Zhongguo

It came from the sky

What a great photo (see video). Also looks a bit like observers near a black hole!

NYTimes: When an asteroid exploded over the Russian city of Chelyabinsk in February, shattering windows for miles and injuring well over 1,000 people, experts said it was a rare event — of a magnitude that might occur only once every 100 to 200 years, on average.

But now a team of scientists is suggesting that the Earth is vulnerable to many more Chelyabinsk-size space rocks than was previously thought. In research being published Wednesday by the journal Nature, they estimate that such strikes could occur as often as every decade or two.

... A 450-foot-wide asteroid, Dr. Lu said, would be equivalent to 150 million tons of TNT. “You’re not going to wipe out humanity,” he said, “but if you get unlucky, you could kill 50 million people or you could collapse the world economy for a century, two centuries.”

Dr. Lu said astronomers had found only 10 to 20 percent of the near-Earth asteroids of that size.

Sentinel would also spot many smaller ones that could still be devastating. “What we’ve been talking about are the ones that would only destroy a major metropolitan area — all of New York City and the surrounding area,” Dr. Lu said.

He said only about 0.5 percent of these smaller asteroids, roughly the size of the 1908 one, have been found.

Because telescope surveys have counted so few of the small asteroids, Dr. Brown and his colleagues instead investigated what has actually hit the Earth. In one of the articles in Nature, they examined United States Air Force data from the 1960s and 1970s and later data from sensors verifying a ban on aboveground nuclear weapons testing.

The recordings captured the low-frequency atmospheric rumblings generated by about 60 asteroid explosions. Most came from small asteroids, but their data suggested that the somewhat larger ones hit more frequently than would be expected based on the estimates from sky surveys. That could mean the Earth has been unlucky recently, or that the estimates on the number of Chelyabinsk-size asteroids are too low.

“Any one of them individually I think you could dismiss,” Dr. Brown said, “but when you take it all together, I think the preponderance of the evidence is there is a much higher number of these tens-of-meters-size objects.”

Some ominous analysis here:

A 500-kiloton airburst over Chelyabinsk and an enhanced hazard from small impactors (Nature)

Most large (over a kilometre in diameter) near-Earth asteroids are now known, but recognition that airbursts (or fireballs resulting from nuclear-weapon-sized detonations of meteoroids in the atmosphere) have the potential to do greater damage1 than previously thought has shifted an increasing portion of the residual impact risk (the risk of impact from an unknown object) to smaller objects2. Above the threshold size of impactor at which the atmosphere absorbs sufficient energy to prevent a ground impact, most of the damage is thought to be caused by the airburst shock wave3, but owing to lack of observations this is uncertain4, 5. Here we report an analysis of the damage from the airburst of an asteroid about 19 metres (17 to 20 metres) in diameter southeast of Chelyabinsk, Russia, on 15 February 2013, estimated to have an energy equivalent of approximately 500 (±100) kilotons of trinitrotoluene (TNT, where 1 kiloton of TNT = 4.185×1012 joules). We show that a widely referenced technique4, 5, 6 of estimating airburst damage does not reproduce the observations, and that the mathematical relations7 based on the effects of nuclear weapons—almost always used with this technique—overestimate blast damage. This suggests that earlier damage estimates5, 6 near the threshold impactor size are too high. We performed a global survey of airbursts of a kiloton or more (including Chelyabinsk), and find that the number of impactors with diameters of tens of metres may be an order of magnitude higher than estimates based on other techniques8, 9. This suggests a non-equilibrium (if the population were in a long-term collisional steady state the size-frequency distribution would either follow a single power law or there must be a size-dependent bias in other surveys) in the near-Earth asteroid population for objects 10 to 50 metres in diameter, and shifts more of the residual impact risk to these sizes. [ Italics mine ]

Learning can hurt: UK edition

This BBC Radio 4 discussion focuses on the recent kerfuffle in the UK over revelations that *gasp* genes affect cognitive ability. "Combative, provocative and engaging debate" :-)  @28:28 ad hominem attack on Plomin and BGI by leftist activist David King, well handled by moderator.

Genetics and education (BBC Radio 4)

Duration: 43 minutes
First broadcast: Wednesday 30 October 2013

For centuries philosophers and theologians have wrestled with the question of nature versus nurture. Increasingly and for some controversially, the science of behavioural genetics is starting to come up with some of the answers. The argument is perhaps at its most sensitive when applied to education. When it was revealed that Education Secretary Michael Gove's outgoing special advisor, Dominic Cummings, called for education policy to incorporate the science behind genes and cognitive development he broke a modern taboo and there was a predictable outcry. In a wide ranging paper Mr Cummings cited the work of Professor Robert Plomin who's about to publish a book with psychologist Dr Kathryn Asbury which calls for "genetically sensitive" schooling. It's based on a study of how genes and environment have shaped the development of over 10,000 twins who were studied from birth to early adulthood. The scientists say their work is about probability not prophecy and can be used to personalise education and create better outcomes for all, but fears of genetic determinism are deeply ingrained. How should we use genetics in education? Science is a very long way from knowing exactly which genes influence individual differences in learning but as knowledge in this field advances that time will surely come. We already use genetics to screen for various medical conditions, so why not for learning abilities? And what happens if, or when, the science of genetics becomes so powerful that we can identify different populations that are endowed with different genetic make-ups that we believe are more or less desirable? Is that just a scientific inevitability that we have to come to terms with, or does it open the door to eugenics? How should we use the science of genetics?

Combative, provocative and engaging debate chaired by Michael Buerk. With Michael Portillo, Melanie Phillips, Giles Fraser and Matthew Taylor. DR KATHRYN ASBURY - York University, co-author of 'G is for Genes', DR ANDERS SANDBERG - Research Fellow at the 'Future of Humanity Institute', Oxford University, DR DAVID KING - Founder and Director of the campaign group 'Human Genetics Alert', STEVE DAVY - Teacher at the Wroxham school, Potter's Bar.

It was Dominic Cummings' long essay Some Thoughts on Education and Political Priorities ("leaked" by the Guardian!) that sparked this controversy. See fun Endnote and a few references to this blog.

Dominic Cummings, Genius or Menace? (Guardian): Dominic Cummings is arguably the most brilliant and most controversial special adviser in the coalition. Long seen as a driving force in the Department for Education, he demanded more from a bureaucracy that Conservatives believe was temperamentally sympathetic to Labour.

Some see him as a menace, given to frank exchanges with civil servants and, on occasions, journalists. Others see him as a genius, consistently driving higher standards, clearer exams and taking on the "blob", the Tory term of abuse to describe the educational establishment.

Born and educated in Durham, he secured a first in ancient and modern history from Oxford and worked for Iain Duncan Smith when he was leader of the opposition. In his portrait of the coalition, In It Together, Matthew d'Ancona describes Cummings as being "mild mannered by temperament except when he was not. His volcanic outbursts had astonished Duncan Smith in 2002 when he had briefly been the party's director of strategy". Post-election, Andy Coulson, director of communications, blocked Cummings' appointment as a government adviser on the basis that he might be too independent and a disruptive force. But Michael Gove continued to rely on him from afar and, when Coulson resigned, the education secretary rapidly appointed him.

