Do you think my son has the qualifications for ivy league?

Parent Cafe
201

<p>That may be true, cellardweller, but I believe that it is erroneous to assume that a good choice has to be a Top 20 university. Why has this whole discussion ignored LAC's? I know that the OP didn't mention any, but being that the OP seems to have posted once about a week ago and then not returned, his concerns don't seem to play much into the conversation anymore. </p>

<p>I also don't understand why it is wrong for an applicant to rule out schools they way you have described. I suspect that this is where we disagree, anyway--I do not think that it is usually a good idea to apply to 15 or more schools. Most people can and should rule out schools that simply aren't a good fit for them--sometimes "Duke is in the south" is a very big deal, and sometimes those little snappy, seemingly compulsive rulings are covering a whole litany of reasons why a school is being tossed aside. </p>

<p>Plus, speaking on a personal level, I thought applying to nine schools was more than enough to deal with, and I spread my applications out over several months. I cannot imagine completing the 4 to 8 more applications that I would have needed to really maximize my chances at a top university, and my apps were NOT particularly difficult (only had to write a few essays, etc.). </p>

<p>In my opinion, you could acheive the same effect as applying to a whole boatload of top 20 universities with a few well chosen applications to top LAC's. Obviously, you would have to find some that you liked, but most applicants could, if they gave LAC's a chance. Being that both Cubsfan and myself had a lot more success at our LAC's, it seems like university tunnel vision is not a great strategy.</p>

202

<p>Mombot, to me, the sentiments in your last post here point back to the underlying secondary question that should be asked by the OP and anyone else privately or publicly thinking it: "Why" should my S/D apply to an ivy league U? (Or why should I, the student?) (Of fill in, reachy school --non-Ivy --but reachy in at least one aspect, including financial)</p>

<p>There are many evident reasons why some students not only might want to, but absolutely should, seek Elite school admission, or that "reachy" but expensive school. It has everything to do with the student's needs -- including the location of the U, the "prominence" of the U, the connections available that might be unique to a certain category of school, as well as some other features that are more abundant in most privates than in most publics. (Cf. a couple of much earlier threads on the latter comparisons.) And of course, depending on the student and the student's State of residence, there may not even be the option of a State U that really has the equivalent student body and the quality of programs and the manageable size that the student may need psychologically -- compared to "top 20."</p>

<p>I absolutely do not have a judgment about decisions of students (and the supporting opinions of their parents) who choose to limit themselves to "top 20" (some portion of those), unless I were to know such a student quite well. However, I also think that there are hundreds of students who limit themselves to this category needlessly. I say that based partly on the subsequent return to CC of many who admit that they have fully embraced a less reachy alternative (public or private), are prospering there, admit to not previously having a clearly defined reason for applications to so many reaches, & cannot figure out why they once insisted on limiting their list to a particular strata (which is often more about rankings than excellence), etc. Typically, these are students who never did have the specific orientations I listed in an above paragraph. Many of them, as capable, as accomplished as they are, will take quite awhile to determine even their major or concentration, let alone their options beyond the 4-yr degree. Many of them will be seeking a career in academics -- which opens up a huge variety of excellent undergrad institutions for them. Others know only that will end up in the sciences, but have not narrowed their direction beyond that. And in many, if not most, of these cases, the graduate institution or the professional school will be more determinant of that student's future success than the undergrad, AND where the graduate school will be looking beyond "top 20" graduates to fill their graduate or professional school admissions.</p>

<p>The proportion of students whose future is truly dependent on particular undergrad institutions is, I would guess, less than half of those who apply to top 20. But the peer pressure to apply to top 20 is enormous at most of the private high schools. </p>

<p>I think at least 3 questions, or categories of questions, should be asked:
"Does my S or D have the qualifications?" (for reachy school)
"Why should S/D [or I] apply to these (reachy) schools?"
"Should my S/D limit his or her choices to these schools, and if so, why?"</p>

203

<p>advantagious:</p>

<p>Who said LACs can't be part of the strategy? My D applied to 12 top universities and 6 LACs. Got into 5 Us and 3 LACs. She probably should have applied to more LACs as she was waitlisted but not rejected from half. LACs seem to like the waitlisting game to protect their yield, particularly with strong candidates who may opt for Ivies. At LACs the number of admitted students is very small and the selectivity at the top ones is just as high as the top universities. So you can't apply to just a couple and expect to be admitted either. My main point is simply that you need to apply to many schools to be admitted to a few, whether LACs or Us.</p>

204

<p>
[quote]
My main point is simply that you need to apply to many schools to be admitted to a few, whether LACs or Us.

[/quote]
</p>

<p>Unless you can afford to apply ED and can convince yourself to apply ED to a realistic choice (a match or slight reach) rather than throwing away your one and only ED chance on an extreme reach school.</p>

205

<p>JHS: </p>

<p>I will give you one example that proves the point. Harvard has reported a 25/75 range of SAT scores of enrollees of 2080/2375. Based on a simple computer model, where you fit the curve at each individual point in between the 25th and 75th percentile you get a median of around 2260/2270. (It is non-linear as the percentage of admits rises with the score). A median of 2260/70 by definition means that at least 50% of the students score over 2260/70. (Combined multiple sitting score). An analysis of YPSM would show the same and even signficantly greater for MIT where the median combined score is around 2,300. </p>

<p>In order to prove that these students are part of the 99th percentile of scorers, you just need to show that a 2250 combined is roughly equivalent to a 2200+ single sitting score. </p>

