Holy mackerel are people misinformed about the SAT. And boy howdy do they like to disseminate their misinformation as though it’s gospel truth. I’ve heard some really dumb things coming even from people who should know much better, like high school college counselors/guidance counselors. In fairness, mastering the SAT (not just the test itself, but the system in which it resides) is a complex and ever-evolving task. Then again, The College Board’s own website is a veritable wealth of information, so if it’s your job to know this stuff and you don’t know it, you really don’t have a good excuse. The bottom line, though, is that I’d just like to hear more stories of a guidance counselor telling a student: “You know, I’m not sure. Let me check and I’ll get back to you,” instead of sputtering half- and untruths from a position of authority. I’m getting off-track. All of this is to introduce what will probably be a recurring feature on this site: dumb (but persistent) rumors about the SAT. Oh, here go Hell come.

When I hear people say this it drives me crazy: “Take the January test, because fewer other kids take it so the curve isn’t so bad.” Let me say this without equivocation: the SAT is not curved in the classic sense, and it doesn’t matter when you take it.

If grades in a class you’re taking are based on a curve, that means your teacher looks at the distribution of grades in your class, and assigns letter grades to ranges based on that distribution. So maybe instead of saying at the outset that a 90% or higher is an A, he’ll say “the top 10% of the class will get an A.” If the class is really hard (or he’s just not that great a teacher), maybe the top 10% includes scores all the way down to 75%. By that curve, a 79% is still an A, but that’s not decided until all the scores are in and the teacher can see the distribution. Curved grades aren’t finalized until all the grades are in, because curved grades depend on the performance of the entire group that took the test that day.

That’s not how it’s done on the SAT. Yes, it’s true that scoring tables vary by test. But that’s a function of the difficulty of the test as revealed in the equating* process, not of the ability of the test takers that day. Equating ensures that results are reliable across tests. Scoring tables are not based on a curve of the results from that single sitting. Although scoring tables are finalized after you take the test, there’s no looking at the results afterwards and assigning scores based on percentages. There’s no guarantee (although it usually happens) that anyone on a particular test date will come out with a perfect 2400. Your performance on the SAT has nothing to do with the performance of the kid sitting next to you on test day. If the best performance in the whole country on a particular test day is a 2390, that 2390 is the best score. It doesn’t get curved up. 

Still not convinced? Have a look at the year-by-year average SAT score data (page 3 of this PDF). It changes! I don’t know any kind of curving system that would allow this to happen: the same score is a different percentile depending on the year. The very existence of this table is proof that the SAT is not curved in the classical sense.

Let’s just look at this from one other angle. Pretend you’re a college admissions officer. Do you want to have to worry about what month a kid took the SAT because some months are easier than others? Wouldn’t that defeat the entire purpose of a STANDARDIZED test? I can assure you that no admissions officer has ever said: “Oh, John Smith got his 1900 in March and Katie Jones got her 1900 in January, so John’s actually a little smarter.” A 1900 is a 1900, period. That’s what equating does. It ensures that, no matter when you take the test, your score indicates (within a fairly small range*) your aptitude.

The bottom line: stop worrying so much about WHEN to take the test, and start worrying about getting prepared so that you know HOW to take the test.

ADDITIONAL READING

The legendary Erik the Red put together a super-thorough chart of the curves of all released (i.e. QAS) tests, sorted by ease and by month. There’s no discernible pattern.

Here’s what the College Board itself has to say about how an SAT is scored:

How is the SAT scored?

Scoring is a two-step process:

1. A raw score is calculated.

  • One point is added for each multiple-choice question answered correctly.
  • For multiple-choice questions answered incorrectly, 1/4 point is subtracted:
    • No points are subtracted for incorrect answers to the mathematics questions requiring student-produced responses.
    • No points are subtracted for omitted questions.
  • Then, the total points answered wrong are subtracted from the number answered correctly. If the resulting score is a fraction, it is rounded to the nearest whole number—1/2 or more is rounded up; less than 1/2 is rounded down.
  • Questions in the SAT equating section do not count toward the score.

2. The raw score is converted to the College Board 200- to 800-point scaled score by a statistical process called equating.

  • Equating adjusts for slight differences in difficulty between test editions, and ensures that a student’s score of, say, 450 on one edition of a test reflects the same ability as a score of 450 on another edition of the test.
  • Equating also ensures that a student’s score does not depend on how well others did on the same edition of the test.



*Here’s a link to a really dense and boring white paper from the college board that still doesn’t really tell you how equating is done, but reinforces what I’ve said above, for the most part. Thanks to “fignewton” over at College Confidential for the link and clarification. I didn’t have this completely right when I first published this post.

Comments (3)

Great explanation.  Here’s another related question (sort of):  I’ve heard that you can gain 10 pounds in one meal from eating too much spaghetti while studying for the SAT.  

Is there any truth to that?

Leave a Reply