<p>Well, yeah, I guess that they haven't changed the nature of the test too much. A system where USAMO qualifiers sometimes get 790s (I've seen this happen) and then people who struggle in calc get 800s (this happens too) is erratic and nonsensical.</p>
<p>
[quote]
Well, yeah, I guess that they haven't changed the nature of the test too much. A system where USAMO qualifiers sometimes get 790s (I've seen this happen) and then people who struggle in calc get 800s (this happens too) is erratic and nonsensical.
[/quote]
</p>
<p>Each individual component section of the SAT (such as math) has an error margin of +-30 points; a 790 and 800 are statistically equivalent. Furthermore, this test is designed to measure all levels of high school students; the sample is for all intents and purposes nonexclusive and not self-selected. Most high school students don't take calculus until college. Those who do take it, even if they struggle, are far above average. Since this is not a test of calculus (combined with the fact that statistical anomalies can and do occur), it's entirely unsurprising that some get 800s.</p>
<p>This is hardly nonsensical, unless you don't believe in statistics and stupid mistakes.</p>
<p>I didn't read all the responces, but the minimum area is definately less than three. If the two legs of lengths 6 and 7 were close to parallel, then the third leg would be almost 13 units long. This would make an area of close to zero. The larger (closer to 180) the angle is between the two 6 and 7 legs, the smaller the area.</p>
<p>"The 24 is a trap, because people will think 6*8 = 48 and then end up with 24, but this is not possible, since you can't multiply the hyp by a leg.
The 13 is another trap in that someone might do 6+7."</p>
<p>"lol I know...I just felt I would explain it."</p>
<p>Indeed that was a great explanation! :(</p>
<p>General Rak, I understand that there's a "margin of error", but my bigger point is that a test should include both above-average math students and math prodigies in the same score range. If the SAT is to be truly useful for top math-oriented universities like say, Caltech, it has to better distinguish ability at the upper end of the spectrum.</p>
<p>780 - 800 or so is a statistically equal score range. Any score in that range should be treated as statistically equal to any other score in that range, and as far as I'm aware that's what colleges do. Any test designed to be taken by everyone will by definition be less accurate at the margins. It's the same on IQ tests, for instance.</p>
<p>To get a more precise IQ reading, psychologists have more specialized tests. Same thing here. Colleges can and do get a better read of mathematical ability from SAT IIs, AP classes, and competitions (like AIME / USAMO, etc.).</p>
<p>That's why, at the top, SAT I score differences are mostly irrelevant. The difference between a 2320 and a 2380 is statistically insignificant, but that's the way it has to be in order to accommodate every high school student.</p>
<p>Exactly - I just think that the system needs to be improved. The AIME and USAMO serve well to separate the often vastly different ability levels of students who score in the top range on the math SAT, but not everyone knows about them. If we make the AIME and USAMO critical in the identification of top math talent, we need to make it clear that they are the standard (since these tests are certainly "studyable", and otherwise the playing field isn't nearly level). Same goes for the physics olympiad, chem olympiad, and everything else.</p>
<p>shoot xiggi i'm dumb***...yeah I forgot to do 26/4 which is def. less than 7. alright its 21 and 13.</p>
<p>Random, yea, I've qualified for the AIME the past two years, but I didn't get an 800 on this past SAT (I made a reading error). I know two other kids like that, it's kind've maddening.</p>
<p>Did anyone mention heron's formula O_O</p>
<p>don't even think 2c needs that...</p>
<p>To solve the problem of the SAT's limited ability to measure math abilities, they should simply include some really difficult bonus questions so that the top score is somewhere around 820-830. </p>
<p>That would make it more accurate, right?</p>
<p>:-D</p>