SAT prep for all. Why bother?

The New York City Department of Education is starting a new program to provide online SAT prep for all high school juniors in the city. Their stated purpose is to do something about the steadily falling average SAT scores in the city, and if they’re lucky, maybe they’ll have some success with that (though, to be honest, I’m doubtful). Even if they do manage to raise some test scores, in the long run, are they doing anyone any good? (Anyone besides whoever gets paid to design and run the program, of course).

I have to wonder why anyone should care about the SAT at all. The SAT mostly tests how good you are at taking the SAT, and the few actual skills it does test–the ability to pick out useless information from a boring passage, a few basic math concepts (e.g. the angles of a triangle add up to 180 degrees), a bizarre assortment of supposedly crucial vocabulary words, and a few incorrect “rules” of English grammar–are either so pointless or so basic that it should be a national embarrassment that we waste high school students’ time by making them take this test. Still, given how much influence the SAT has over the lives of soon-to-be college students, New York City can hardly be blamed for wanting to improve students test scores. In any event, it’s the colleges that need to come to their senses and realize the pointlessness of the SAT (though a critical mass of parents and students might do), not the public school districts.

Still, let’s leave aside both the absurd pointlessness of the SAT and the probable failure of New York’s plan and assume that they do succeed in raising the average SAT scores of their students. In fact, let’s go a step farther and imagine that not only all New York students, but all students get access to a program like this. Will this do anything to encourage equality? Or to make sure that the “best” students get to go to the “best” schools? Or to ensure success for everyone? NO!

  1. Test prep can only do so much. Though no one is sure exactly how the connection between socioeconomic status and test scores works, as far as I know, no one doubts that there is at least a correlation. That graphic comes straight from The College Board, the people who sell the SAT (made for them by ETS) and who spend a great deal of time insisting that the test is in no way biased towards high income kids. No one is sure exactly why this is the case–Smarter parents? More test prep? Better schools? Higher expectations? A test taking gene that allowed those parents to be successful in the first place?–but you can bet that it’s not all due to test prep, at least not test prep on the scale that New York is planning to provide. A few students who are both motivated and get hours and hours of $400-per-hour private tutoring might make spectacular gains in prep scores, but even students who take expensive test prep classes don’t go from a 300 to a 700 (I know. I teach these classes.). In other words, I have trouble believing that this amount of test prep is going to make that big a difference for the kids who are limited to it. And even if it does make a difference, parents who can afford it are still going to pay for more test prep for their kids. The more free test prep there is out there, the more test prep companies are going to try to convince rich parents that their high priced programs will lead to more score improvement than that schlock the city is providing–and they’re probably right. Whatever gains can be had from test prep, rich kids will get more. In other words, the average score might rise, but the distribution of scores won’t change.
  2. However much we might talk about success for everyone, the SAT is explicitly designed to make it impossible for everyone to be successful. It is designed to produce a normal distribution or scores with a mean of about 500 (though I think the mean for the verbal section is actually something like 508) and a standard deviation of 100. If too many students get too good at the test, the ETS will simply pull the same trick they pulled in the mid-90s in reverse: they’ll “recenter” the scores. That’s the great benefit of their goofy 200-800 point scaled scoring system. Since it doesn’t directly reflect how many problems the test-taker answered correctly, they can simply change the scale any time they want.
  3. I suppose they could run into trouble in the unlikely event that so many students got so good at the test that scoring on a curve turned into complete absurdity (miss one problem, you get a 750, miss two problems you get a 500). In that case, they’ll just rewrite the test with harder problems. They have to keep the normal distribution of scores or else their whole argument that the SAT provides a standard way of comparing students from different schools descends into pointlessness.
  4. Even if the ETS does nothing, if average scores rise, schools will simply raise their expectations for SAT scores.

Regardless, the SAT can never allow everyone to be successful, and no amount of test prep is going to make a serious dent in the pattern of who ends up on top and who ends up on the bottom. The solution? Throw out the SAT. Throw out homogenized universal education to a pathetically minimal standard. Focus on education to help every student reach their own best potential in every area of life, not just the three Rs.

[Edited to add: Disclaimer: I do occasionally teach classes for major test prep companies. I don’t intend this to represent their views, and I don’t intend to argue for their best interests. I’d happily sacrifice that income stream to see the death of standardized testing. Make of that what you will.]


2 thoughts on “SAT prep for all. Why bother?

  1. What about the signals that a good score sends? While I’ll agree with you that SAT scores are not indicative of intelligence, they do send signals. Not everyone gets the same high school education as not everyone went to the exact same high school. However, making kids take a standardized test does send signals about how they compare to everyone else across the country.

    Yes, the SAT and ACT are simple tests, similarly the GRE is a simple test and I’m willing to bet that the MCAT and CPA exams are also not *that* difficult for those within those respective fields. However, it’s a standardized metric by which to measure people. Is it an absolutely perfect way to measure an individual? No, not by any means. But it is one way to compare people across states/regions.

    I highly doubt that any college/university would base its decision entirely off of a student’s SAT/ACT scores, but surely you agree that it plays some role in the decision making. Would you expect a person who got a 200 to get into Harvard?

    Further, ETS likes to use an adaptive test to figure out your level of aptitude. They “zero in” on your score by asking you progressively harder questions on the exam until you get something wrong. Then they adjust the difficulty much more finely until they get what they assume to be an accurate score for you. It’s not like if you miss 2 questions you’ll automatically go down 150 points like you suggest. More likely, if you miss the first 2 questions your score may go down 150 points whereas if you miss the last 2 questions your score may only go down 20 points.

    While I’ll agree that test scores and grades are definitely not great indicators of intelligence (I scored very poorly on the verbal sec,tion of the GRE, for example), I will disagree with your (presumed) opinion of them being entirely useless. They do give decision makers some additional information about their applicants which does help them make better decisions on average.

    • superplexa says:


      you’re absolutely right that SAT scores do say something about you. The question is what do they say? I’ve heard the exact same argument you give (that it provides a convenient way to compare students from different schools and different places) from college admissions officers (heck, I used to work in college admissions — I’ve MADE that argument), but just because we have an easy way to rank students from all over the country (or world) doesn’t mean that it’s a ranking we ought to be using. Why not just use socioeconomic status, or zip code, or school ranking?

      I’d also add that, even though college admissions officers (and everyone else), talk about the usefulness of a universal metric, I’m not convinced that it’s all that necessary. Admissions people know the schools they’re dealing with. They know which ones offer a bazillion AP classes and the International Baccalaureate degree and which ones don’t even offer calculus. They know which ones use a 4.0 GPA scale and which ones use a weighted 6.0 GPA scale. They know which ones inflate grades and which ones are tough graders. They know which ones give glowing recommendations for every one of their students and which ones give honest assessments. And mostly, they know which ones tend to produce students that do well at their particular college.

      The suggestion that they need SAT scores to compare students from different schools is an insult to the intelligence of admissions officers. The SAT is nice as a convenient check on all the other information on an application, but I think colleges would get along just fine without it. At least, smaller schools and more competitive schools would. Big state schools that use a more formulaic system for admitting students might have trouble, but I bet even they could manage.

      (By the way, the ETS doesn’t use the adaptive system for the SAT. It’s still the plain ‘ol pencil and paper test. They use it for the GRE and the GMAT, which are for grad school and business school respectively. Also, the MCAT actually is that hard. I’ve taught MCAT classes too. They’re about five times longer than your average SAT class, and the students usually put in hundreds of hours of study in order to get decent scores.)

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s