September 22, 2010

Is the SAT getting easier?

It's interesting to look at California SAT scores over time. The farthest back data on the College Board site for the state of California is 1998, so I'll contrast 1998 to 2010. Overall, the mean score in California has dropped 3 points from 1520 to 1517 on a 600 to 2400 point scale. (Because the 1998 test was reported on a 400 to 1600 scale, I'm multiplying 1998 scores by 1.5 to match them up with 2010 scores).

A three point drop doesn't sound like much but that stability masks all sorts of things going on beneath the surface. For example, mean test scores have gone up for most ethnic groups: whites up 33 points, Asians up 57, blacks up 29, Mexican Americans up 14. (The total score went down because lower-scoring groups grew so fast.)

And that pattern of rising scores within groups is especially noteworthy because the number of high school seniors taking the test has gone up. (This data looks at each senior in the class of 1998 or 2000 and counts only the last time he or she took the SAT, so taking the test multiple times isn't a factor in these numbers.)

California SAT 1998 v. 2010






College Bound seniors 1998 # 2010 # Chg 1998 Mean 2010 Mean Chg 1998 v NHW 2010 v. NHW
Total 142,139 210,926 48% 1520 1517 -3 -89 -124
White 56,217 69,969 24% 1608 1641 33 0 0
Asian, As-Am, or Pac Isl 29,889 44,932 50% 1557 1614 57 -51 -27
Black or Af Am 8,868 14,476 63% 1292 1320 29 -317 -321
Mex or MA 18,494 42,380 129% 1341 1355 14 -267 -286
Other Hispanic 6,606 20,735 214% 1359 1325 -34 -249 -316
Puerto Rican 489 699 43% 1434 1489 55 -174 -152
American Indian 1,415 1,256 -11% 1479 1488 9 -129 -153
Other 7,863 8,498 8% 1566 1561 -5 -42 -80
No Response 12,298 7,981 -35% 1520 1566 47 -89 -75

For example, 24 percent more whites in the class of 2010 took the SAT than whites in the class of 1998, although I would guess that there were fewer white 17-year-olds in California in 2010 than in 1998. (Correct me if I'm wrong.) Black SAT-takers went up by 63 percent, even though lots of black families left California between 1998 and 2010. Clearly, there is a big push to get more marginal kids to take the SAT. (It's a great era to be in the testing racket!)

All else being equal, the higher the percentage of a group who takes the SAT, the lower the expected mean score (the scraping-the-bottom-of-the-barrel effect). But, except for the "Other Hispanic" category, where test takers exploded by 214%, we don't see that in the big groups here.

So, what's really going on with mean SAT scores in California? A few possibilities:

- Scoring is getting easier. 
- People are getting smarter.
- People in California are getting smarter relative to the rest of the country.
- Students are more familiar with test-taking because of all the other tests they take now.
- Students are better test-prepped for the SAT in 2010.

Considering the rank order of the size of the effect -- biggest gain among Asians, next biggest among whites, then blacks, then Mexicans, finally a bad dropoff among Central Americans, most of whom have recently arrived and don't know about SATs -- that would be about what I'd expect from average enthusiasm among parents for test prepping, so I don't think we can rule out the last possibility.

22 comments:

Wilson said...

Off topic, but take a gander at the GOP's "Pledge to America." It's basically a George W Bush Redux. It advocates no useful policy on enforcing the borders. It merely advocates (buried way down on page 20) to reaffirm the right of states to enforce immigration laws - a right they already have.

And I'm gunna go out on a limb and wager that it says nothing about eliminating racial preferences or CRA.

But it's opening lines are the best. "America is more than a country. America is an idea...America is an inspiration to those [immigrants] who yearn to be free..."

Aunt Zeituni couldn't have said it better.

Republicans have learned nothing from the Bush years. It's not a question of whether or not GOP policies will win votes, it's a question of whether they will actually make America better. We already know the answer to that.

Anonymous said...

Back in 1991, Asians nationally outscored whites only by 5 points on the SAT. Today, that number is substanially larger. So it does seem that the growing Asian-white and Asian-NAM gaps are not entirely influenced by IQ, or else we'd expect the gaps to remain constant. Not grow larger.

