next Einstein we could find all the kids that are 4SDs above average and drill them to become geniuses. One high IQ person has an idea for an iPhone, and the other umbrella that opens up from your shirt, one musician makes a great album, and the other goes to Guinness book of world records for being able to play most notes per second on a guitar.
Method of finding the next Einstein from a subset of people with IQ > 4SD+average is like trying to find next NBA player by taking people who can jump as high 4SD above average. And if you look at NBA, they all can jump high, but Jordan, Bird, Johnson, Nowitzki, LeBron have a broad range of ability to jump high (but nowhere near Javier Sotomayor). So yes, even if it may not be true that Feynman had an IQ of 125 but somewhat higher, it's not so improbable as you think that he is not part of 4SD club, but only 2SD club.
And the bigger problem is that education system is going in the way of constantly measuring IQ's , and giving grades for ability to play most notes per second while not trying to select people who make good music.
To conclude, IQ is relevant to some extent, next Einstein will probably not have an average IQ, and nations with higher average IQs will probably have more of good engineers and less crime on their streets BUT start looking for other relevant parameters, if you don't know them, and if it's not possible to know them, or you're just not smart enough to find them, it doesn't mean they don't exist.