Friday, May 25, 2012

The common app

 A short science fiction story by one of my collaborators :-)

The common app 
Robert Scherrer   Nature 485, 540 (24 May 2012)  
... Vanessa set her drink down on the bar. “Eastern Southwest State's a perfectly good school, Anne. And if Larry doesn't get in, they have a northern branch campus that would certainly take him.” Vanessa glanced over at George. “Our Chester will be going to Harvard, just like his father.” 
“But what if he doesn't get in?” asked Anne. 
Vanessa shot Anne a poisonous look. “Oh, we've made alternate plans. Princeton is an excellent back-up choice.” Vanessa smoothed back her peroxide hair. “Now, George, it's almost time. Be a sweetheart and turn on the wall screen.” 
... But back in the old days, you had to fill out an application for college — not like now, where you just send in a cheek swab. ...

How long before this comes true?


Fred__R said...

Wouldn't they have already got their kids genes processed before deciding where to apply?

David Coughlin said...

It begs an empirical question.  How does capacity scale with scarcity?  I'm curious [and I think that it must happen], at which echelon does the task of managing subordinates become overwhelming?

tractal said...

Your tone seems awfully cheeky. If this ever happened (and it will probably happen, at least in substance) it would bring unimaginable social despair and who knows what else. Real social mobility may be low, but the imagined possibility of social mobility is solace enough to keep the thing together. Take that away and life is pretty grim, for the losers and winners both. It really isn't a :)ing matter.

Although I guess the point of your post to wonder at how far this could go. Isn't G only 70-80% heritable?

Robert Sykes said...

What does the cheek swab imply? Does it mean that race and sex are the only admission criteria?

MtMoru said...

The author must think that all psychological traits are determined at conception and that their value is independent of the culture. So far as there is any such thing as a psychological trait twin studies have proved this to be false. Maybe he doesn't have the cognitive ability to interpret them?

US unis and colleges do not select as strongly for IQ as those in other countries. But legacies are only a small part of the reason. The overwhelming reason is that grades and extracurriculars do not correlate well with IQ especially at America's inexcusably shitty public high schools.

US graduate schools are similarly much less selective for IQ than those of other countries. The reason again is that the quality of an UG degree in the US is measured by gpa and the rank of the UG institution rather than by cummulative exams as it is measured in all other countries.

MtMoru said...

"Isn't G only 70-80% heritable?"

Not even that. The best study of IQ heritability was done on British twins. It is the best because despite its small size the range of environments in the UK is much greater than those in Minnesota or Colorado or whatever. There the heritability was < .7.

And of course US unis would have to drop all other criteria. I would have qualified with a cheek swab if SAT scores were so heritable, but I had no chance given my high school and background.

Father BA Harvard '66. His father BA, PhD Princeton 193?.

MtMoru said...

It implies that the author has dirty underwear and narrow shoulders.

Even among the cognitive elite the understanding and even the ability to understand what the heritability figure really means is limited. So what heritability there is is exaggerated and simplified to "social rank reflects like a well polished mirror the innate ability of the individual".

This is ridiculous when made explicit, but it rarely is. It is a very large part of ideology today.

The syllogism:

IQ is heritable.
Rank correlates with IQ.
Therefore, rank is heritable.

But it should be:

IQ is imperfectly heritable within the narrow range of adoptive homes within a certain region, culture, society, etc.
Rank correlates with IQ to some extent.
Therefore, rank is heritable a little bit.

Obviously the first syllogism is much easier to think.

MtMoru said...

The above google results should be enough for those with sufficient cognitive ability :) to realize that twin studies showing high heritability for IQ are close to meaningless.

Coronary artery disease is still the rich world's biggest killer, but the Japanese are immune despite having high bp. Is it "genetic"? I don't know if it's true but here ya go:

Environment varies so much between Japan and the US that risk of CVD quadruples, but it doesn't vary at all as far as gene environment interactions and IQ? Twins separated at birth one raised in Japan the other in Stockton CA will at 18 have the same IQ score within 5 pts? Really?

Behavioral genetics and the genetics of the diseases of civilization are for people best described as products.

Christopher Chang said...

This is the unrealistic premise of Gattaca as well.

If you want to estimate a child's adult height, you look at their current age and height, you don't just look at family history (though you should take the latter into account as well).

LaurentMelchiorTellier said...

There's a key difference between this and Gattaca, though. 

Gattaca uses genetic profiles to perform negative screening (high probability of heart failure = unacceptably risky candidate for interplanetary spacecraft pilot, given stable-hearted alternative candidates). 

The above fiction uses genetic profiles to perform positive screening (high probability of hard working brain = hard working brain, without testing later to see how the probabilities pan out).

IMHO, the difference between these two is the difference between realistic scifi and bad scifi.
The Gattaca program has good reason to minimize risk of catastrophe, based on raw probabilities.
But why should Harvard and Yale accept students blindly by stochastic DNA predictions, when they could accept students based on both DNA, and later measurements of how the brain turned out? It's not really the same thing.

LaurentMelchiorTellier said...

That's true. In Gattaca, they did both, though. In two stages. 