His 250-page screed sprawls across a vast canvas about the future, education, Britain's place in the world and disruptive forces ahead. Quite frankly, much will pass over the average reader's head. It is either mad, bad or brilliant – and probably a bit of all three.

Ideas and scientific evidence matter, especially (hopefully!) in the long run.

Keynes: "Practical men, who believe themselves to be quite exempt from any intellectual influence, are usually the slaves of some defunct economist. Madmen in authority, who hear voices in the air, are distilling their frenzy from some academic scribbler of a few years back. I am sure that the power of vested interests is vastly exaggerated compared with the gradual encroachment of ideas."

Single cell sequencing in PGD and cancer treatment

Note, the BGI Cognitive Genomics group with which I am associated is not involved in the work described below. Aneuploidy means an abnormal number of chromosomes within a cell, indicative of chromosomal abnormality. The most common type is Down Syndrome.

Single-cell Sequencing Makes Strides in the Clinic with Cancer and PGD First Applications (Genomeweb, October 02, 2013)

Single-cell sequencing is quickly entering the clinic with initial applications in cancer and pre-implantation genetic diagnosis and screening, researchers reported this week at the Beyond the Genome conference in San Francisco, Calif., which was sponsored by Genome Biology and Genome Medicine.

Within the field of pre-implantation genetic diagnosis and screening, BGI is already using single-cell sequencing to screen for aneuploidies prior to in vitro fertilization, and a team from Peking University is testing both single-cell transcriptome sequencing and single-cell whole genome sequencing for applications in IVF.

Meantime, a team from Harvard University has demonstrated through single-cell sequencing that circulating tumor cells from lung cancer patients show unique copy number variation profiles, while another group from Cold Spring Harbor Laboratory has tested single-cell sequencing methods in prostate cancer patients to monitor response to treatment and identify biomarkers and drug targets.

BGI's Fei Gao said that BGI has been testing a method published earlier this year in PLoS One for detecting copy number variants from single-cell, low-pass, whole-genome sequencing on couples undergoing in vitro fertilization.

In August, the first IVF baby that was sequenced before implantation was born healthy, he said, and since then more than 20 healthy babies have been born healthy following pre-IVF single-cell sequencing to screen for aneuploidies and large copy number variants. [ Italics mine. ]

Gao said that the BGI team first tested several kits for whole-genome amplification including ones that used multiple displacement amplification, degenerate oligonucleotide primed PCR, and a technique known as MALBAC developed by Sunney Xie's group at Harvard University. ...

... Gao said the team analyzed the samples for chromosomal aneuploidies and large copy number variants, and showed that the results were concordant with microarrays.

Next, they conducted a study of 41 couples that were undergoing IVF either because they were carriers of chromosomal abnormalities or had already had repeated miscarriages.

From those 41 couples, the team biopsied and sequenced 150 blastocysts. While 71 were identified as euploid, 25 had chromosomal aberrations, 40 had imbalanced structural aberrations, and 14 had both chromosomal and structural aberrations.

The sequencing test enabled the physician to choose only euploid blastocysts for implantation, Gao said.

... Separately, a team from Peking University is testing single-cell whole-genome sequencing using Xie's MALBAC technique, published in Science last year (IS 1/2/2013).

Fuchou Tang, an assistant professor at Peking University's Biodynamic Optical Imaging Center, said this week that his group is testing the technique on the 1st and 2nd polar bodies — by-products of the IVF process from which chromosomal numbers in the female pronucleus can be deduced.

The advantage of sequencing the polar bodies, as opposed to cells from the blastomere, is that there is no risk in harming a potentially viable embryo.

Tang's group has been collaborating with Xie's group, who presented at this year's Advances in Genome Biology and Technology meeting in Marco Island, Fla.

At the meeting, Xie said that in a pilot of six female donors, the technique could correctly infer embryo aneuploidy by sequencing to 0.1-fold depth (CSN 2/27/2013).

Since then, Tang's group has demonstrated that sequencing depth can be as low as 0.03-fold to accurately call aneuploidies, and he is now testing the technique to call point mutations that cause Mendelian disease.

... Aside from IVF applications, researchers are looking to single-cell sequencing to aid in cancer prognostics, diagnostics, and disease monitoring. Harvard's Xie has been using MALBAC to look at circulating tumor cells in lung cancer patients.

Circulating tumor cells are believed to be indicative of metastasis, which "accounts for 90 percent of cancer mortality," Xie said. "We need single-cell techniques to tackle this problem," particularly because cancer is so heterogeneous, and even more so after it metastasizes.

In a proof-of-concept study, Xie used MALBAC to do single-cell exome sequencing and in some cases whole-genome sequencing as well, of eight circulating tumor cells from one patient. He also sequenced the patients' primary and metastatic tumor and compared the mutational profiles from each. ...

Project Einstein

I met Jonathan Rothberg, a real pioneer in genetic sequencing technology, at Scifoo back in 2008 (see Gene machines). Jonathan's foundation is now backing an effort similar to the BGI Cognitive Genomics project. He may not remember, but we had a long conversation about this topic on the bus from the hotel to the Googleplex.

I've agreed to participate in Project Einstein (I am not worthy!) as a DNA donor, and I hope that our projects will someday share data and resources. Rothberg's attitude is typical of a true innovator: damn the critics, full speed ahead!

Nature: He founded two genetic-sequencing companies and sold them for hundreds of millions of dollars. He helped to sequence the genomes of a Neanderthal man and James Watson, who co-discovered DNA’s double helix. Now, entrepreneur Jonathan Rothberg has set his sights on another milestone: finding the genes that underlie mathematical genius.

Rothberg and physicist Max Tegmark, who is based at the Massachusetts Institute of Technology in Cambridge, have enrolled about 400 mathematicians and theoretical physicists from top-ranked US universities in a study dubbed ‘Project Einstein’. They plan to sequence the participants’ genomes using the Ion Torrent machine that Rothberg developed.

The team will be wading into a field fraught with controversy. Critics have assailed similar projects, such as one at the BGI (formerly the Beijing Genomics Institute) in Shenzhen, China, that is sequencing the genomes of 1,600 people identified as mathematically precocious children in the 1970s (see Nature 497, 297–299; 2013).

... Rothberg has long been interested in cognition. He is also in awe of the abilities of famous scientists. “Einstein said ‘the most incomprehensible thing about the Universe is that it is comprehensible’,” he says. “I’d love to find the genes that make the Universe comprehensible.”

There is precedent to the concept of sequencing extreme outliers in a population in the hunt for influential genes. Scientists have used the technique to sift for genes that influence medical conditions such as high blood pressure and bone loss. Some behavioural geneticists, such as Robert Plomin at King’s College London, who is involved with the BGI project, say that there is no reason that this same approach won’t work for maths ability. As much as two-thirds of a child’s mathematical aptitude seems to be influenced by genes (Y. Kovas et al. Psychol. Sci. 24, 2048–2056; 2013).

