May 12, 2013

How do Hispanics score on grad school admissions tests?

Here are a couple of useful tables from my 2009 VDARE article on the results of admissions tests mandated by graduate and professional schools (the Graduate Record Exam, Law School Admission Test, Medical College Admission Test, Graduate Management Admission Test, and Dental Admission Test). 

The worthy organizations that administer these tests issue huge reports most years on how applicants have done by race, ethnicity, sex, and so forth. I took the most recent reports available in 2009, covering years from 2005 up to 2008). To make the median scores comprehensible and comparable, they are all displayed as percentiles vs. the white distribution. 
    Mean Score as Percentile of White Distribution
Test Degree White Black Asian Tot Hisp Mex-Am
GMAT M.B.A. 50% 13% 55% 27% 24%
GRE-Verbal Ph.D./M.A. 50% 18% 47% 29% 28%
GRE-Quant Ph.D./M.A. 50% 14% 66% 29% 28%
LSAT J.D. 50% 12% 47% 19% 29%
MCAT-Verbal M.D. 50% 10% 36% 19% 21%
MCAT-Phys Sci M.D. 50% 14% 61% 24% 25%
MCAT-Biol Sci M.D. 50% 10% 54% 24% 25%
DAT D.D.S. 50% 16% 60% 27% NA
For example, on the GMAT, the median white applicant scores, by definition, at the 50th percentile of the white distribution, while the median black applicant scores at the 13th percentile, Asians at the 55th percentile, all Hispanics at the 27th percentile, and (within Total Hispanics, Mexican Americans at the 24th percentile.

As you'll notice, blacks and Hispanics don't do all that well on these tests, which is why they get affirmative action.

As I wrote in 2009:
Note that language is a surprisingly small problem for Hispanics—they score no worse on the GRE Verbal subtest than on the GRE Quantitative, and only moderately worse on Verbal portion of the MCAT. Why? Because Hispanics who have problems with English generally don't finish college, or even high school.

Here's a second table showing what percentage of each ethnic group, relative to whites, take each test:
White Black Asian Tot Hisp Mexican
% of 20-24 ages 61.5% 15.5% 4.9% 17.5% 10.2%
White Shr Share of Test-Takers v. White Share
GMAT 68.7%     100 49205 29             18
GRE 74.0%     100 47  93 29             21
LSAT 66.1%     100 62 154 47             15
MCAT 57.0%     100 51 434 44             26
DAT 62.0%     100 44 489 36  NA
This table helps you evaluate the first table because it shows how hard the bottom of the barrel is being scraped for each group for each test. What percentage of each cohort of 20-24 year olds takes each test? Whites are set, by definition, at 100. For example, blacks are 49 percent as likely to take the GMAT as whites, Asians 205%, Total Hispanics 29%, and Mexican-Americans only 18%.

Mexican-Americans are no more than half as likely as blacks to take one of these tests. As I wrote in 2009:
If you are wondering why America's white elites aren't more worried about their kids facing competition from the huge number of Mexican immigrants they've let in, this educational indolence is one answer—at the highest levels of American society, Mexican-Americans just aren't much competition.

You can read the whole article here.


Anonymous said...

If you disagree with a lib, he/she says, "you HATE me, so I can hate YOU."

Anonymous said...

The part about Hispanics doing fine on the GRE verbal section and struggling on the quantitative section fit my experience to a tee. I scored in the 97th percentile on the verbal section and only in the 56th percentile on the quantitative section.

Anonymous said...

DR said...

At the high end "hispanic" gets seriously skewed by people that aren't exactly your every-day off the street hispanics. The only hispanics in my ivy league dorm were whites from Mexico, South America and Spain. Many of these "hispanics" had blonde or red hair and were much paler than me. Most were fluent in English and came from very wealthy families.

bjdubbs said...

"The part about Hispanics doing fine on the GRE verbal section and struggling on the quantitative section"

that's not what was said. in fact I'd bet the whole population of Hispanics is slightly better at math than verbal (Amerindians are to some degree Asian). rather it's the case that Hispanics who are good at verbal self-select for these tests. perhaps Hispanic academic underperformance vs. test scores is explained by poor verbal, ie being good but not great at math is no advangtage, because a 550 on the SAT math isn't going to cut it in an engineering dept., whereas 550 verbal will suffice in the english dept.

Anonymous said...

Asians specialize in the medical professions and tend to avoid grad school

Looking at the chart you put up, I don't see that Asians avoid grad school. At 93, it's not so far off from the White rate, especially when you take into account that it's also not popular with Blacks or Hispanics.

Anonymous said...