In the movie, Vincent (posing as Jerome) uses the Jerome genetic identity to pass the first round of negative screening, and get access to the candidate group. Presumably, this was in addition to the rest of "Jeromes" CV.

He is then positively selected as a pilot, from among a large pool of negative screened candidates, over the course of the film, due to his performance. "Jerome" was chosen for "not a single mistake in a million keystrokes", excellent health, and not having murdered anyone. Also, the doctor liked his dong.

Not everyone in the negatively screened Gattaca pool gets to be a pilot in the end. Uma Thurman eventually gets screened out - albeit due to further, negative screening on her genetic risk factors.

Richard Seiter said...

That is a strong assertion.  Can you supply a reference please?

Richard Seiter said...

For pre-implantation selection of embryos my recollection is that Gattaca presented the outcomes as much more well-defined than would be justified by anything other than an extremely (unrealistically IMHO) high heritability.  I agree using genetic tests as a sole criterion for personnel selection is unrealistic as you and others have mentioned.  How realistic would it be to use genetic testing as an additional criterion though?  For college admissions do you think a genetic test could add significant predictive power someday?

LaurentMelchiorTellier said...

It's the difference between personnel selection and personnel DE-selection.
What we above call positive and negative selection, respectively.
It's possible to both select FOR skill, and to select AGAINST probability (without certainty) of heart attack in the middle of the mission. The movie makes it clear that the probability of Vincent getting a heart attack does not mean that he gets a heart attack. Maybe his heart attack risk is just 5%, this is never stated - only the risk of having a "disorder". But I don't see anything questionable about the Gattaca program nonetheless finding this to be an excellent reason for de-selecting him as personnel (negative screening), when they have such a glut of stable-hearted altnerative candidates.

Richard Seiter said...

I might need to rewatch the movie (it has been a while) to get my details correct, but IIRC they implied that the pre-implantation selection gave a very high probability of eliminating myopia.  I believe there is enough nurture component to myopia that this was not realistic (I believe it would be possible to reduce the incidence of myopia, but unlikely to reduce it to the degree they assert).  Given that myopia is an observable trait I don't think knowing a genetic estimate of its probability is useful given that you can just observe it directly.  This reasoning does not apply as clearly to heart disease, but I wonder how well we be able to predict heart disease from observed physical characteristics (e.g. measured thickness of the carotid intima) in the time frame where we will have enough knowledge of genetics to make Gattaca type genetic judgments.  I definitely agree with your point about the rationality of negative screening when there is a glut of candidates.
I agree about the possibility of both positive and negative selection in the Gattaca context (though as previously noted I don't think using a genetic test as the sole criterion is realistic, and question its value when in conflict with observed data).  I am curious about the applicability in other areas (hence my college admissions question.  I think a genetic test could add some value there, but can not begin to estimate how much).

The place where I think genetic screening gets incredibly complicated is where traits can not be cleanly divided into good/bad categories.  A good example of this is Hans Eysenck's idea that creativity is a product of the (uncommon) combination of high ego-strength and high psychopathology (see figure on page 122 of and surrounding text).  If this sort of thing is true (another, simpler example would be the value of a recessive sickle cell gene in a malaria prone area) then I think making screening judgments has the potential to become very complex (and very position dependent--like psychological screening today).

I understand the notion of acceptable error, but for embryo selection I think people tend to underestimate the possibility of unintended consequences.  I also think if people resulting from embryo selection have the idea that they are invulnerable to various things there will be much potential to engage in behavior that will demonstrate just how much effect nurture has ;-)

dwbudd said...

I just wish that the producers had done a DNA test prior to casting.  Maybe - just maybe - a film with a better actor than Ethan Hawke in the lead would have been the result.

David Coughlin said...

 I should add some clarification.  As I read this excerpt, I think about the social ramifications.  The social strata are not necessarily self-organizing [as is clear in the excerpt], so there must be some relationship between the social orders, enforced by one level's superiority over the lower.  Because of relative scarcity, at some stratum, the task of 'managing' the strategic relationship overwhelms it.  If you were going to make a WAG, where would you say that is?

LaurentMelchiorTellier said...

Sure, any human has a certain probability of developing myopia. I do think it's fair to say that optimized genetics and optimized nurture might make it highly unlikely, though. Even so, In Vincents case, he did not have to hide his myopia because it would give him away as an invalid, but because it would give him away as not being Jerome. Likewise, I agree with your point about measuring gross heart anatomy - I think they just thought it would be easier to do a sound recording. They're not even thinking of checking for Vincents, they're measuring and comparing Jeromes.

As for unintended consequences of screening... the whole movie is sort of about them missing out on a heck of a pilot. ^_^ I figure they'd be somewhat cognizant of this risk, but simply at peace with it. If the crew of Vincents ship smashes into Titan as he goes into cardiac arrest, their prejudice might be proven correct. Then again, maybe he'll never make a single, mistaken keystroke, and win his gold medal.

tractal said...

Yeah, the plot is a typical intellectual cheap shot. The test ended up not mattering because its an imperfect predictor and this one time a guy who would have been excluded by the test ended up doing really well. Seen that argument  before somewhere... I liked the movie though, but it would've been better had they manned up and had the guy have a heart attack midflight. 

Blog Archive