... The Rothberg Institute for Childhood Diseases, Rothberg’s private foundation based in Guilford, Connecticut, is the study’s sponsor. But Rothberg won’t say who is funding the project, which other geneticists estimate will cost at least US$1 million. Some speculate that Rothberg is funding it himself. In 2001, Fortune estimated his net worth to be $168 million, and that was before he sold the sequencing companies he founded — 454 Life Sciences and Ion Torrent, both based in Connecticut — for a combined total of $880 million.

Rothberg is adamant that the project is well worth the time and the money, whoever is paying for it. “This study may not work at all,” he says — before adding, quickly, that it “is not a crazy thing to do”. For a multimillionaire with time on his hands, that seems to be justification enough.

Let me repeat the scientific motivations for this type of project. The human brain is arguably the most complex object we know of in the universe. Yet, it is constructed from a blueprint containing less than a few gigabits of information. Unlocking the genetic architecture of cognition is one of the greatest challenges -- now feasible in the age of genomics that Rothberg and others helped bring into existence.

For a discussion of previous GWAS results on general cognition, and their implications for the prospects of studies like Project Einstein, see First GWAS hits for cognitive ability. For general background on the science, watch this video. Or read these: MIRI interview, FAQ.

Nabokov on teaching

Nabokov was professor of literature at Cornell from 1948-1959. The excerpt below is from a 1964 Playboy interview, reproduced at longform.org (a site I highly recommend).

Nabokov: I gave up teaching—that’s about all in the way of change. Mind you, I loved teaching, I loved Cornell, I loved composing and delivering my lectures on Russian writers and European great books. But around 60, and especially in winter, one begins to find hard the physical process of teaching, the getting up at a fixed hour every other morning, the struggle with the snow in the driveway, the march through long corridors to the classroom, the effort of drawing on the blackboard a map of James Joyce’s Dublin or the arrangement of the semi-sleeping car of the St. Petersburg-Moscow express in the early 1870s—without an understanding of which neither Ulysses nor Anna Karenina, respectively, makes sense. For some reason my most vivid memories concern examinations. Big amphitheater in Goldwin Smith. Exam from 8 a.m. to 10:30. About 150 students—unwashed, unshaven young males and reasonably well-groomed young females. A general sense of tedium and disaster. Half-past eight. Little coughs, the clearing of nervous throats, coming in clusters of sound, rustling of pages. Some of the martyrs plunged in meditation, their arms locked behind their heads. I meet a dull gaze directed at me, seeing in me with hope and hate the source of forbidden knowledge. Girl in glasses comes up to my desk to ask: “Professor Kafka, do you want us to say that…? Or do you want us to answer only the first part of the question?” The great fraternity of C-minus, backbone of the nation, steadily scribbling on. A rustic arising simultaneously, the majority turning a page in their bluebooks, good teamwork. The shaking of a cramped wrist, the failing ink, the deodorant that breaks down. When I catch eyes directed at me, they are forthwith raised to the ceiling in pious meditation. Windowpanes getting misty. Boys peeling off sweaters. Girls chewing gum in rapid cadence. Ten minutes, five, three, time’s up.

The first paragraph of Lolita, one of my favorites in all of literature:

Lolita, light of my life, fire of my loins. My sin, my soul. Lo-lee-ta: the tip of the tongue taking a trip of three steps down the palate to tap, at three, on the teeth. Lo. Lee. Ta.

Annals of psychometry: adult cognitive skills by country

You may have read (NYTimes; enjoy the 500 expert comments!) about the recent OECD study of adult skills, which showed Americans lagging behind most other advanced countries. The outcomes more or less recapitulate the PISA results, which are produced by the same organization, led by Andreas Schleicher (Without data, you are just another person with an opinion; Schleicher was trained in physics before moving into educational assessment). Studying adults rather than students makes generational differences apparent: young Koreans scored near the top in the world, whereas older Koreans, who grew up in a much poorer country, do less well. This was also true for Finns, but countries like the UK, US and Australia showed almost no difference between the young and old. It would appear that the Flynn Effect has abated for the current adult population in those countries.

The most striking findings in the report to someone already familiar with this kind of country level data, are that:

(I) Job growth in areas requiring superior ability is high, whereas there is almost no growth in other sectors (see figure below and figure 1.3 in the report). The biggest decline in jobs is in manufacturing-related areas, which are characterized by low, but not the lowest, scores. The number of menial jobs, characterized by the lowest scores, held their own. In other words, in developed countries jobs for factory workers declined while maids, gardeners and servants held their own... a grim future of income inequality based on cognitive ability?

(Difference between mean scores of skilled and "elementary" occupations is about 1 population SD -- Table A3.19. This is about the same as mean score difference between college graduates and those that did not reach upper secondary education! Table A3.9 (l))

(II) Scores on numeracy and literacy showed high correlations: 0.8 -- 0.9 in every country. This is much higher than, for instance, M and V correlation on the SAT.

Here are national score ranges for math skills:

This figure shows that Japanese high school graduates have better literacy skills than Italian college graduates. There is plenty more data like this. Interestingly the standard deviations in scores for Japan and Korea are on the low end compared to other countries.

Roughly every fifth Finn and Japanese reads at high levels (Level 4 or 5 on the Survey of Adult Skills). This means, for example, that they can perform multiple-step operations to integrate, interpret, or synthesise information from complex or lengthy texts that involve conditional and/or competing information; and they can make complex inferences and appropriately apply background knowledge as well as interpret or evaluate subtle truth claims or arguments. They are also good at numbers: they can analyse and engage in complex reasoning about quantities and data, statistics and chance, spatial relationships, change, proportions and formulae; perform tasks involving multiple steps and select appropriate problem-solving strategies and processes; and understand arguments and communicate well-reasoned explanations for answers or choices.

How does someone who can't do these things obtain a good college degree? If only 20% of Finns and Japanese can do these things (the largest fraction among all nations), what fraction of US students are ready for serious college work?

The Fate of Empires

John Bagot Glubb, a British officer in the first and second world war, and British Commander of the Arab Legion during the Arab-Israeli war of 1948, wrote a number of books including The Fate of Empires, which examines regularities in the rise and fall of 11 empires over 3000 years. The empires Glubb studied had a lifespan of about ten human generations, or two hundred and fifty years, despite changing factors such as technology. Glubb describes a pattern of growth and decline, with six stages: the Ages of Pioneers, Conquest, Commerce, Affluence, Intellect and Decadence. He pointedly avoided writing about India or China, focusing rather on middle and western Eurasia, stating that his knowledge was inadequate to the task.

Note that six stages in 10 generations means that significant change can occur over one or two generations -- a nation can pass from one stage to the next, as I believe we have in America during my lifetime.

The Fate of Empires

... There does not appear to be any doubt that money is the agent which causes the decline of this strong, brave and self-confident people. The decline in courage, enterprise and a sense of duty is, however, gradual. The first direction in which wealth injures the nation is a moral one. Money replaces honour and adventure as the objective of the best young men. Moreover, men do not normally seek to make money for their country or their community, but for themselves. Gradually, and almost imperceptibly, the Age of Affluence silences the voice of duty. The object of the young and the ambitious is no longer fame, honour or service, but cash. Education undergoes the same gradual transformation. No longer do schools aim at producing brave patriots ready to serve their country. [ Or to discover great things for all mankind! ] Parents and students alike seek the educational qualifications which will command the highest salaries.


