<p>I am not taking word of mouth as data. I am looking at school reported data from a number of highschools. As I said, these kids do get into fine schools, but have been rejected from the ivies. Now, few kids apply to every single ivy too, and that can be an issue. But I've seen where they have applied for most ivies and other top selective schools. Not at all unusual here to be rejected by HPYMS and Columbia too. That is a usual mix I see. Also there are definitely more kids who get a 2300 when you mix test attempts. Alot more. My son went up significantly on one part of the SAT, but went down on another at a second seating. However, by taking the best of each seating he does not do poorly at all. Pretty well, in fact. And he has only taken the test twice. It is not unusual here to take it 3 or even 4 times. Also the years can be mixed as some seniors will take the test for one last time in January and those results will be used.</p>
<p>Don't know how the to get that data. If you do the math, you can figure the odds .... Anyone with a certain score in the Math, for example can get a 2300 regardless of how low they scored on the other parts as long as they get a certain threshhold in future seatings. Seems difficult to me as my kids really got some decent scores by mixing and matching. The oldest actually broke 2200 when he took the test in January, because he had gotten a very high score in the math when he first took the test the previous March, and then took the writing twice before getting a high score in it (that was before the 3 part SAT).</p>
<p>
[quote]
I can't answer your question regarding number of sitings etc.
[/quote]
</p>
<p>I have asked the College Board research department about this through a College Board representative I met at the National Association for Gifted Children in Minneapolis earlier this November 2007. It may be a while before I hear back about this point. My sense of the evidence is that the number of students with a given level of SAT scores </p>
<p>which is reported by College Board each year on the basis of showing the best SINGLE-SITTING composite score on the SAT Reasoning Test for each distinct individual in a given high school graduating class (regardless of what year the individual took the test) would NOT change a lot if "superscored" scores were considered. For a lot of repeat test takers, who make up a modest majority of all test-takers, their best sitting has their best score on ALL test sections--in other words, they gain nothing by "superscoring" as to their composite scores. Colleges announce a policy of superscoring, and students are relieved by that policy </p>
<p>and then when they retake, they gain ground on ALL sections of the test. </p>
<p>But when College Board answers me specifically on this point, which many students and parents on CC are curious about, I'll let everyone know with plenty of posts on superscoring.</p>
<p>Let's say that someone has a statistical 30% chance at admission to any given Ivy based on past admissions of students with the same SAT scores and GPA. The calculation used near the beginning of the thread is flawed if it is intended to allow someone to correctly conclude chances at admission to at least one Ivy if all are applied to. The acceptance/rejection at any one Ivy has indicative effect on the probabilies that should be assumed for admission to another school. </p>
<p>For example, if Bob is rejected from Columbia, then we can assume his chances at other schools need to change. (They don't actually change, only what we think them to be does.)</p>
<p>What this should lead us to conclude is that if all the schools had similar admissions, a 30% chance at one school would very rarely lead to a single accceptance. Getting rejected at every school is entirely possible, certainly more so than was stated. </p>
<p>Hopefully this enlightens. </p>
<p>Thanks for reading!</p>
<p>P.S. (Afterthought: Perhaps this questions the direct correlation of probabilty of future occurence and past percentages...)</p>
<p>tokenadult. I did and they agree with my statement, although we don't have the most qualified teachers at my high school... but that's another story. I stand by my statement though and if you don't agree with me I believe you are misinterperting my statement. What the first 8 colleges do to the person does not affect his chance of admission at the last, much like a FAIR coin doesn't get affected by the first 8 tosses. Again, this is for a FAIR coin and for an applicant that actually has a 50% chance of getting in at each and every school (as my simulation explains).</p>
<p>
[quote]
although we don't have the most qualified teachers at my high school... but that's another story.
[/quote]
</p>
<p>That is a problem in a lot of parts of the country. Okay, let me propose this way of looking at the problem, and then you can, please, tell me if it accords with your own interpretation of your earlier statement. Let's find two students, A and B, each of whom has all the characteristics specified by the OP and who differ in other characteristics unknown to you and to me but apparent in their admission files. Then let's learn their admission results, in order from the least selective college they both applied to on up to the most selective college they both applied to. My contention is that if the results are </p>
<p>A: reject, reject, reject, reject, reject, reject, </p>
<p>and </p>
<p>B: admit, admit, admit, admit, admit, admit, </p>
<p>and then we had to guess about how they fared at two more colleges, the smart way to guess is to say that A was rejected at each and only B was possibly admitted at one or both. (Remember I said we worked up the selectivity scale as we heard the news.) That's not 100 percent certain, but what I WOULDN'T say is "they both have the same chance to be admitted to the last two colleges." Nope.</p>
<p>"Let's find two students, A and B, each of whom has all the characteristics specified by the OP and who differ in other characteristics unknown to you and to me but apparent in their admission files."</p>
<p>This is where we seem to misunderstand each other. I am not saying its only those charcteristics they have in common. Please do look at my definition of chance (the simulation). I am saying that GIVEN the fact that both A and B have a 50%, those string of rejections/acceptances don't affect the other two. </p>
<p>And again, I understand fully what you are saying, and agree that statistically, new information revealeved about the students whose chances we are UNSURE about, does skew what their chances are. But this is because we didn't know what chances of A and B were, whereas my statement referred to a person with an ACTUAL 50% chance of getting in (again, based on my explanation of chance, not on a chart of SAT/GPA/etc.)</p>
<p>^^^^^^^ yep, I agree 100%. makes perfect sense to me. exactly what i was trying to articulate but my head is too messy
besides, I don't pick fights with moderators lol ;P</p>
<p>
[quote]
I am not saying its only those charcteristics they have in common. Please do look at my definition of chance (the simulation). I am saying that GIVEN the fact that both A and B have a 50%, those string of rejections/acceptances don't affect the other two.