By the way, there's been a huge proliferation in SAT Asian cram schools in the past couple decades. So this is consistent with the growing Asian-White/NAM gap.

Now you could add that Asians are closing the gap because they are acculturating and their verbal skills are getting better. Only this ignores that the Asian scores on both the math and verbal have improved to similar extents, so it doesn't appear that languge acculturation is the major driver of score improvement. It suggests that better test prep is a more likely reason for the improvement.

The mix of Asian immigrants does seem to be skewing more elite, so perhaps that has something to do with improving scores too.

More test prep + elite immigration = higher Asian SAT scores.

Anonymous said...

it is possible the highest IQ achievers in silicon valley, who give it the reputation for greatness, mostly arrived as adults after being statistical outliers from wherever they originated. the people born and raised there are normally distributed as everywhere else?

Mitch said...

Hell yes, the SAT got easier in 2005 when it added the writing section.

Math went from 60 to 54 questions. Reading went from 77 (ack, I think it was 77) to 66. Writing on the Subject test was 60 questions, it went to 49.

You think they took those questions off the easy end of the test? They couldn't--the achievement gap was already huge.

2005 was when analogies and quantitative comparisons went away. I believe, but am not sure, that these sections had the larger achievement gap. Both had a high percentage of difficult questions.


I did test prep at Kaplan from 2003 to 2009, so I was there for the cutover. Test prep instructors are one of the few job categories that are legally allowed to test for IQ, so we're a smart group. I didn't know a single test prep instructor who didn't know in their bones that the test was easier. We used to joke that the test got longer, easier, and more expensive.

However, at first scores dropped, as each wrong answer carried more weight. The kids most penalized were those who usually scored in the low 700s. These are, typically, kids with high IQs but low attention to detail. So they'd make a few unforced errors on the test, but make it up by getting a number of difficult questions that more detail-oriented kids might miss (of course, at the top of the scale were kids who were both bright and detail oriented).

I actually hunted down the data on this a few years ago and proved to my own satisfaction that the decrease in test scores noticed right after 2005 was caused by the decrease in scores from 700-750 (per section), which was accompanied by an increase in the 650-700 range. I think I remember another increase upwards on the 400-500 score area, but I can't swear to it. And I'm not the data geek that you are, but I did run it by a poli sci prof at Berkeley who said it was convincing.

I'm pretty sure the College board changed the scale slightly after that to mask the drop. It's still very hard to get a low 700 score. Anecdotally, I know more 2400s who are bright but not extraordinary, but highly detailed. In 2003-2005, the only 1600s I knew were exceptionally bright kids.

Go take a look at the College Board red book with old SATs and compare them to the new SATs in the blue book. It's completely obvious.

Incidentally, the kids who would normally get in the low-mid 700s on the SAT are now better served by the ACT--lots of hard questions, no wrong answer penalty, and the ACT doesn't test for detail attention the way the SAT does.

Anonymous said...

Psst...percentiles are all the matter to people.

Anonymous said...

Steve I'm too lazy to check the data and figure this out on my own, so can you explain how this comparison deals with the fact that the SAT was still on the 1600 scale in 1998?

Steve Sailer said...

Just multiply the 1998 scores by 1.5.

Anonymous said...

Yes, it's getting easier. So is the LSAT.

Anonymous said...

Stanley Kaplan has been around for a while, but I feel like test prep becoming a multibillion dollar industry is a fairly recent thing: every college town now has multiple test prep centers, including Princeton review, Sylvan, Stanley Kaplan, PowerScore, etc. This is because, in spite what ETS says about prep not mattering for the SAT or PSAT, it certainly does. However, as everybody preps, the advantage naturally declines, leaving out in cold only the last dumb fools who don't get the word i.e. Spanish speaking immigrants. The reason you can prep for the SAT or PSAT is just because it tests so little in the way of math and so much in the way of vocab. After you become an expert of simple geometry and algebra problems, all that's left is to swallow a thesaurus,and presto, your board scores are up by 150 -- 200 points. But don't get me wrong. Raising board scores is probably a 1-2 year project involving at least 2 hours of effort daily so only those with absolute dedication to the process can make test prep work for them.