The inadequacy of intellect

Perhaps the most dangerous by-product of the Age of Intellect is the unconscious growth of the idea that the human brain can solve the problems of the world. Even on the low level of practical affairs this is patently untrue. Any small human activity, the local bowling club or the ladies’ luncheon club, requires for its survival a measure of self-sacrifice and service on the part of the members. In a wider national sphere, the survival of the nation depends basically on the loyalty and self-sacrifice of the citizens. The impression that the situation can be saved by mental cleverness, without unselfishness or human self-dedication, can only lead to collapse.

Thus we see that the cultivation of the human intellect seems to be a magnificent ideal, but only on condition that it does not weaken unselfishness and human dedication to service. Yet this, judging by historical precedent, seems to be exactly what it does do. Perhaps it is not the intellectualism which destroys the spirit of self-sacrifice—the least we can say is that the two, intellectualism and the loss of a sense of duty, appear simultaneously in the life-story of the nation.  [  Correlation != Causation  :-)  The point is that the ages of intellectualism and decadence occur at a similar stage of development. ]

Indeed it often appears in individuals, that the head and the heart are natural rivals. The brilliant but cynical intellectual appears at the opposite end of the spectrum from the emotional self-sacrifice of the hero or the martyr. Yet there are times when the perhaps unsophisticated self-dedication of the hero is more essential than the sarcasms of the clever.


... Neither is decadence physical. The citizens of nations in decline are sometimes described as too physically emasculated to be able to bear hardship or make great efforts. This does not seem to be a true picture. Citizens of great nations in decadence are normally physically larger and stronger than those of their barbarian invaders ...

See also Duty, Honor, Country:

The unbelievers will say they are but words, but a slogan, but a flamboyant phrase. Every pedant, every demagogue, every cynic, every hypocrite, every troublemaker, and I am sorry to say, some others of an entirely different character, will try to downgrade them even to the extent of mockery and ridicule.

The 21st century American reality (the Age of Decadence):

"Yeah, I calculated the NPV, and, you know, it's just not worth it for me. I really believe in your project, though. And, I share your passion. Good luck."

Number sense and math ability

This is consistent with my experience as a parent and educator: my guess is that number sense is a cognitive module, at least somewhat distinct from general intelligence, and somewhat hardwired.

Number sense in infancy predicts mathematical abilities in childhood (PNAS)

Abstract: Human infants in the first year of life possess an intuitive sense of number. This preverbal number sense may serve as a developmental building block for the uniquely human capacity for mathematics. In support of this idea, several studies have demonstrated that nonverbal number sense is correlated with mathematical abilities in children and adults. However, there has been no direct evidence that infant numerical abilities are related to mathematical abilities later in childhood. Here, we provide evidence that preverbal number sense in infancy predicts mathematical abilities in preschool-aged children. Numerical preference scores at 6 months of age correlated with both standardized math test scores and nonsymbolic number comparison scores at 3.5 years of age, suggesting that preverbal number sense facilitates the acquisition of numerical symbols and mathematical abilities. This relationship held even after controlling for general intelligence, indicating that preverbal number sense imparts a unique contribution to mathematical ability. These results validate the many prior studies purporting to show number sense in infancy and support the hypothesis that mathematics is built upon an intuitive sense of number that predates language.

Halloween

My son wanted a mutant cyclops jack o lantern :-)

Did I mention my kids love Calvin and Hobbes?

Compressed sensing and genomes

For more discussion of our recent paper (The human genome as a compressed sensor), see this blog post by my collaborator Carson Chow and another on the machine learning blog Nuit Blanche. One of our main points in the paper is that the phase transition between the regimes of poor and good recovery of the L1 penalized algorithm (LASSO) is readily detectable, and that the scaling behavior of the phase boundary allows theoretical estimates for the necessary amount of data required for good performance at a given sparsity. Apparently, this reasoning has appeared before in the compressed sensing literature, and has been used to optimize hardware designs for sensors. In our case, the sensor is the human genome, and its statistical properties are fixed. Fortunately, we find that genotype matrices are in the same universality class as random matrices, which are good compressed sensors.

The black line in the figure below is the theoretical prediction (Donoho 2006) for the location of the phase boundary. The shading shows results from our simulations. The scale on the right is L2 (norm squared) error in the recovered effects vector compared to the actual effects.

Perhaps we are approaching a D-T moment in genomics ;-)

... a Donoho-Tao moment in the Radar community at the next CoSeRa meeting :-). As a reminder the Donoho-Tao moment was well put in this 2008 IPAM newsletter: .... It’s David Donoho [5] reportedly exclaiming [to] a panel of NSF folks “You’ve got Terry Tao (a Fields medalist [6]) talking to geoscientists, what do you want?” ....

In previous discussions I predicted that of order millions of phenotype-genotype pairs would be sufficient to extract the genetic architecture of complex traits like height or g. This estimate is based on two ingredients:

1. The sparsity of these traits is probably no greater than s ~ 10k (evidence for this comes from looking at genomic Hamming distance as a function of phenotype distance).

2. The compressed sensing results suggest that good recovery can be achieved above a data threshold of roughly n ~ 30 s (assuming 1E06 SNPs and additive heritability h2 = 0.5 or so).

Including an extra order of magnitude to be safe, this leads to n ~ millions.

Tap or Snap

The dreaded heel hook in action. Palhares has been banned from the UFC after this fight, for holding the submission too long. It doesn't actually seem that egregious to me in the video -- when the stakes are high, the fighter should not release the hold until instructed by the referee. See also Snap, Crackle, Pop.

Back in the day when grappling and BJJ were still fringe activities, I often had to travel to strange clubs to find training. It was intimidating to visit a new school where I didn't know anyone, even more so to spar with people who could easily injure me. The one submission I was most afraid of was the heel hook. The two serious injuries I sustained in years of training were from a straight armbar (juji gatame) and a heel hook, which sprained the tendons around my knee. The heel hook is much more effective on the street, where the opponent is likely to be wearing shoes and pants (escaping by pulling the leg out is much harder than in MMA), although there are also reasons not to pull guard in a street fight.

Palhares reportedly cried at the weigh-in after a hard cut to 170 lbs. In the past he's fought at 185 and probably walks around at 200 or so.

Bezos quotes

These quotes appear in an excerpt from a new biography by Brad Stone. Stone uncovers some new ground -- tracking down Bezos' long-lost biological father, who was unaware(!) that his son had become a billionaire e-commerce titan.

I had always heard that Bezos has a vulcan or hyper-rational management style. It's good to know he occasionally loses his temper like everyone else.

“Are you lazy or just incompetent?”

“I’m sorry, did I take my stupid pills today?”

“Do I need to go down and get the certificate that says I’m CEO of the company to get you to stop challenging me on this?”

“Are you trying to take credit for something you had nothing to do with?”

“If I hear that idea again, I’m gonna have to kill myself.”

“We need to apply some human intelligence to this problem.”

[After reviewing the annual plan from the supply chain team] “I guess supply chain isn’t doing anything interesting next year.”

