<p>What would happen if the number of perfect scorers increased dramatically (say by 2x in one year)?</p>
<ol>
<li><p>The fact that more people get a perfect score than the next highest score (2400 vs. 2390) creates statistical bias, and this bias diminishes the importance of any signal attached to a perfect score. Not necessarily the usefulness of a perfect score to the student. But it provides less information to colleges. If this trend persists, the test will need, somehow, to start discriminating among that group; perhaps with some very hard questions that don’t currently appear on the test. </p></li>
<li><p>In general: the test is starting to demonstrate more of a problem known as endogeneity. You all know the basics of this problem: a given score demonstrates nothing but the ability to prepare for the SAT, and it (arguably) becomes less useful as a predictor of success in college. </p></li>
</ol>
<p>A corollary is that more perfect scores tend to reflect the ability to prepare for a perfect score. At some point, that skill becomes disconnected from predicting success.</p>
<p>You might say (and I think Silverturtle has said this) that the real goal of colleges looking for perfect scores isn’t to predict collegiate success. The goal is to present the right statistical profile to the world, in terms of yield and average SAT. I’m just suggesting that if, hypothetically, the number of perfect scores were to double, any of these goals would be harder to achieve. And it would necessitate a response from CB or the top colleges. Either a harder test or, at some point, less incremental weight attached to a perfect score.</p>
<p>Wow, that’s a lot - it basically jumped 100 people since my year. I’m guessing the growth of test prep companies as well as population will make an upward trend inevitable.</p>
<p>lol, Only 382? There’s like 1000 people on this forum alone that claim to have a 2400. </p>
<p>It’s really not a crowning achievement - the written test is subject to the opinion of the reviewers (eg, an “acceptable” response), and the math section consists of geometry and some basic algebra. </p>
<p>Getting ~2000 on a first try without having ever seen an SAT before sounds fine to me (and this is the point of the SAT), but spending a year studying some rather crappy topics is insanity.</p>
<p>How many people got a 2400 on their first try?</p>
<p>Also, what is with all this rubbish that the SAT is just about how well you prepared and not a real test of intelligence? I haven’t prepared for the SAT at all (besides taking a few practice tests) but have consistently scored 2400’s on all of them (without the essay, which I skip). Also, my little brother, who didn’t even know what the SAT was beforehand and is a certified genius, got a 100% on the multiple choice portion when I gave him a practice test. I haven’t done any studying at all for the SAT or ACT or any similar test and am reasonably confident of a 2400. However, I may practice a few essays before the SAT to get a feel for them, and look at some essay guides. That’s it, however.</p>
<p>We all know this already, but the SAT is being gamed by prep courses, tutors, prep books. The number is only going to get higher, no? In any case, it’s certainly not going to become a more accurate/reliable(?) standard of “measuring intelligence”</p>
<p>I think the reason the average is getting higher is more to do with population increases than anything. The class of 2011 is going to be larger than 2010. You have to do a stats test, such as the “Two proportions with samples in 2 categories” test. Or a simple ratio haha.</p>