[/quote]
</p>
<p>Oh, well if you are stipulating away all the real-world characteristics of the problem, then maybe the result is as you calculate, but then it provides no guidance to an actual high school student who is trying to figure out how to plan his application list.</p>
<p>A couple years ago a friend emailed me a link that led me to a home page of a site about statistics, and I found a GREAT article there </p>
<p><a href="http://statland.org/MAAFIXED.PDF%5B/url%5D">http://statland.org/MAAFIXED.PDF</a> </p>
<p>about how mathematics is different from statistics, and what kind of issues one has to pay attention to in order to reason about statistics correctly. I LOVE this article--I reread it every once in a while to make sure I am picking up a correct understanding of statistics from my other reading. Statistics is all about data. There are a LOT of jokes about mathematicians based on the premise that mathematicians disregard data when thinking about real-world problems. Statisticians don't do that.</p>
<p>Wow, I can't believe I didn't read this thread before posting my "feeling</a> unexceptional" thread! I'm not sure if I'm in the Ivy League... erm... league... but the past few weeks I've been feeling really "ehh" about my 4.0/2390, since all I ever hear is that "colleges REALLY DON'T CARE about the numbers!" But this is reassuring.</p>
<p>"Oh, well if you are stipulating away all the real-world characteristics of the problem, then maybe the result is as you calculate, but then it provides no guidance to an actual high school student who is trying to figure out how to plan his application list." -tokenadult</p>
<p>Right, I think we finally understand each other. My calculations were purely mathematical in nature.</p>
<p>Although, I would point out that people CAN make somewhat reasonable estimates of their chances based on SAT,GPA, ECs, recs, essays and then perform the calculations I suggested, which although are clearly not guarantees, do provide 'guidance'</p>
<p>Oh, and one more thing regarding the calculations. I don't believe the stats' teachers statements have much meaning because we are basically saying that based on the results of 8 rejections, the person probably did not have a 40% chance at ANY of those schools, and thus any calculations are futile. In other words are 40% was totally arbitrary in that case and carries no meaning.</p>
<p>argh.
i wanna smack my self for that 2290.
gosh. would it kill me to get 1 more question right? </p>
<p>LOL.</p>
<p>Let me throw out two questions that may shed some light on this:
First, what are the factors that probably explain the seemingly "random" patterns of acceptances we see here anecdotally (ie, in at H, out at P, or vice versa)?
Second, what are the factors that probably explain some people being rejected by all their selective schools despite having stats and ECs that seem to place them well within the range?
It is my opinion that the answer to the first question is that there are many factors that affect the decision and that differ from school to school--some of them objective (geographical diversity, for example) and others subjective (how a particular adcom reacts to an essay). It is this variation, in my opinion, that makes the OP's idea somewhat true--and that makes it reasonable to apply to more, rather than fewer selective schools if one is within the stats range. In other words, if your stats say you have a reasonable chance of getting into an Ivy League school, your actual chances will be greater or lower at various schools, depending on factors unknowable by you.
For the second question, I always suspect that there is some unknown disqualifier connected with the student--perhaps he's an obnoxious jerk, and that comes through in his essays and interviews. But if it's somebody who applied only to HYPMS and Columbia, that's not even a very good test of the OP's thesis, because the somewhat less selective Ivies aren't included.</p>
<p>What if an applicant meets the OP's criteria, but has no extracurriculars?</p>
<p>^ frankly most dont, since almost all people that good usually do something outside of school (usually math circle or science olympaid or whatever) and the small few that dont are outliers, just as you will find people w/ 2.0s and 1500 SATs and No ECs admitted to HYPM as well.</p>
<p>
[quote]
you will find people w/ 2.0s and 1500 SATs and No ECs admitted to HYPM
[/quote]
</p>
<p>You will? Majoring in what? Do you personally know such a person?</p>
<p>I think that grades and ECs have such an overwhelming impact on admissions that the extra stuff (essays, recs) are what account for the 30% chance of admittance. It's like the statistics you get in psychology...like black males have a higher chance to go to jail than white males...obviously there are alot of other factors than just race, but the statistic itself is still helpful in predicting outcomes.</p>
<p>Almost every chance calculation has many factors outside of the factors which are considered...but hey, that didn't stop MIT from calculating a way to beat the house.</p>
<p>"w/ 2.0s and 1500 SATs and No ECs admitted to HYPM as well."
LOL! Yeah right! Like who, pray tell?</p>
<p>Since so much of this thread is debating the math of it, let me post a simple, concrete counterexample that shows why the OP's math is flawed. (This is roughly the same one I posted in the other thread.)</p>
<p>Suppose there are ten students. Five of them have spectacular EC's, or are triple-super-mega-legacies, or whatever, and these five each have an 80% chance of getting into any given Ivy league school. The other five are convicted felons who don't plan to graduate from high school, but still managed 2400's on their SATs and got straight A's due to grade inflation.</p>
<p>We would expect that four of these ten students--specifically, four of the first five--would get into any given school, and thus the group will have a 40% admit rate at each school. Fine.</p>
<p>But if each student applies to nine schools, how many will get into at least one of the nine? The convicted felons, with a 0% admit rate at each school, will not get in anywhere. The superstars will each have a 100% chance (minus a tiny rounding error) of getting into at least one of the nine.</p>
<p>What's the "Ivy league admit rate" for this group? 50%.</p>
<p>Not 99%. 99% is the chance that a student whose chances are 40% at each school will get into one of the nine. However, the 40% admit rate per school that the OP cited is an average across thousands of students. Each student's chances will vary depending on other factors (EC's, hooks, whatever--the minutiae of the admissions process are irrelevant).</p>