[After reading a start-of-meeting memo] “This document was clearly written by the B team. Can someone get me the A team document? I don’t want to waste my time with the B team document.”

[After an engineer’s presentation] “Why are you wasting my life?”

... To the amazement and irritation of employees, Bezos’s criticisms are almost always on target. Bruce Jones, a former Amazon supply chain vice president, describes leading a five-engineer team figuring out ways to make the movement of workers in fulfillment centers more efficient. The group spent nine months on the task, then presented their work to Bezos. “We had beautiful documents, and everyone was really prepared,” Jones says. Bezos read the paper, said, “You’re all wrong,” stood up, and started writing on the whiteboard.

“He had no background in control theory, no background in operating systems,” Jones says. “He only had minimum experience in the distribution centers and never spent weeks and months out on the line.” But Bezos laid out his argument on the whiteboard, and “every stinking thing he put down was correct and true,” Jones says. “It would be easier to stomach if we could prove he was wrong, but we couldn’t. That was a typical interaction with Jeff. He had this unbelievable ability to be incredibly intelligent about things he had nothing to do with, and he was totally ruthless about communicating it.”

See also Bezos on the big brains ;-)

Jeff Bezos: Yeah. So, I went to Princeton primarily because I wanted to study physics, and it's such a fantastic place to study physics. Things went fairly well until I got to quantum mechanics and there were about 30 people in the class by that point and it was so hard for me. I just remember there was a point in this where I realized I'm never going to be a great physicist. There were three or four people in the class whose brains were so clearly wired differently to process these highly abstract concepts, so much more. I was doing well in terms of the grades I was getting, but for me it was laborious, hard work. And, for some of these truly gifted folks -- it was awe-inspiring for me to watch them because in a very easy, almost casual way, they could absorb concepts and solve problems that I would work 12 hours on, and it was a wonderful thing to behold.

"Abstract geniuses" like the ones Bezos encountered at Princeton might not have the common sense or practical inclination necessary to run an organization like Amazon, but on the other hand, perhaps a few do! 8^)  In any case, credit to Bezos for being so brutally honest and logical about his own abilities and limitations. Most people, when confronted by an obviously superior intellect (even if confined to some narrow subset of abilities), resort to comforting rationalizations: "I could understand quantum mechanics if I really wanted to. But I don't, so who cares!"

Global innovation and entrepreneurship

Photos from the World Summit on Innovation and Entrepreneurship, held at the Museum of the Moving Image, NYC.

It seems that every region in the world is trying to replicate the Silicon Valley model. There were delegations at this event from the EU, Norway, Jordan, Egypt, Japan, you name it. Lots of interesting NYC startups represented.

The NSA will note that I do a lot of blogging while traveling -- in particular while waiting for a flight  8^/

Creation, Myths and Twitter

Great article by Nick Bilton on the creation myth (and true story) behind Twitter. To see that luck plays an unimaginably huge role in life you just need to look carefully at the story behind any successful company or entrepreneur.

NYTimes: ... Soon, the question of a name came up. Williams jokingly suggested calling the project “Friendstalker,” which was ruled out as too creepy. Glass became obsessive, flipping through a physical dictionary, almost word by word, looking for the right name. One late afternoon, alone in his apartment, he reached over to his cellphone and turned it to silent, which caused it to vibrate. He quickly considered the name “Vibrate,” which he nixed, but it led him to the word “twitch.” He dismissed that too, but he continued through the “Tw” section of the dictionary: twist, twit, twitch, twitcher, twitchy . . . and then, there it was. He read the definition aloud. “The light chirping sound made by certain birds.” This is it, he thought. “Agitation or excitement; flutter.” Twitter.

... Whatever his reasons, Dorsey had recently met with Williams and threatened to quit if Glass wasn’t let go. And for Williams, the decision was easy. Dorsey had become the lead engineer on Twitter, and Glass’s personal problems were affecting his judgment. (For a while, portions of the company existed entirely on Glass’s I.B.M. laptop.) After conferring with the Odeo board, around 6 p.m. on Wednesday, July 26, 2006, Williams asked Glass to join him for a walk to South Park. Sitting on a green bench, Williams gave his old friend an ultimatum: six months’ severance and six months’ vesting of his Odeo stock, or he would be publicly fired. Williams said the decision was his alone.

... Williams and Dorsey started meeting for weekly dinners to discuss the problems, but one night Dorsey became defensive. “Do you want to be C.E.O.?” he said abruptly. Williams tried to evade the question, but eventually replied: “Yes, I want to be C.E.O. I have experience running a company, and that’s what Twitter needs right now.” ... told him that they were replacing him as C.E.O. with Williams. Dorsey sat before a bowl of uneaten yogurt and granola as he was offered stock, a $200,000 severance and a face-saving role as the company’s “silent” chairman. No one in the industry had to know that he was fired. (Investors would not want to be seen as pitting one founder against another anyway.) But Dorsey had no voting rights at the company. He was, essentially, out.

... Access to the tech blogosphere and press can help percolate a fledgling start-up into a multibillion-dollar business. But this access often relies on having a narrative — being an entrepreneur with just the right creation story. ... After he was stripped of his power at Twitter, Dorsey went on a media campaign to promote the idea that he and Williams had switched roles. He also began telling a more elaborate story about the founding of Twitter. In dozens of interviews, Dorsey completely erased Glass from any involvement in the genesis of the company. He changed his biography on Twitter to “inventor”; before long, he started to exclude Williams and Stone too.

... Without Williams and Stone influencing its development with the lessons they learned from Blogger, it still would not have taken off. Making it a company required Williams’s money, then Wilson, Sabet and Fenton’s and dozens of other investors, not to mention Costolo, who turned it into viable business, and 2,000 employees who helped shape it into one of the biggest social networks on the planet. Such is the case with every company in Silicon Valley, though you never hear it in their creation myth. Dorsey will make $400 million to $500 million when Twitter goes public. Glass stands to make about as much as Dorsey’s secretary at Square. ...

This is from a 2009 post Me and Twitter. I had met Williams at Foo Camp, probably in 2007. In the 2009 post I didn't mention that most of the conversation was about Odeo, Williams' podcasting startup from which Twitter sprang as an almost accidental creation. His description to me of how the Twitter idea originated was a bit different than what Bilton reports.

I met Twitter founder Evan Williams a few years ago, before Twitter was anywhere near a big thing. He told me about Blogger, which he sold to Google, and then the inevitable "So what are you working on now?" question came up.

He described Twitter to me, and two thoughts entered my mind. The first shows I am old, or out of touch, or have no feel for Web 2.0 consumer startups: "Who would use that?" I said to myself.

The second thought, which I actually verbalized, turns out to be a good question (still unanswered) and shows I may have VC potential: "How are you going to monetize that?" :-)

The human genome as a compressed sensor

Compressed sensing (see also here) is a method for efficient solution of underdetermined linear systems: y = Ax + noise , using a form of penalized regression (L1 penalization, or LASSO). In the context of genomics, y is the phenotype, A is a matrix of genotypes, x a vector of effect sizes, and the noise is due to nonlinear gene-gene interactions and the effect of the environment. (Note the figure above, which I found on the web, uses different notation than the discussion here and the paper below.)