Now who might such people be, hmmm?

Probably the next escalation in the "bellum omnium contra omnes" annual college admissions contest will be high school students trying to ace the LSAT, GMAT, and MCAT before they enter their freshman year, thereby giving the college some sense of the applicants likely post graduate income.

I believe the best way to reform college admissions is to eliminate standardized testing in favor of individual colleges like Harvard and Yale administering their own tests and requiring applicants to travel to that college to take it. That way, Harvard doesn't have to slog through the applications submitted by every Chinese,Korean, and Indian National Merit Scholar in the United States.

Sheila said...

All such tests and licensing exams have been getting easier, one more gift of affirmative action. There's a fascinating discussion of this at Alt Right Radio (at Alt Right blog; I couldn't manage to embed the link) which is right up your alley, Steve. Richard Spencer discusses this with a Professor Farron; the points he makes about the percentage of black Harvard med school grads (pass/fail grades only) who failed their medical boards three out of three times is particularly interesting.

Dutch Boy said...

Mensa hasn't accepted SAT scores to qualify for membership in years.

Dr. Foo Man Chu said...

The mix of Asian immigrants does seem to be skewing more elite, so perhaps that has something to do with improving scores too.

This is untrue based upon my 20yrs in California, mostly the SF Bay Area.

When I was at Berkeley over 20yrs ago the Asian immigrants I knew were a lot more elite than the ones I've seen recently: children of academics, international businessmen and Communist party leaders.

Today, drive through the mid-peninsula or East Bay (or Irvine, the Valley) and see hordes of entirely new unwashed Asian immigrant communities dropped almost whole from abroad. These are not the elites - they are the middle class and below who didn't have access to the US 20yrs ago.

This applies mainly to NE Asian and especially Chinese and somewhat to Koreans.

If your claiming that Asians scores increased due to the greater percentage of NE Asians rather than the elite status of said NE Asians, perhaps you have a point.

Mote In Brother's Eye said...

Anonymous said...

Psst...percentiles are all the matter to people.

Did you mean 'all that matters?'

Kind of ironic, being an SAT thread and all...

Wilson said...

"Back in 1991, Asians nationally outscored whites only by 5 points on the SAT. Today, that number is substanially larger. So it does seem that the growing Asian-white and Asian-NAM gaps are not entirely influenced by IQ, or else we'd expect the gaps to remain constant. Not grow larger."

Negative. This does not consider the changing backgrounds of Asian students today versus Asian students of 19 years ago. Even back in the early 90s Asians did signifcantly better than whites. Today an even larger fraction of Asian students are the children of parents who came here on a skilled worker visa, mostly tech workers. So of course they will be getting smarter.

In Middle Age Europe the elites essentially willed the Ashkenazi overclass into existence by banning them from grunt work and restricting them to high-IQ, high profit professions. Our elites haven't changed today. They are willing a new Asian overclass into existence by pricing bright, scientifically and mathematically inclined Americans out of the market. In the space of just a generation or two they have created a minority with an IQ gap nearly the same as that between Jewish and non-Jewish whites.

Of course the increased gap could also be partially explained by shrinking birthrates among well-edumacated whites...

"Yes, it's getting easier. So is the LSAT."

If the SAT's getting easier wouldn't the gap between Asians and whites be decreasing, since the ceiling is more attainable? I pulled up some SAT stats the other day and at least 3 times as many Asians as whites scored perfect 800s on the math. An 800 math score is 97th percentile among Asians. Nearly 1 in 33 Asians scores a perfect 800 in math. Fewer than 1% of whites do so.

Anonymous said...

I don't think the multiply by 1.5 method is apt for this comparison; the writing section is not just a linear extrapolation of the combined math/verbal scores. The writing section of the new SAT has no close analog on the old format test, so it would be better to compare just the verbal and math parts that are largely the same.

Since (it's believed) that the writing section is easier than the math and verbal sections, multiplying the math/verbal score by 1.5 should under-predict the math/verbal/writing score.

