<p>Keep in mind that the percentage of the population with an IQ in the 150+ range is minute. Sure, the SAT was more distinguishing between the highest of the highest scorers but how useful is this when considering the bulk of the applicants. With only 3802 test takers CURRENTLY scoring above a 1550 on the SAT, the amount of students scoring in this range pre-recentering was probably less than 1000. This means that more students of disparate capabilities towards the middle of the pack were receiving the same score because the fringe 120 points were devoted to a minutia of applicants, thus making it harder for colleges to differentiate between more average applicants.</p>
<p>btw, what do you really care about the recentering? That's the way it has been for years now and it seems to have worked just fine. The only way I could see a problem with it is if schools start reporting middle fifty percent SAT scores with the top quantity near 1600. (please do not tell me that harvard's 75th percentile is 1590: it is not)</p>
<p>^ Yes. Its partially because a 800 in Math today is much easier to get than an 800 in Math in 1994. You can still miss one or two today and get 800. Back then if you missed 1 you got a 780 - 750 on SAT Math. </p>
<p>So some people who get 1600 nowadays could have gotten higher than 1480 on the SAT in 1994. Yes this is true. Also, percentiles is the most accurate way of looking at it. The College Board made it easier to get a 800 on the SAT Math because in 1994, 80% of the 800 SAT Math Scores were Asian. Now that they made it easier, that percentage dropped down to about 60% since it is so easy for anyone to get 800 on SAT Math.</p>
<p>There were about 300 students who scored 1520 or above in 1994. However, a 1520 on the SAT is considered to be 158 IQ in 1994, while a 1600 today is considered to be a 152 IQ.</p>
<p>The recentered SAT makes more sense for elite colleges since there's better differentiation whithin the top 1% (120 points spread out across 99th percentile). It also helps with middle tier students since you don't have all those low-high 90th percentile students crammed in together.</p>
<p>
[quote]
The recentered SAT makes more sense for elite colleges since there's better differentiation whithin the top 1% (120 points spread out across 99th percentile). It also helps with middle tier students since you don't have all those low-high 90th percentile students crammed in together.
[/quote]
</p>
<p>NO THE DIFFERENTIATION IS WORSE!</p>
<p>Why is everyone having such a difficult time understanding this? They lowered the ceiling, They've dumbed it down. The differentiation of the top 1% used to be across 190 points (in 1994), now its 120 points. ITS WORSE AT DIFFERENTIATING, IS EVERYONE CLEAR ON THAT NOW? GOOD, IM DONE.</p>
<p>That's why they recentered it. Differentiating between them at that high a level is stupid as most customers of the collegeboard (i.e. colleges) are most interested in the 90th percentile and up. As a result, if you waste 120 points of your spread on the top .1% of scorers, it's gonna be more difficult for colleges to differentiate between the rest of the people in the 90th percentile and up.</p>
<p>don't know if that made sense...but whatever</p>
<p>"The differentiation of the top 1% used to be across 190 points (in 1994), now its 120 points. ITS WORSE AT DIFFERENTIATING, IS EVERYONE CLEAR ON THAT NOW? GOOD, IM DONE."</p>
<p>120 of those 190 points were used on the top .1% of scorers. How is that better differentiation?</p>
<p>THIS IS IN 1994. Today, it is a joke to score 1500 or above on the SAT, tens of thousands do it. It used to be about 700-900 scored 1500 or above on the SAT. Now its like in the tens of thousands. Its ridiculous, there is no differentiation among the top 1% anymore.</p>
<p>Like I said before Khalid, there is a difference between a 140 IQ and a 150 IQ. Its as simple as that. There are noticable differences, now not everyone is smart enough to notice those differences. But intelligence is not the only measure of success.</p>
<p>I don't think the SAT is necessarily a good measure of intelligence, since it's a coachable test and people improve in it all the time. Unless, of course, you could show me some actual research that says otherwise.</p>
<p>^ You may be correct. The Stanfurd Binet IQ test for up to 150 IQ and the L&M Supplement for IQ above 150 is probably a better measure. After all, its an administered test, and its most accurate when taken at age 8. </p>
<p>SAT is coachable somewhat, but I believe that reaching your limit potential becomes easier with coaches.</p>
<p>sat cannot fairly gauge a person's iq, i have an iq of... well i don't really want to share, but above the range mentioned before and all i managed was a 1470's on the sats... and just as an opinion the cattell scale is more accurate than the stanfurd scale for iqs greater than 150, westside is quite astute about that*</p>
<p>The scale is probably better today, since the SAT has never been a difficult enough exam to distinguish effectively among the top 0.5 to 1.0% of the test-takers. Essentially, some students achieved 1500+ scores pre-1995 through pure luck (when they could have very well scored a 1470-1480, let's say). The difference between a 1500 and 1600 was a mere <em>3 to 5</em> questions at most on the old scale. Think of a fairly easy exam at school, and let's say only 2 out of 200 students scored 95 or above on it. Is the student who scores 100 on the exam necessarily much more intelligent than one who scored a 95, but simply made a few careless errors or did not care about the exam as much as the top scorer did? You may argue that the school exam distinguishes well among the top 1% of the students in this case as well, but the argument would be extraordinarily fallacious. The exam was simply not good enough to distinguish between these two top students, even though one did seemingly score much higher than the other. A friend of mine from high school took the SAT in 1994, and scored a 1590 on it. The second time he took it (trying for a 1600), he only scored a 1510 or 1520. And he had only scored a 1460 or so on the PSAT the previous year! I bet the difference between the two scores was no more than 2 or 3 questions. The score difference was not statistically significant! The 1600 scorers (pre-1995) simply had a really, really good test day and were not really much better than 1500 scorers.</p>
<p>The SAT may be a test that distinguishes well among so-called "average" test-takers, but it's definitely not an effective tool to separate the top from the very top. In other words, it's too EASY an exam (for the very top scorers, at least). Hence, the present scale is more than adequate for the top test-takers -- and all test-takers, for that matter. The solution, then, is not to mess with the conversion scale again, but to make the exam even more difficult, but then you would lose a degree of differentiation with the middle-range students, who do, after all, represent a majority of the test-takers. One potential resolution of this dilemma, which I'm sure the College Board would NEVER adopt, is to have TWO levels of the exam, one being the "entry" one, which everyone takes, and the second being a much more difficult version, which only, let's say, 1450+ scorers would take to distinguish more effectively among themselves.</p>
<p>So stop griping about the recentered scale! :-P</p>