Let p be the number of variables (i.e., genetic loci = dimensionality of x), s the sparsity (number of variables or loci with nonzero effect on the phenotype = nonzero entries in x) and n the number of measurements of the phenotype (i.e., the number of individuals in the sample = dimensionality of y). Then  A  is an  n x p  dimensional matrix. Traditional statistical thinking suggests that  n > p  is required to fully reconstruct the solution  x  (i.e., reconstruct the effect sizes of each of the loci). But recent theorems in compressed sensing show that  n > C s log p  is sufficient if the matrix A has the right properties (is a good compressed sensor). These theorems guarantee that the performance of a compressed sensor is nearly optimal -- within an overall constant of what is possible if an oracle were to reveal in advance which  s  loci out of  p  have nonzero effect. In fact, one expects a phase transition in the behavior of the method as  n  crosses a critical threshold given by the inequality. In the good phase, full recovery of  x  is possible.

In the paper below, available on arxiv, we show that

1. Matrices of human SNP genotypes are good compressed sensors and are in the universality class of random matrices. The phase behavior is controlled by scaling variables such as  rho = s/n  and our simulation results predict the sample size threshold for future genomic analyses.

2. In applications with real data the phase transition can be detected from the behavior of the algorithm as the amount of data  n  is varied. A priori knowledge of  s  is not required; in fact one deduces the value of  s  this way.

3.  For heritability h2 = 0.5 and p ~ 1E06 SNPs, the value of  C log p  is ~ 30. For example, a trait which is controlled by s = 10k loci would require a sample size of n ~ 300k individuals to determine the (linear) genetic architecture.

Application of compressed sensing to genome wide association studies and genomic selection          

http://arxiv.org/abs/1310.2264

Authors: Shashaank Vattikuti, James J. Lee, Stephen D. H. Hsu, Carson C. Chow
Categories: q-bio.GN
Comments: 27 pages, 4 figures; Supplementary Information 5 figures

We show that the signal-processing paradigm known as compressed sensing (CS)
is applicable to genome-wide association studies (GWAS) and genomic selection
(GS). The aim of GWAS is to isolate trait-associated loci, whereas GS attempts
to predict the phenotypic values of new individuals on the basis of training
data. CS addresses a problem common to both endeavors, namely that the number
of genotyped markers often greatly exceeds the sample size. We show using CS
methods and theory that all loci of nonzero effect can be identified (selected)
using an efficient algorithm, provided that they are sufficiently few in number
(sparse) relative to sample size. For heritability h2 = 1, there is a sharp
phase transition to complete selection as the sample size is increased. For
heritability values less than one, complete selection can still occur although
the transition is smoothed. The transition boundary is only weakly dependent on
the total number of genotyped markers. The crossing of a transition boundary
provides an objective means to determine when true effects are being recovered.
For h2 = 0.5, we find that a sample size that is thirty times the number
of nonzero loci is sufficient for good recovery.

Nobels for Higgs and Englert

Congratulations to Peter Higgs and François Englert on their Nobel prize. A bit of background from an earlier post How the Higgs boson became the Higgs boson:

IIRC, I met Peter Higgs in Erice in 1990. He was quite a nice fellow, but the story below by Steve Weinberg illustrates how capricious is the allocation of credit in science.

NYBooks: (Footnote 1) In his recent book, The Infinity Puzzle (Basic Books, 2011), Frank Close points out that a mistake of mine was in part responsible for the term “Higgs boson.” In my 1967 paper on the unification of weak and electromagnetic forces, I cited 1964 work by Peter Higgs and two other sets of theorists. This was because they had all explored the mathematics of symmetry-breaking in general theories with force-carrying particles, though they did not apply it to weak and electromagnetic forces. As known since 1961, a typical consequence of theories of symmetry-breaking is the appearance of new particles, as a sort of debris. A specific particle of this general class was predicted in my 1967 paper; this is the Higgs boson now being sought at the LHC.

As to my responsibility for the name “Higgs boson,” because of a mistake in reading the dates on these three earlier papers, I thought that the earliest was the one by Higgs, so in my 1967 paper I cited Higgs first, and have done so since then. Other physicists apparently have followed my lead. But as Close points out, the earliest paper of the three I cited was actually the one by Robert Brout and François Englert. In extenuation of my mistake, I should note that Higgs and Brout and Englert did their work independently and at about the same time, as also did the third group (Gerald Guralnik, C.R. Hagen, and Tom Kibble). But the name “Higgs boson” seems to have stuck.

[ Note that to Higgs' credit his is the only paper that clearly works out the properties of the excitation now known as the Higgs boson. ]

Jeffrey Goldstone showed (1961) that when rigid ("global") continuous symmetries are spontaneously broken by the vacuum (the vacuum configuration is not invariant under the symmetry), a massless boson necessarily results. This boson is the eponymous Goldstone boson: the particle excitation corresponding to small perturbations of the vacuum state in the direction of the symmetry. The natural next step is to ask what happens if the broken symmetry is a gauge (local) symmetry. This is the problem that Higgs et al. solved. But Goldstone had one of the first cracks at the problem. Indeed, Jeffrey deduced the existence of a massive excitation (i.e., the Higgs boson), but its physical reality was in question -- only apparent in certain "choices of gauge"; gauge theory was not then very well understood. According to legend, Sidney Coleman convinced Goldstone that the boson was only a gauge artifact. For years afterward Goldstone would say that Sidney, despite his obvious brilliance, was, when it really counted, always wrong!

I met Englert for the first time in 2008 at a workshop in Paris on the black hole information problem. Over coffee, he explained to me some mysterious comments 't Hooft had made in his talk. A real gentleman, and still very sharp.

A photo from the summer school in Erice, Sicily 1990. Higgs is in the blue socks and sandals, holding a glass of wine. I'm in a maroon shirt two rows back.

A portrait of Higgs in the physics department of the University of Edinburgh.

UK to sequence 100k genomes

Here's to the NHS! The plan is to complete this in 5 years.

Welcome to Genomics England

We are a new company set up by the Department of Health to help deliver the 100k Genome Project first announced by the Prime Minister, David Cameron, in December 2012.

This project will sequence the personal DNA code – known as a genome – of up to 100,000 patients over the next five years. This unrivalled knowledge will help doctors’ understanding, leading to better and earlier diagnosis and personalised care. Based on expert scientific advice, we will start by tackling cancer, rare diseases and infectious diseases.

The company will manage contracts for sequencing, data linkage and analysis, and set standards for patient consent.

Fuzzballs, black holes and firewalls

Yesterday Samir Mathur gave a colloquium here on the black hole information paradox. I've known Samir for many years; he was an assistant professor at MIT when I was a postdoc up the river. I've always found him to be a very precise and clear thinker.

On his web page there is a very simple introduction to the paradox. The initial presentation emphasizes the role of negative binding energy in black hole physics, which is related to the question of monsters: configurations in classical general relativity with more entropy than black holes of the same mass. (Slides.)

Here is a recent paper in which Samir discusses the black hole firewall problem and subadditivity of entropy.

