<p>Okay. But that doesn’t give you reason to discourage me from going to top schools. I didn’t ask for that advice, nor do I need it.</p>
<p>
</p>
<p>No, actually they are. Resorting to limited anecdotal “evidence” does not refute the existence of published statistical studies demonstrating a steepening nonlinear relationship between admission probability and standardized test scores at certain distinct thresholds (at the strategically active universities that Kameron is looking into). However, data from elite institutions relating high school GPA and class rank and admission probability do not substantially increase once a certain threshold is met to the same extent as standardized test scores. For instance, Princeton categorically publishes its admission statistics for viewers to obtain a basic numerical understanding of the qualities demonstrated by the class. For students enrolled in the class of 2013, those with 4.0 GPAs were admitted at a 16.9% rate, while those within the 3.90-3.99 range were admitted at a 11.2% rate ([Princeton</a> University | Admission Statistics](<a href=“http://www.princeton.edu/admission/applyingforadmission/admission_statistics/]Princeton”>http://www.princeton.edu/admission/applyingforadmission/admission_statistics/)), which is fairly flat and influentially incongruous relative to the more pronounced increase in admission probability as SAT scores surpass the 98th percentile. Thus, [url=<a href=“http://www.infogoaround.org/CollegesChinese/RevealRanking.pdf]this[/url”>http://www.infogoaround.org/CollegesChinese/RevealRanking.pdf]this[/url</a>] is not merely a patently defective and dimensionless study that represents basic correlation. I have yet to see a study completed that supports the non-causal claims of escalating nonlinear dependence of test scores on admittance through the inclusion of all relevant evaluative criteria. Yet such an investigation is profoundly difficult to be statistically convincing given the purported holistic evaluation scheme and the inherent degree of subjectivity that affects admission choices.</p>
<p>mifune, I don’t think the available data in any way supports the proposition that, for any given individual (such as kameron), the chances of admission to Harvard (or any other school) are greater with a 2400 than with a 2360. While you make some interesting points, one would need a lot more data to support the strong form of causality that you are claiming. Here are just a few issues that come to mind:</p>
<ol>
<li><p>The chart in the study you cite is not sufficiently detailed to show the difference, if any, in admit rate between, say, a 2350 and a 2400. The chart was being used for a completely different purpose and was only trying to demonstrate a general trend for admit rates above the 98th percentile. The data that QEDdad pulled from the CC stats profiles at post 57 of this thread - <a href=“http://talk.collegeconfidential.com/sat-preparation/864093-ivy-league-level-sat-score-4.html[/url]”>http://talk.collegeconfidential.com/sat-preparation/864093-ivy-league-level-sat-score-4.html</a> - suggests that the admit rate at Harvard in the 2350-2390 range (60%) is virtually identical to the admit rate for 2400 scorers (63%). Of course it’s a small and potentially unrepresentative sample size, but it’s an interesting comparison. I also note that the data used for the revealed preference study is over a decade old, so even if it were presented with a usable level of detail, who knows whether it’s still valid.</p></li>
<li><p>To show causality, one would have to hold all other factors constant. It’s reasonable to assume that higher SAT scores are positively correlated with other attributes that admissions committees find attractive. So even if one had sufficiently detailed data to show the correlation between higher SAT scores and higher admit rates, the causality might be due partially, or even primarily, to these other factors. But for any given individual, all other factors, by definition, are held constant. So a positive correlation between higher SAT scores and higher admit rates in general doesn’t mean that, for any given individual, an increase of, say, 50 points, will have any impact on that individual’s admissions chances. (Obviously, once the gap gets great enough, it’s almost a certainty that it will improve the individual’s chances - e.g., there’s almost certainly a difference in odds for a given individual with a 2200 v. the same individual with a 2400 - but I’m not aware of data that would permit a valid calculation of the magnitude of the improved odds even for a sizeable gap.)</p></li>
<li><p>Reported SAT scores do not distinguish between scores achieved on the first try versus scored achieved in multiple tries, and may not even distinguish between single sittings and superscoring from multiple sittings. Admittedly, I have no data to support this, but I have to believe that an admissions committee would find a 2350 achieved in a single sitting on the first try at least as impressive, if not moreso, than a 2400 achieved by taking the CR score from one sitting, the M score from a second sitting and the W score from a third sitting. And a rational admissions committee, if they saw a 2350 from a single sitting on the first try, would have to conclude that the candidate was capable of scoring a 2400 if he/she had taken the exam enough times (and, perhaps, gotten a little lucky). The college board’s score ranges would support this conclusion.</p></li>
<li><p>Finally, and yes, this is completely anecdotal, I have spoken with the dean of admissions at Harvard about test scores and my conversation with him supports my conclusion (expressed in my earlier post in this thread) that even if kameron retakes the SAT and gets a 2400 (and as noted before, getting the 2400 is by no means a certainty), his/her odds of admission to Harvard will not increase one iota.</p></li>
</ol>
<p>@Mifune, you’re missing the point:
<a href=“http://talk.collegeconfidential.com/1064513065-post60.html[/url]”>http://talk.collegeconfidential.com/1064513065-post60.html</a>
<a href=“http://talk.collegeconfidential.com/1064516015-post70.html[/url]”>http://talk.collegeconfidential.com/1064516015-post70.html</a></p>
