Friday, September 12, 2008

On Acing the SAT

NPR's Morning Edition ran a fascinating snippet recently about a program at Baltimore Polytechnic Institute, which is a magnet high school in Baltimore. The school's "Ingenuity Project" has a good track record of students earning perfect scores on various sections of the SAT. (NPR sounded like everyone was getting 2400s or 1600s -- the old perfect score -- but the school web page is more toned down -- it says from 2001 to 2006, 12 students scored 800s on various sections, and one scored a 1600; you can listen to the NPR story here).

The most fascinating thing for me about this story was that NPR made it sound like the point of the program was to get students to ace the SAT, which isn't the case (though the project does like to publicize such scores). The program is designed to have Baltimore students achieve on the national levels in various competitions and gain acceptances to top colleges. To do that, the Ingenuity Project has students accelerate through the high school math curriculum and do calculus by their junior years. The program also leaves plenty of time for AP science courses, labs, and independent research (see here for more details). I've been coming across more and more public schools that have these research programs -- partly to have students win Davidson Fellowships, Intel nods and the like. What is rewarded gets done. We will talk about this more in the near future.

But anyway, back to the SAT question. The Ingenuity Project is a great example of a high expectations public school program. In 2005, three participants earned finalist nods in the Intel Science Talent Search. But with all the accelerated math these students are doing, you might think that more than 12 students would have achieved perfect SAT section scores. As one young man told NPR, there was nothing on the SAT they hadn't seen.

Fundamentally, a test like the SAT tests both content knowledge and your ability to call it up quickly and apply it to new problems. This is one of the reasons people dislike the SAT -- if your geometry class grades show you know the subject, why does it matter how quickly you can conjure it up? Why does it matter that you can solve a geometry problem that's next to an algebra problem? Others dismiss the test altogether, saying it can be coached.

But while coaching might raise a kid's 1800 to a 1920, coaching alone will not get anyone a perfect score. If a program like Baltimore's Ingenuity Project, which likes to publicize its perfect scores, can't get more than 12 of them in a 5 year period, that shows how difficult such a score is to achieve. It's not just a matter of having an excellent high school curriculum -- though obviously that's a big part of it.

For many years there's been a cultural debate over the place of the SAT. Once it was viewed as more of an intelligence test -- designed less to cover the high school curriculum, and more to identify people with high intelligence, regardless of what their high schools covered. Over the years this idea has fallen out of favor. Some schools are now even moving away from requiring the SAT.

But I still think this idea of fast critical thinking has a place. It may not be the most critical factor for predicting high grades in college -- it would make sense that one's academic grades in high school would be a better predictor of this. But it may predict some of one's future success in life. A Vanderbilt study done two years ago studied kids who did well on the SAT, taken in 7th or 8th grade. They were more likely than a comparison group of graduate students to earn an annual salary of more than $100,000 or achieve tenure in a top-50 institution. That's not what the SAT sells itself as predicting - so it's quite interesting if that's the case.


InTheFastLane said...

As a predictive test in 7th & 8th grade, I would think that might be the case in the reading and critical writing areas, but for math? Math requires specific skills to be learned and if a student has not taken Geometry, it is unlikely they will do extremely well on the math portion until they have been taught geometry.

Anonymous said...

The math section of the SAT does not require much training---my son took it after 6th grade and got 720. The reading is not hard either (he got a 730). The writing section does seem to be harder---he only got 560 on that, and his essay scored 3 out of 12---nearly as low as it gets.

I've no idea whether middle-school SAT scores are predictive of future success. So few kids take the SATs then that it would be hard to get an unbiased sample.

hschinske said...

There has been a *lot* of research on those who do very well on the SAT at young ages (and also on those who do a good deal less well -- keep in mind a substantial proportion of talent search participants score in the 400s, which is considered fairly good performance for that age, and another big wodge in the 500s, which is considered very good).

About Baltimore Polytechnic: I think the focus on the 800 scores is misguided. What you really want to know is how many got above 750 or so (not sure where the cutoff should be, as I'm not that familiar with recentered SAT scores). Above some level, in a population of people who've studied all the math formally, the ranking is generally a question of speed, which at that level often has more to do with reading speed than anything else. While reading very quickly is undoubtedly a helpful ability to have, and may be related to quick processing in other areas, it's not math.

I don't think there's anything wrong with using a test that loads information processing speed so heavily, as long as it's understood that it's only one TYPE of achievement, and that others count just as much or more.

It is certainly possible to answer a lot of the math questions on the SAT without having had formal exposure to the material. Two of my kids scored 500 and 600 without having taken much high-school-level math at all, nor having done it on their own (the one who scored 500 had not had any high school math at all). As many talent search participants score below those levels, it's not just that the test is easy. I think for such students the SAT really does function as the aptitude test it once was, rather than solely an achievement test.

Anittah Patrick said...

"What is rewarded gets done."

Indeed. Incentives are a powerful, powerful thing.

Anonymous said...

"What is rewarded gets done."

Indeed. Incentives are a powerful, powerful thing.


I'd beg to disagree. I was influenced by Alfie Kohn and his marvelous book, "Punished by Rewards," and eleven years later, it's still an inspiration. It's how I parent and guide. My daughter is a CTY student and attends a highly selective math science technology school. I must have done something right!


Anittah Patrick said...

The premise of Punished by Rewards is that what gets rewarded gets done, and thus encourages us to be mindful of what kinds of extrinsic behaviors we reward. So, given that you are inspired by this book, it seems that in fact, you do agree that what gets rewarded gets done, and that's why you have chosen to opt out of deploying certain incentives.

Also, it seems that view your daughter's attendance at CTY as a "reward" for you having done "something right". So, you doing "something" in a manner defined as "right" is what is done, and the "reward" is your daughter going to CTY and a math/science magnet.