Samir in action giving a more technical seminar earlier today:

Deemed Naughty by Nature

Nature editorial condemns research in cognitive genomics. A slight exaggeration, but consistent with the level of reporting in the accompanying article. (Click through and vote in their poll!)

I am quoted as follows:

After this summer's furore over Miller's interview [in Vice Magazine, of all places; Miller referred to Chinese embryo selection for "genius babies"], Hsu played down the potential for abuse. “There's a big gap between finding a few hits and finding thousands of hits — enough to predict the trait on the basis of the genotype — and we were never saying we were going to get to that point,” he says. But in 2011, before the uproar over the study, Hsu told Nature: “I'm 100% sure that a technology will eventually exist for people to evaluate their embryos or zygotes for quantitative traits, like height or intelligence. I don't see anything wrong with that.”

The first quote refers to the discovery power of our sample of 2000 gifted individuals. We would be quite happy to find even one genome-wide significant hit. The second quote refers to my prediction for what will be possible eventually (perhaps decades from now). Juxtaposing the quotes this way is deliberately misleading.

Beanbag genetics: blood pressure

As I wrote in an earlier post Beanbags and Causal Variants:

Not only do these results implicate common causal variants as the source of heritability in disease susceptibility, but they also suggest that gene-gene (epistasis) and gene-environment interactions are of limited impact. Both the genetic and environmental backgrounds for a particular allele vary across Eurasia, so replicability puts an upper limit on their influence. See also Epistasis vs Additivity.

How can it be? But what about the marvelous incomprehensible beautiful sacred complexity of Nature? But But But ...

In the blood pressure (BP) study cited below, the data include East and South Asians, African Americans and Europeans. The effect sizes of variants in one population are well correlated with effect sizes in other populations, despite changes in the genetic background (i.e. other genes) and environments with which they interact. This suggests the interaction effects are small.

Genome-wide Association Analysis of Blood-Pressure Traits in African-Ancestry Individuals Reveals Common Associated Genes in African and Non-African Populations

Abstract: ... We also demonstrate that validated EA BP GWAS loci, considered jointly, show significant effects in AA samples. Consequently, these findings suggest that BP loci might have universal effects across studied populations, demonstrating that multiethnic samples are an essential component in identifying, fine mapping, and understanding their trait variability.

(COGENT = African Americans, ICBP = European Americans)

Long live "beanbag genetics"! :-)

A Defense of Beanbag Genetics

JBS Haldane

My friend Professor Ernst Mayr, of Harvard University, in his recent book Animal Species and Evolution1, which I find admirable, though I disagree with quite a lot of it, has the following sentences on page 263.

The Mendelian was apt to compare the genetic contents of a population to a bag full of colored beans. Mutation was the exchange of one kind of bean for another. This conceptualization has been referred to as “beanbag genetics”. Work in population and developmental genetics has shown, however, that the thinking of beanbag genetics is in many ways quite misleading. To consider genes as independent units is meaningless from the physiological as well as the evolutionary viewpoint.  [Italics mine]

... In another place2 Mayr made a more specific challenge. He stated that Fisher, Wright, and I “have worked out an impressive mathematical theory of genetical variaion and evolutionary change. But what, precisely, has been the contribution of this mathematical school to evolutionary theory, if I may be permitted to ask such a provocative question?” “However,” he continued in the next paragraph, “I should perhaps leave it to Fisher, Wright, and Haldane to point out what they consider their major contributions.” ...

Now, in the first place I deny that the mathematical theory of population genetics is at all impressive, at least to a mathematician. On the contrary, Wright, Fisher, and I all made simplifying assumptions which allowed us to pose problems soluble by the elementary mathematics at our disposal, and even then did not always fully solve the simple problems we set ourselves. Our mathematics may impress zoologists but do not greatly impress mathematicians. Let me give a simple example. ...

See also Eric, why so gloomy?

Fisher's Fundamental Theorem of Natural Selection identifies additive variance as the main driver of evolutionary change in the limit where selection timescales are much longer than recombination (e.g., due to sexual reproduction) timescales. Thus it is reasonable to expect that most of the change in genus Homo [traits which have been under selection] over the last millions of years is encoded in a linear genetic architecture.

Too much homework?

I don't recall doing much homework until I was in high school. My grade school kids already have quite a bit to do, although I wouldn't say it has reached an excessive level. It is however clear that school is more serious than when I was a kid. In elementary school they didn't really know what to do with me so I spent a lot of time reading in the library. For some reason our library had all six volumes of Gibbon's History and Decline of the Roman Empire and Shirer's The Rise and Fall of the Third Reich. I could read pretty well but unfortunately no one tried to teach me any advanced math.

Atlantic Monthly: My Daughter’s Homework Is Killing Me

What happens when a father, alarmed by his 13-year-old daughter's nightly workload, tries to do her homework for a week

... whenever I bring up the homework issue with teachers or administrators, their response is that they are required by the state to cover a certain amount of material. There are standardized tests, and everyone—students, teachers, schools—is being evaluated on those tests. I’m not interested in the debates over teaching to the test or No Child Left Behind. What I am interested in is what my daughter is doing during those nightly hours between 8 o’clock and midnight, when she finally gets to bed. During the school week, she averages three to four hours of homework a night and six and a half hours of sleep.

... My daughter has the misfortune of living through a period of peak homework.

It turns out that there is no correlation between homework and achievement. According to a 2005 study by the Penn State professors Gerald K. LeTendre and David P. Baker, some of the countries that score higher than the U.S. on testing in the Trends in International Mathematics and Science Study—Japan and Denmark, for example—give less homework, while some of those scoring lower, including Thailand and Greece, assign more. Why pile on the homework if it doesn’t make even a testable difference, and in fact may be harmful?

“It’s a response to this whole globalized, competitive process,” says Richard Walker, a co-author of the book Reforming Homework. “You get parents demanding their children get more homework because their children are competing against the whole world.”

The irony is that some countries where the school systems are held up as models for our schools have been going in the opposite direction of the U.S., giving less homework and implementing narrower curricula built to encourage deeper understanding rather than broader coverage. ...

Dept. of Nobody Knows Anything: redshirting children

See also Expert Predictions and Medical Science?

New Yorker: ... Redshirting is the practice of holding a child back for an extra year before the start of kindergarten, named for the red jersey worn in intra-team scrimmages by college athletes kept out of competition for a year. It is increasingly prevalent among parents of would-be kindergartners. In 1968, four per cent of kindergarten students were six years old; by 1995, the number of redshirted first- and second-graders had grown to nine per cent. In 2008, it had risen to seventeen per cent. The original logic of the yearlong delay is rooted in athletics: athletes who are bigger and stronger tend to perform better, so why not bench the younger, smaller ones for a year? The logic was popularized in “Freakonomics,” in which the authors, Stephen J. Dubner and Steven D. Levitt, pointed out that élite soccer players were much more likely to have birthdays in the earliest months of the year—that is, they would have been the oldest in any group of students that used a January 1st cutoff for enrollment.