<p>True, a trend line can be drawn from this data as the people at <a href=“http://www.infogoaround.org/CollegesChinese/RevealRanking.pdf[/url]”>http://www.infogoaround.org/CollegesChinese/RevealRanking.pdf</a> have, however, due to the volatile nature of the data and its associated high deviations, a correlation test must be done in order to warrant that the correlation between SAT scores and admission still exists past the 21/2200+ window. There are no significant correlations between SAT and admission probability past a certain threshold for all colleges tested, therefore the admission data that your study was based off of cannot be representative of the mean acceptance probability for applicants with 21/2200+ SAT scores.</p>
<p>While it doesn’t directly support noncausality between SAT scores and admission probabilities at this extreme, it does claim that there are no likely correlations between SAT and admission probability at the 21/2200+ level.</p>
<p>Has anyone successfully assembled that non-causal model yet or is it just mere conjecture that one exists? The study is not completely insignificant nor can it be simply dismissed as correlation. Other factors may be merely continuations of the same evaluation scale. But grades, work ethic, dedication, for instance, are much more difficult to evaluate and standardize on one scale; thus, it is logical that the sensitivity applied to those may stabilize beyond a certain threshold (which it seemingly does) due to the unavoidable skew associated with their evaluation on a non-standardized continuum. Cognitive assessment, however, is the one main factor that applies to the entire applicant pool and can properly discriminate differences in academic aptitudes, clearly and with some apparent prominence. </p>
<p>Moreover, the demonstration of merit in college admissions can never simply be considered “good enough” – as if some evaluative ceiling exists or that universities choose to disregard the test’s ability to differentiate between those achieving at high levels. Each ten-point disparity constitutes a difference – but, admittedly, the extent to which each difference may affect a college admission decision is difficult to discern with any bit of resolution. Elite institutions claim that test scores may become irrelevant beyond a certain point, but considering their emphasis as a criterion in published ranking systems (in which higher scores increase their ranking and subsequent prestige among the public) and the reality that data consistently points in the opposite direction, it is hardly sensible to indiscriminately believe their assertion.</p>
<p>Oh, another thing - how sensible is it to assume that a 2100 or a 2200 is as equally attractive to a university as a 2400? Are universities that deeply apathetic about the effect of standardized test scores on national and world rankings to consider the demonstration of merit beyond a certain point irrelevant? Taking limited self-reported data from a public message board does not properly suit the degree of rigor necessary to refute nationwide studies.</p>
<p>I feel so inferior 
lol</p>
<p>You guys are taking the situation to the two extremes. A 2100 is definitely not equally attractive to a university as a perfect score, but nor is 2360 significantly less desirable to a college than a 2400.</p>
<p>
</p>
<p>I feel the same way. failboat and mifune are around my age, and yet I can’t argue like them.</p>
<p>
</p>
<p>I am not resorting to any extreme. There is a difference between a 2360 and a 2400 – albeit not one that is as significant as greater disparities.</p>
<p>@jerrry: Me too :p. Lol, thank goodness you don’t argue like them though. =)</p>
<p>I acknowledge you and cosar too :D. Yeah, but I wish I have that kind of skill. Maybe I should join Debate team at my school, but I would fail due to my stuttering.</p>
<p>@mifune: of course there is a difference, but my point was that the difference is not significant enough to warrant multiple retakes or to sway the admission decision from a rejection to an acceptance.</p>
<p>^ Your subjective evaluation of what warrants a retake or not is pretty moot. I’m guessing you base your views off only anecdotal evidence so you can’t really support a claim that a jump from a 2360 to a 2400 won’t sway an admissions decision.</p>
<p>@xrCalico23: Sure, that’s a fair assessment but a forty-point difference might make a positive contribution. Kameron may also have a personal goal regarding his SAT performance and, although this is directed at no one in particular, it is no one’s responsibility to dismiss that. Whatever his choice, I would fully support it.</p>
<p>@jerrry: haha, you should take lessons from mifune or something. He can seriously turn everything into a debate.</p>
<p>
</p>
<p>Do remember that the burden of the proof falls upon he who makes the assertion. In light of the lack of correlation between test scores and admission probability at these higher corridors, I am simply stating that the data that you are basing your observations off of are flawed in the context of your conclusion, in which case it’s flawed to ask for objective proof of noncausality between SAT scores and admission probabilities at those scoring levels.While I’m in no position to interpret such patterns, if you truly want a reason as to why there exist such a lack of correlation between SAT scores and admission probabilities, then I would venture to guess that, even if there are no “score ceilings”, after a certain SAT threshold, it no longer matters to the AdCom how well prepared you are for the SAT’s. While colleges do tend to emphasize the need to have a high median test score, they don’t take it to the extremes.</p>
<p>
</p>
<p>The lower bound of 2100/2200 are dependent upon the school surveyed. The window of the research was chosen in order to mimic those used in your study. By narrowing the range of test scores that are allowed into the test, the correlation between test scores and admission probabilities decreases further.</p>
<p>@mifune: yes, it is his personal choice and his choice alone. But then again, since he created a thread specifically to ask other posters the value of retaking the SAT, he should be able to navigate the deluge of opinions that we hereby offer with both an open mind and a certain degree of confidence in what he himself believes. If he has a personal goal of reaching a 2400 or something, then that is his own goal. We’re arguing here whether SAT scores after a certain point make any difference for admission purposes, which is not specifically directed toward any single person.</p>
<p>By the way, did you just chance your location? :)</p>
<p>I remain steadfast in my belief that kameron should focus on the “full package”, his GPA and extracurriculars, rather than shift his focus and attention to a trivial 40 points on the SAT.</p>
<p>^“I remain steadfast in my belief…” …aww, why are we all sounding like mifune. 
kidding.</p>