Looking at the chart you put up, I don't see that Asians avoid grad school. At 93, it's not so far off from the White rate, especially when you take into account that it's also not popular with Blacks or Hispanics.

Steve, do you have the total number of test takers for each test? That would provide more context. I would imagine that GRE would have the greatest number and DAT the smallest.

Anonymous said...

But this combination of average score and sample size data suggests that blacks tend to have inflated ambitions, especially compared to the under-ambitious Mexican American population.

Considering blacks used to be able to fly and built the pyramids by levitation, I don't think they have inflated ambitions at all.

Recently arrived Mexicans, having come from a socially immobile society, mostly operate on that paradigm.

LemmusLemmus said...

Combining the information from the two tables, and assuming normally distributed abilities in the population, couldn't you estimate hypothetical test results in the populations (white, black, etc.) as a whole? I'm not interested enough in the topic to do it myself.

AMac said...

> If you are wondering why America's white elites aren't more worried about their kids facing competition from the huge number of Mexican immigrants they've let in, this educational indolence is one answer... Mexican-Americans [don't offer much high-level] competition.

That helps with specifics about elite liberals (and conservatives) throwing Jason Richwine under the bus.

"Hispanics aren't as smart as as PLU, on average? Duh! Exhibiting hypocrisy or cognitive dissonance on this subject is virtuous. Explicitly noticing the Hispanic/White gap is evil. Comes too close to implying self-serving and self-dealing behavior on my part. I'm offended at the very notion!"

Explicit quotas for African Americans and Native Americans to redress past wrongs would be much healthier than the current, corrosive embodiment of Affirmative Action. Of course, from the elites' perspective, Gaming The System is a feature, not a bug.

Pat Boyle said...

This is just about what you would expect from simple IQ scores. If blacks score about one SD below whites, that would put them at the 15.87%tile.

These figures also clearly show that the Asian mental capabilities are more quantitative and less verbal than the white.

The question remains is verbal intelligence more important or less important than quantitative intelligence? Obviously the field of endeavor matters. Verbal ability is more important in Law and less important in engineering.

But what about in general? La Griff du Lion published an analysis of the Lynn/Vanhatten IQ and national wealth data in which he got the best fit with verbal IQ transformed as a threshold variable. This seems to suggest that verbal ability is more important economically. If that's true then maybe European descended peoples can hold their own against the East Asians.


Anonymous said...

Given that there is no "sight" test for whether someone is Hispanic (see Jorge Ramos), nor have I heard of any strict documentation requirements to prove it, it would behoove as many white people as possible to claim they are Hispanic through whatever means they can. Name your kid a Hispanic-sounding name. Maybe you were a resident of a Spanish-speaking country or territory at some point in your life (I resided in Puerto Rico for 4 years as a child). Learn to speak Spanish. You get the point. If enough of us do this, perhaps we can overwhelm the system and bring it crashing down.

Deckin said...

What I don't understand is why Richwine isn't being hailed up one side and down the other for finally, finally, demonstrating conclusively that IQ tests are NOT culturally biased?

After all, if it's false that Hispanic scores on IQ tests are, on average, lower than Caucasian and Asian scores, then it's obvious that Richwine has managed to bring out what has eluded so many: Proof that IQ tests are definitively culturally neutral.

Anonymous said...

But Asians ARE serious competition and elite whites don't seem to be super worried about Asian immigration either.

Anonymous said...

Sword said...

When once compares Hispanics in general with Mexican-americans, one sees that their median test results (compared to those of whites) are quite similar - with the exception of the LAT test, where the Mexicans outscore the general Hispanics 29-19. That is quite a lot, and it is only seen on one test. Anyone have an explanation for that aberrant-looking result?

Steve Sailer said...

The terrible LSAT scores for non-Mexican Hispanics are largely due to the English-language LSAT being required in Puerto Rico to attend local Spanish-language law schools.

Anonymous said...

The header of the first table says "Mean Score as Percentile of White Distribution", while your discussion says median

Jefferson said...

If your average Hispanic had a high I.Q, than Los Angeles for example would be just like Portland, Oregon.

Meaning Los Angeles would have very few gangs,a low crime/homicide rate, tons of affluent Hispanic hipsters, and almost everybody would be fluent in English.

Anonymous said...

Wow, the Whites actually beat the Asians in a few tests. Go figure.

Florida resident said...

Not that Mr. Sailer tries to hide this (non-existing) statistics,
but putting
Chinese, Koreans and Japanese, on one hand, and
Filipinos, Cambodians and Indonesians, on the other,
into common group "Asians",
makes the discussion not that informative.
Respectfully, F.r.