On the surface, redshirting seems to make sense in the academic realm, too. The capabilities of a child’s brain increase at a rapid pace; the difference between five-year-olds and six-year-olds is far greater than between twenty-five-year-olds and twenty-six-year-olds. An extra year can allow a child to excel relative to the younger students in the class. “Especially for boys, there is thought to be a relative-age effect that persists across sports and over time,” said Friedman. “Early investment of time and skill developments appears to have a more lasting impact.” Older students and athletes are often found in leadership positions—and who can doubt the popularity of the star quarterback relative to the gym-class weakling?

It’s this competitive logic, rather than genuine concern about a child’s developmental readiness, that drives redshirting. Many parents decide to redshirt their children not because they seem particularly immature or young but because they hope that the extra year will give them a boost relative to their peers. In light of modern competitive demands, why wouldn’t you want your child to have that edge? The psychologist Betsy Sparrow calls it “gaming the system”—and the data on who chooses to redshirt bears out that classification: the people most likely to redshirt their children are those who can most afford to do so—that is, the white and the wealthy. Families in the highest socioeconomic quintile are thirty-six per cent more likely to redshirt their children than those in the lowest, and while close to six per cent of white children are redshirted, the figure falls to two per cent for Hispanic children, and less than one per cent for their black peers.

The data, however, belies this assumption. While earlier studies have argued that redshirted children do better both socially and academically—citing data on school evaluations, leadership positions, and test scores—more recent analyses suggest that the opposite may well be the case: the youngest kids, who barely make the age cutoff but are enrolled anyway, ultimately end up on top—not their older classmates. When a group of economists followed Norwegian children born between 1962 and 1988, until the youngest turned eighteen, in 2006, they found that, at age eighteen, children who started school a year later had I.Q. scores that were significantly lower than their younger counterparts. Their earnings also suffered: through age thirty, men who started school later earned less. A separate study, of the entire Swedish population born between 1935 and 1984, came to a similar conclusion: in the course of the life of a typical Swede, starting school later translated to reduced over-all earnings. In a 2008 study at Harvard University, researchers found that, within the U.S., increased rates of redshirting were leading to equally worrisome patterns. The delayed age of entry, the authors argued, resulted in academic stagnation: it decreased completion rates for both high-school and college students, increased the gender gap in graduation rates (men fell behind women), and intensified socioeconomic differences.

As it turns out, the benefits of being older and more mature may not be as important as the benefits of being younger than your classmates. In 2007, the economists Elizabeth Cascio and Diane Whitmore Schanzenbach decided to analyze the data of Tennessee’s Project STAR—an experiment originally designed to test the effects of classroom size on learning—with a different set of considerations: How would the relative class composition affect student performance? Their approach differed from most studies of redshirting in one crucial way: the students had been assigned totally randomly to their kindergarten classrooms, with no option for parents to lobby for, say, a different teacher, a different school, or a class in which the child would have some other perceived or actual relative advantage. This led to true experimental variation in relative age and maturity. That is, the same student could be relatively younger in one class, but relatively older in another, depending on his initial class assignment. The researchers discovered that relatively more mature students didn’t have an academic edge; instead, when they looked at their progress at the end of kindergarten, and, later, when they reached middle school, they were worse off in multiple respects. Not only did they score significantly lower on achievement tests—both in kindergarten and middle school—they were also more likely to have been kept back a year by the time they reached middle school, and were less likely to take college-entrance exams. The less mature students, on the other hand, experienced positive effects from being in a relatively more mature environment: in striving to catch up with their peers, they ended up surpassing them. ...

Few researchers would dispute that, in the immediate term, being relatively bigger, quicker, smarter, and stronger is a good thing. Repeatedly, the studies have found exactly that—older kindergarten students perform better on tests, receive better teacher evaluations, and do better socially. But then, something happens: after that early boost, their performance takes a nosedive. By the time they get to eighth grade, any disparity has largely evened out—and, by college, younger students repeatedly outperform older ones in any given year.

You must visit us in Sabah

Biotechnology meeting with Malaysian Prime Minister.

New Mission Bay campus of UCSF.

Espresso in Berkeley.

View from Treasure Island.

Nobel prizes by undergraduate institution

[ See this 2015 post for updated statistics, normalized to size of school population. ]

Commenter LY linked to this CollegeConfidential analysis in an earlier thread. I can't vouch for its accuracy but I find it quite interesting. Keep in mind this is a lagging indicator with relatively low statistics. Nevertheless a simple analysis of population-normalized (i.e., per individual) probability of great alumni accomplishment (e.g., as calculated below) would show very strong correlation with cognitive ability of student body over time. Must be a coincidence -- we all know those tests don't measure anything real  ;-)

Many throw around Nobel Prize affiliations as proof of quality of education. So I decided to compile a list of colleges and universities according to where Nobel Prize winners completed their undergraduate education.

First, the ground rules:

1. I only included winners of the prize for Chemistry, Economics, Literature, Medicine and Physics. I did not include winners of the Peace prize.

2. I decided to add Fields Medalists since their is no Prize for Mathematics and the Fields medal is extremely prestigious.

Most of the results were expected, although I was surprised at how few prize winners had graduated from several elite universities (especially Brown, Duke, Johns Hopkins, Michigan, Northwestern, Princeton and Stanford).

The conclusion, I think most will agree after seeing the numbers, is that one cannot rely on the production of Nobel Prize winners as an indicator of quality of undergraduate education because only 7 universities have produced more than 5!

The summarized list of universities according to the number of Nobel Prize winners (undergraduate degree only):

Harvard University: 21
Columbia University: 16
University of Chicago: 13
Massachusetts Institute of Technology: 11
University of California-Berkeley: 11
Yale University: 11
California Institute of Technology: 8
Cornell University: 5
Swarthmore College: 5
University of California-Los Angeles: 5
University of Illinois-Urbana Champaign: 5
University of Michigan-Ann Arbor: 5
Amherst College: 4
Carnegie Mellon University: 3
Case Western Reserve University: 3
Dartmouth College: 3
Princeton University: 3
University of Pennsylvania: 3
University of Rochester: 3
University of Washington: 3
University of Wisconsin-Madison 3
Brandeis University: 2
Haverford College: 2
Johns Hopkins University: 2
Oberlin College: 2
Purdue University-West Lafayette: 2
Rice University: 2
Rutgers University: 2
Stanford University: 2
University of Minnesota-Twin Cities: 2
Williams College: 2
Brown University: 1
College of the Holy Cross: 1
Georgia Institute of Technology: 1
Grinnell College: 1
Hamilton College: 1
Michigan state University: 1
New York University: 1
Northwestern University: 1
Ohio State University: 1

Further down the thread on CC, someone notes

... The most impressive is Cal Tech. 120 years old, fewer than 1000 undergraduates, and 8 prizes. My guess is that they have by far the highest Nobel / alumnus ratio.  

My graduating class was 186 kids! See also The Normaliens.

The secret that both Caltech and Ecole Normale understand is simple: sample as hard as possible from the right tail of the g and conscientiousness distribution (that's about all you can reliably measure; maybe throw in some maverick wacky personality or "I looove science!" for good measure), hire the best professors you can, and cross your fingers... Note this methodology optimizes the school's contributions to humanity, but sadly not its endowment. See Defining Merit.

If I had time I'd make a slight adjustment by removing Literature and Economics from the prize list (sorry!), and perhaps add the Turing Prize or something similar.