All across the country, applicants to graduate and professional schools have been receiving fat letters of acceptance or thin letters of rejection.
They have a right to feel nervous. They’ve sweated through college and through rigorous standardized exams, which they hope will open the door to their chosen professions. But the prestigious postgrad programs are ruthless about selecting the best candidates (at least if they are white or Asian). So, applicants obsess over whether their 165 LSATK-12 education or 680 GMAT is good enough to get in.
But, paradoxically, professors at the top schools seldom preach what they practice when it comes to K-12 education or immigration. They are fiercely selectionist about whom they let in to their institutions. Yet they lecture American citizens about how we should be lax about whom we let in to our country.
There is much that can be learned from the study of average test scores from the major postgrad exams. The idiosyncratic scoring systems do make them seem impenetrable to outsiders, but fortunately, they are all graded on the bell curve, so I’ve come up with a handy table that makes them easy to understand.
I’ve accumulated recent data on the average scores by race for five exams: the GRE for grad school, the LSAT for law school, the MCAT for medical school, the GMAT for business school, and the DAT for dental school.
To make all the numbers comprehensible, I’ve converted them to show where the mean for each race would fall in percentile terms relative to the distribution of scores among non-Hispanic white Americans. Most of us have some sense of what the distribution of talent is among whites—political correctness doesn’t demand we avert our eyes when it comes to whites—so I’ll use whites as benchmarks:
|Mean Score as Percentile of White Distribution|
Thus, for example, on the Graduate Management Admission Test (GMAT), the gatekeeper for the M.B.A. degree, the mean score for whites falls, by definition, at the 50th percentile of the white distribution of scores. The mean score for black test-takers would rank at the 13th percentile among whites. Asians average a little better than the typical white, scoring at the 55th percentile.As you know, I like social statistics the way Bill James likes baseball statistics.
Most of these tests break out separate nationalities among Hispanics. Thus, my table has columns both for “Total Hispanics” (27th percentile on the GMAT) and “Mexican-Americans” (24th percentile). In the 2000 Census, Mexicans made up 58 percent of the Total Hispanic population.
I noticed that the New York Times ran another long, intelligent article about baseball statistics today. How much brainpower does America devote to solid thinking about baseball statistics compared to social statistics?
For example, here's a featured article in the Washington Post today that would been ripped to shreds for obvious methodological flaws before ever being published in a baseball statistics journal:
If they had given the working memory IQ subtest first at age 9 and then again at age 17 and shown that those under more stress had seen bigger IQ declines, then it would be suggestive that stress might lower IQ. But by waiting until age 17 to give the IQ subtest for the first time, the study is of almost no use. How do we know that individuals who are exposed to a lot of stress because they and/or their families made a lot of stupid decisions are stupid because of the stress or were they exposed to a lot of stress because they were always stupid?
Now, research is providing what could be crucial clues to explain how childhood poverty translates into dimmer chances of success: Chronic stress from growing up poor appears to have a direct impact on the brain, leaving children with impairment in at least one key area -- working memory.
"There's been lots of evidence that low-income families are under tremendous amounts of stress, and we know that stress has many implications," said Gary W. Evans, a professor of human ecology at Cornell University in Ithaca, N.Y., who led the research. "What this data raises is the possibility that it's also related to cognitive development."
With the economic crisis threatening to plunge more children into poverty, other researchers said the work offers insight into how poverty affects long-term achievement and underscores the potential ramifications of chronic stress early in life.
"This is a significant advance," said Bruce S. McEwen, who heads the laboratory of neuroendocrinology at Rockefeller University in New York. "It's part of a growing pattern of understanding how early life experiences can have an influence on the brain and the body."
Previous research into the possible causes of the achievement gap between poor and well-off children has focused on genetic factors that influence intelligence, on environmental exposure to toxins such as lead, and on the idea that disadvantaged children tend to grow up with less intellectual stimulation.
"People have hypothesized both genetic and environmental factors play a role in why poor children don't do as well in school," said Martha Farah, director of the center for cognitive neuroscience at the University of Pennsylvania. "Experiential factors can include things like having fewer trips to museums, having fewer toys, having parents who don't have as much time or energy to engage with them intellectually -- to read to them or talk to them."
But Evans, who has been gathering detailed data about 195 children from households above and below the poverty line for 14 years, decided to examine whether chronic stress might also be playing a role.
"We know low-socioeconomic-status families are under a lot of stress -- all kinds of stress. When you are poor, when it rains it pours. You may have housing problems. You may have more conflict in the family. There's a lot more pressure in paying the bills. You'll probably end up moving more often. There's a lot more demands on low-income families. We know that produces stress in families, including on the children," Evans said.
For the new study, Evans and a colleague rated the level of stress each child experienced using a scale known as "allostatic load." The score was based on the results of tests the children were given when they were ages 9 and 13 to measure their levels of the stress hormones cortisol, epinephrine and norepinephrine, as well as their blood pressure and body mass index.
"These are all physiological indicators of stress," said Evans, whose findings were published online last week by the Proceedings of the National Academy of Sciences. "The basic idea is this allows you to look at dysregulation resulting from stress across multiple physiological systems."
The subjects also underwent tests at age 17 to measure their working memory, which is the ability to remember information in the short term. Working memory is crucial for everyday activities as well as for forming long-term memories.
"It's critical for learning," Evans said. "If you don't have good working memory, you can't do things like hold a phone number in your head or develop a vocabulary."
When the researchers analyzed the relationships among how long the children lived in poverty, their allostatic load and their later working memory, they found a clear relationship: The longer they lived in poverty, the higher their allostatic load and the lower they tended to score on working-memory tests. Those who spent their entire childhood in poverty scored about 20 percent lower on working memory than those who were never poor, Evans said.
"The greater proportion of your childhood that your family spent in poverty, the poorer your working memory, and that link is largely explained by this chronic physiologic stress," Evans said. "We put these things together and can say the reason we get this link between poverty and deficits in working memory is this chronic elevated stress."
And, are we so sure that upper middle class Korean families are all that stress free when the scion brings home a 700 instead of an 800 on his SAT? Judging from Portnoy's Complaint, I would say that Philip Roth did not grow up in a low stress environment, but he seems pretty smart. I volunteer to lead an expedition to a low stress culture, such as Maui, and tabulate all the brilliant intellectual achievements the local kids have come up with.
By the way, why did they just use one IQ subtest, on working memory, instead of an entire IQ test? First, it's probably quicker and easier, especially for giving it to kids who speak exotic languages. Second, by just giving one IQ subtest, you don't have to mention the dread letters IQ. Third, blacks do relatively better on average on working memory than most other subtests -- Jensen says the white-black gap is only half a standard deviation on working memory.
Also, blood tests of hormones tied to emotions can give very different responses depending upon the current emotional state of the patient. For example, a medical clinic with people in lab coats walking around holding clipboards and discussing things in muted tones would be a fairly nonstressful environment for me (assuming I'm just being asked to give a sample for a scientific study, not for a personal diagnosis). But for some homeboy, well, it seems like a long way from home.