The SAT has a kind of weird number system for its scoring. While most of the tests you take in school classes are scored from 0% to 100%, SAT points go from 200 to 800 per section (totaling 2400), representing how well you did in a more abstract way.
Calculating SAT Scores – Your raw SAT score
There’s a penalty for wrong answers, so you can’t figure out an SAT score based just off of how many answers you answered correctly; you also have to know how many you skipped and how many you got wrong.
In any given section (other than the essay), your raw score is counted like this:
(# correct) – (# incorrect)(.25)
But if there are seventy math questions altogether and you get thirty-five points, that doesn’t mean that you’re going to score a 500 on math (halfway between 200 and 800).
Calculating SAT Scores – Score percentiles
If you just want to know how good your score is, then, it’s not so easy to just think about it in terms of number of correct answers or in terms of percent answered correctly. If the score “1640” means nothing to you on its own, then how do you know how good it really is?
The easiest way is by percentiles. Using the College Board’s table of SAT score percentiles, you can see how well you did in comparison to other SAT takers. If you’re in the 50th percentile, that means you scored better than 50 percent of people. If you picked a hundred people at random and lined them up in order of highest to lowest SAT scores, your 50th percentile ranking would put you right around the middle of the line. If you’re in the 60th percentile, you’d be ten people closer to the high scorers in that same line.
That table definitely helps put some perspective on your scores and what they look like to colleges.
How scaled SAT scores are calculated
Unfortunately, it’s not so easy to explain why and how scores are translated into the 200-800 scale for each section. Why not? First of all, it’s not possible to make two different SATs (with different questions) be on the exact same difficulty level. They work to make it as close as possible, but question creation isn’t an exact science. And since the College Board has to make lots of tests, making sure that every one is fairly scored is tricky.
In order to make different versions of the test equivalent, they relate your scores to the scores of other test takers (the percentile of your raw score) and to scores of past tests. If you score the exact average raw score (which may be different from test version to test version), then you’ll get the average score in the 200-800 scale (which is constant between test versions). But there’s an even trickier part—they need to make sure your score isn’t dragged down if a whole lot of people just happen to be having really good days. If that happened, then the average would go up, and your scaled score would suddenly be lower than it should be.
These problems are taken care of together by equating, which is, thankfully, not something you need to do on your SAT. In order to have a something constant to use to balance the scores, every SAT has one “experimental” section, which isn’t counted in your scores and has questions that are nearly identical to some “experimental” questions on other tests. By checking how testing groups fare on that section, they can check the general ability of the group. Knowing that helps keep people from getting lower scores because of those freak good days. (And no, you won’t be able to tell which section is the experimental one).
To put it simply, if you get a 1500 on your first SAT and a 1600 on your second SAT, it’s not because the second one was easier, and it’s not because of anything anybody else did. That’s all taken into account.
It’s only about how ready you are to take the test.