<p>^^ well said</p>
<p>Show me the data to which you refer in this quote: “They often report that they do not discriminate among scores at the high range. The data convincingly disprove this”</p>
<p>Ah, but silverturtle, the SAT math 2 test already fails completely at the upper end. There is, indeed, a wide range of ability and achievement in mathematics found in applicants who score around 750 and above. But the test doesn’t bring that out. You can only find that difference in other data - test scores for the AMC10/AMC12/AIME/USAMO/Mandelbrot and other high level tests, grades in courses past Calc BC, participation in various math camps and seminars. Again, I have personal knowledge of a good number of kids who are doing math at the real upper levels, not merely the material covered by the SAT math 2 exam. They do not uniformly score 800s on the SAT Math 2 exam. For these kids, the SAT math 2 is mostly an exercise in focusing on the boring for an hour.<br>
Now, if all you can bring to the table is test scores - if you have nothing remarkable to show - then those 800s might carry some modest weight. But equally, admissions officers might wonder why you spent all your energy on test scores and didn’t do something else.</p>
<p>
</p>
<p>Is anyone convinced by neomom’s argument that scores 750+ on the SAT Subject Test in Math Level 2 do not correlate or negatively correlate with mathematics ability? I certainly am not.</p>
<p>
</p>
<p><a href=“http://talk.collegeconfidential.com/college-admissions/651345-race-college-admissions-faq-discussion-3-a-49.html[/url]”>http://talk.collegeconfidential.com/college-admissions/651345-race-college-admissions-faq-discussion-3-a-49.html</a></p>
<p>[Princeton</a> University | Admission Statistics](<a href=“http://www.princeton.edu/admission/applyingforadmission/admission_statistics/]Princeton”>http://www.princeton.edu/admission/applyingforadmission/admission_statistics/)</p>
<p>[Brown</a> Admission: Facts & Figures](<a href=“Undergraduate Admission | Brown University”>Undergraduate Admission | Brown University)</p>
<p><a href=“http://www.infogoaround.org/CollegesChinese/RevealRanking.pdf[/url]”>http://www.infogoaround.org/CollegesChinese/RevealRanking.pdf</a> (page 8)</p>
<p>neomom, what data can you share in support of your position?</p>
<p>
</p>
<p>If I were an admissions officer, I would not want to accept students who are not willing to exert themselves for sixty minutes to demonstrate their skills.</p>
<p>Silverturtle - you keep asserting that there is data that shows that 800s on the SAT math subject test provide more of a chance for admissions at the elite schools than scores of 750 to 790. I have asked for that data.
Your data in “Race” does not split out SAT 2 math 2 subject test data. It is also based on a small, self reporting group and is possibly highly biased. Quite a few of the posters here at CC are highly focused on test scores.
Princeton lists scores for the SAT 1s, not the subject tests. They group 2300-2400 , and do not split out math 1 800s. I’m happy to agree that 2300-2400 on SAT 1 is better for admissions than 2100-2290.
Brown, similarly , does not list scores for the SAT 2 subject tests. They do, indeed, show that 800s are admitted at a higher rate than 750-790s for the SAT 1 math section, but it is a small effect, as I have already noted.
The infogoaround paper, is an unreviewed paper, focused on student perceptions. Page 6 does have a graph which suggests that possibly admissions is directly related to SAT 1 scores, but there is no detailed explanation of how this graph was generated and the scale is fairly crude, making it entirely possible that the uppermost reaches are an estimate. And, again, it has nothing to say about SAT subject tests.</p>
<p>
</p>
<p>I do not recall ever asserting this. </p>
<p>
</p>
<p>Yes, self-selection is a factor. I would likewise prefer a larger sample size, but 148 is not small enough to render the data useless.</p>
<p>
</p>
<p>Biased in such a way that non-URM rejectees with sub-2300 scores are willing to share their results but that non-URM acceptees with sub-2300 scores are not? For those who haven’t checked the link, I’ll summarize my analysis:</p>
<p>Total sample size: 148</p>
<p>Asian sample size: 58
Average SAT score for an Asian acceptee (17 were accepted): 2347</p>
<p>White sample size: 52
Average SAT score for a White acceptee (16 were accepted): 2353</p>
<p>The acceptance rate over various ranges for Whites and Asians: </p>
<p>2310-2400: 47% (29/62)
2210-2300: 10% (3/30)
600-2200: 0% (0/16)</p>
<p>neomom, I again request that you produce data that indicate that there is no admissions advantage in having higher scores (you pick: ACT, SAT, SAT Subject Tests of any range – anything that could lend a shred of objective proof to your claims).</p>
<p>I do not have an valid statistical data concerning the SAT 2 math 2 subject test scores, splitting out scores in the 750-790 range and comparing them with 800s in terms of admissions to elite schools. I can’t find anything like this. As far as I can tell, neither can you, Silverturtle.