Anonymous said...

If anyone is interested in seeing how the elite performers are doing on the SAT, I recommend checking out the new 2010 statistics. Despite the aggregate Asian scores lumping together East Asians and South Asians with Southeast Asians and Pacific Islanders, we can tell that there are still more high performers percentage wise amongst Asians relative to whites on all three portions of the exam.

http://professionals.collegeboard.com/profdownload/sat-percentile-ranks-by-gender-ethnicity-2010.pdf

What's amazing is how 4% of Asians scored an 800 on the math exam. No doubt that percentage would've been even larger if we were to have examined only the East Asian scores. And interestingly enough, roughly twice as many Asians percentage wise scored above a 700 on the writing exam as did whites.

Anonymous said...

Hasn't the percentage of Asian-Americans with college degrees, even among the foreign born, increased? That would suggest more elite Asian immigration.

By the way, there have been pretty substanial quick gains for Asians on the ACT and SAT in recent years. That suggests that something other than more selective immigration is at work. I think that test prep agencies are getting more sophisicated in how they teach.

Anonymous said...

"Kind of ironic, being an SAT thread and all..."

yeah the sat tests typos!

sure does!

champ!

David said...

Isn't studying for the SAT/PSAT/ACT missing the point?

What are these tests intended to test, if not the kids' g - their mental ability to pick up what the school is giving them? If some kids devote their attention to beating the test by grinding, they may memorize and learn a great deal (which is a positive good), but it defeats the purpose of the test. How do we know what their natural mental ability is, if they've been coached by Kaplan and crammed by Kumon?

Either all students must be forced to grind, or else all must be given the test "cold." The input must be equal so we can clearly see the difference in output.

The drive and willpower to grind may be valuable, just like IQ is. But tutors and "help" and pushy parents and other coaching "resources" will not always be there on the job in the real world. We need to measure raw mental ability, and separate this from the ability to "donkey one's way through." An employee who can grasp some new data in one minute is better than an employee who needs to work for nine hours to grasp that same data. Yeah, the nine-hour guy is tenacious and hard-working - good for him! Not so good for the business.

Because it blurs the line between the one-minute brights and the nine-hour donkeys, test prep could be considered a form of dishonesty.

Wanderer said...

Nationwide there's been no serious change 1998 to 2010.

1998 (USA)
Verbal : 505
Math : 512

2010 (USA)
Verbal : 501
Math : 516

Four points when there are 100-point SD's... negligible.

By the way, Steve, it's best to compare apples to apples: Compare Verbal to Verbal year to year, and Math to Math... not all this conversion with the new writing section. It blurs trends.

Wanderer said...

704,000 whites in 1998 took SAT
Verbal: 526
Math: 528

670,000 whites in 2003 took SAT
Verbal: 529
Math: 534

838,000 whites in 2010 took SAT
Verbal: 528
Math: 536

If whites nationally are a "control group", an easier SAT between 1998 and 2010 would show clearly improved white scores. Is 8 points in Math on a 100-SD significant?

-- Whites nationally show an avg. gain of 5 points per section from 1998 to 2010.
-- California shows a white gain of 11 points per section (OP).
-- [Non-CA USA whites: under 4.5 point gain per section].

Perhaps the better question to ask is why did CA whites improve so much more than other whites?

Maybe test-prep is a bigger business there than the for the rest of us.

Wanderer said...

704,000 whites in 1998 took SAT
Verbal: 526
Math: 528

670,000 whites in 2003 took SAT
Verbal: 529
Math: 534

838,000 whites in 2010 took SAT
Verbal: 528
Math: 536

If whites nationally are a "control group", an easier SAT between 1998 and 2010 would show clearly improved white scores. Is 8 points in Math on a 100-SD significant?

-- Whites nationally show an avg. gain of 5 points per section from 1998 to 2010.
-- California shows a white gain of 11 points per section (OP).
-- [Non-CA whites: under 4.5 point gain per section].

Perhaps the better question to ask is why did CA whites improve so much more than other whites?

Test-prep must have risen in CA in the past 12 years. Not so much in the rest of the USA.