As to your comment that " If I were an admissions officer, I would not want to accept students who are not willing to exert themselves for sixty minutes to demonstrate their skills. " - I can say that this is fairly absurd. The kids I know of are AIME/USAMO kids, some of them. I know , approximately , what their USAMO and AIME scores are. These are extremely difficult, long tests . Their scores on these tests and excellent grades in upper level university math courses demonstrate their ability far better than getting a few more simple questions right on the SAT 2 math level 2 subject test.
(For those who are not aware of the AIME and USAMO competitions - a brief summary follows. The AIME is 3 hour exam which students can only take by achieving a high score on the AMC 10 or 12 - lower level exams which are still more challenging than the SAT math 2. The top scoring students from the AIME (using a metric that is slightly complicated) continue to the USAMO which is 9 hours long and takes two days. The top students there go on to the US math team for international competitions. Kids who take these exams work hard. But they don’t all get 800s on the SAT math 2. They do get courted by the best schools in the country.</p>
<p>
</p>
<p>I absolutely agree. However, you have failed to counter any of my claims. By some very reasonable extension in accordance with the mentality you presented, a college should excuse a truly brilliant student’s poor grades because he or she is bored by taking tests (many, many more hours of tests than that required for the Subject Tests, mind you). Moreover, you’re debating my general claim about admissions practices with a very small and extraordinary group.</p>
<p>Isn’t the test out of 50? If you answered 46, didn’t you only miss 4?</p>
<p>
</p>
<p>The student left four blank and missed additional questions.</p>
<p>Kids who get 800s in Math II are more likely to be the ones that go to AIME and USAMO. Kids who get 750-790 are less likely to be the ones that go to AIME and USAMO.</p>
<p>If an admission officer sees 760 Math II and USAMO Gold, he/she will think “He did not study enough for the 800.”</p>
<p>
</p>
<p>It would not be due to a lack of preparation but to apathy.</p>
<p>Silverturtle, you misrepresent my position. I did not and do not assert that the SAT 2 math level 2 subject test does not correlate or negatively correlates with math ability. It is merely an imperfect instrument for measuring math ability. It works quite nicely for a large fraction of the high school students who take it. But, like most broadly based, multiple choice tests, it is flawed. It tells us nothing about ability to do proofs, to do math past roughly the second year high school honors level, to do big problems (by which I mean problems taking 20 to 120 minutes).
So, I would assert that the SAT 2 Math 2 exam is poorly correlated to math ability in the upper reaches. Students who have the ability to do proofs, to handle big problems, to handle calculus, linear algebra, topology, proof geometry and other material as well have no way to demonstrate that ability.</p>
<p>
</p>
<p>
</p>
<p>What else could this mean?</p>
<p>
</p>
<p>RAlec114 and silverturtle (postings 53 and 54)
I have an admittedly small sample set (perhaps 20 to 40) but kids who do well on the AIME and USAMO do not always score an 800 on the math 2.</p>
<p>
</p>
<p>Nor did I claim that they did, nor is this relevant to my arguments.</p>
<p>Silverturtle, re Posting 50.
I am not responding to your broad claim that ‘better is , well, better’. I am responding to the assertion that there is a detectable difference in admissions chances to elite schools based on a 750-790 and and 800 on the SAT subject Math 2 test. Let’s remember the thread and history:
**Title of thread: “Should I cancel my Math Level 2 subject test” under
“SAT Subject tests preparation”
Gadgor opened the thread with a question about the math level 2 subject test. I responded to his/her question and the general discussion following concerning high scores on the math 2 test, asserting that all scores in the upper 700s are treated in the same way. I was not talking about the whole range of SAT subject test scores. I was not talk about about SAT 1 scores. I was disagreeing with your state “ Higher scores are better; there are no thresholds.” . I presumed that you (and any other loyal readers) would understand that all my comments referred largely to the upper reaches of the SAT 2 Math Level 2 test and not to all subject tests, and certainly not to SAT 1 scores. I did mention an example (posting #19) of kids whose SAT 1 scores are very high but not perfect as a general example.
You specifically asserted (post #22) “But to argue that admissions officers should consider, for example, a perfect 50/50 on the SAT Subject Test in Math Level 2 to be the same as 39/50 is logically incoherent. Luckily, all the data indicate that the higher one’s score is, the better.”
However , you did not provide ‘the data’, nor have you provided any date related to SAT subject tests. This is not surprising – I have yet to find any very detailed information involving SAT subject test scores and admissions at the elite schools. You did provide some information about SAT 1 scores, but that is besides the point.</p>
<p>Silverturtle, responding to posting 56 - I think my point is clear. The SAT 2 Math 2 test does not possess the necessary range to detect the extreme high end of math ability. A score of 750 does not indicate the lack of such ability. A score of 800 does not indicate such ability. A score of 750-800, roughly, does indicate high math ability. (ANd, for completeness sake, I would allow that anything over 700 for a test-taker under 12 indicates a strong potential for very high ability as well - but , of course, we are talking here about tests that will be submitted to colleges.)</p>