<p>It is actually relatively easy to model for large cohorts the correlation between single sitting and combined scores, even if the variations may be substantial for each individual. Some schools like UCLA which only accept students based on single sitting scores also report individual ranges which can validate the model. </p>

<p>I maintain that you can show that a best combined score of 2250 is approximately the same as a best single sitting of 2200 which is the also the 99th percentile cutoff. </p>

<p>You can intuitively work through the analysis as follows:<br>
-There are 25,944 verbal SAT scores of 750 or greater reported by the College Board for 2006
-The number of writing scores over 750 is only 17,500. This means that a very susbtantial number of the high verbal scorers (at least 8,400 but more like 12,000) had writing scores of less than 750. A weighted average would give about 21,000 combined score of 1500 or greater. A third score will reduce the number averaging over 2250 even further. So again, the 2250 total combined scorers can't very far off from the 2200+ single sitting number of 19,000.</p>

<p>So again, 2200+ (single sitting) = 2250 combined score= 99th percentile = <20,000</p>

<p>This does not address the probability of admission of 99th percentile scorers. We know that they are just a minority of applicants, estimated at around 20% based on the Harvard and Brown numbers for applicants with individual scores greater than 750. The Brown data already tells that the applicants with a single SAT score of greater than 750 have a 25% chance of admission. Since chance of admission climbs with score and that 2250 is harder to achieve than a single 750 score, then the rate of admission for 99th percentile scorers is significantly higher than 25% at Brown. By curve fitting I find around 33%. </p>

<p>With a small margin of error, you arrive to at least a 25% chance of admission at any individual college part of HYPSM for the 99th percentile and at least 33% at the next 10 colleges. Still not 50% but certainly still about 2.5 times the chances of the average applicant.</p>

206

<p>Cellardweller, the problem with your analysis is that you are assuming that the score itself raises chances for admission, when it is very likely that highly competitive students for the Ivies are usually equally strong on other factors that are far more influential in terms of the decision. For example, if you look at reported class rank, you will see that valedictorians fare better than any other group -- the same mindset that will cause a kid to jockey for #1 place standing at a highly competitive public high school (such as the one the OP's son attends), would also tend to lead that same kid to prep extensively for the SATs and sit for repeated sessions until target scores were achieved. Similarly, that same kid is going to be aiming for top standing and recognition in his EC's. And you are going to see that pattern over and over again, because it is a competitive game. </p>

<p>So your analysis is like using measures physical height to assess likelihood of a kid being recruited to play Div I basketball. It is true that height confers a definite advantage to basketball players, but height alone will never do it. No coach wants a 6'6" player who is slow to react and clumsy with a ball -- the kid has to be an excellent player. And the coach is going to select for the strongest players, not the tallest-- unless the player is so short as to be essentially "out of range" for what the coach needs. </p>

<p>SAT scores are very easy to manipulate, because the system allows for unlimited prep and coaching and multiple retakes of the same exam. So of all the factors on a student's application, a relatively low SAT score is the easiest to rectify. A kid can't erase the effect of a D in trigonometry or take up the oboe as a high school junior and expect to be any good at it .... but the kid can make a concerted effort to turn an 1850 first-sitting score into a 2250. </p>

<p>Obviously college ad coms know this -- so they aren't going to have their eyes bug out at a high score, nor select for scores -- hence the large number of high-scorers who they reject. Class rank and strength of academic record is going to trump SAT score every time, because you can't fake that. The ad coms are going to use a much lower threshhold score as an initial screen - and then select for the rest of the package -- and then the high-end score may be the deciding factor in the end-game round of admission. As between two roughly equivalent applicants with a 200 point score differential, the higher scoring student will get in. The problem is that it is just as likely that neither of the "roughly equivalent" applicants make the grade, because the ad com is honing in on students with more impressive credentials.</p>

207

<p>Calmom:</p>

<p>I am assuming nothing. I am just providing the stats. The results are quite clear. On the lone metric of SAT score, a 99th percentile is associated with a 2.5 greater chance of admission compared to the average applicant. That's it. It also means that if you have a low SAT score, you are up the creek without a paddle. </p>

<p>It clearly does not mean that high scorers are guaranteed admission. If anything 75% will be rejected by any individual HYPSM. </p>

<p>But a key element is that as group they will perform extremely well. About half of the total in the nation and a much greater number (more like 80%) of those actually applying are eventually admitted to a top school. </p>

<p>That is where your analogy breaks down: There are millions of tall people but very few (by definition) SAT scorers in the 99th percentile. If your analogy was correct it would be like saying that half of all tall people play in the NBA and that 80% of all tall players that apply end up playing for some team. A better analogy would be a metric such as percentage baskets from the 3-point line. The 99th percentile in that group is probably a pretty good source of top players, even though other factors will also come into play. But if the player can't shoot from a static position, he is probably not going to do well while running and jumping. </p>

<p>So, the correlation between enrollment and test score remains very strong. It has both high positive and negative predictive values for admission. </p>

<p>The next most important factor would be class rank with top schools enrolling 90% in the top 5% and 95% in the top 10%. Little room is left for the nationally ranked chess player if he is not a top scorer and also at the top of his class. He may be picked ahead of some other kid with similar rank and scores, but probably won't even be in the running unless his scores/GPA are already up there.</p>

<p>Whether rank trumps score remains to be seen. At many schools, the GPA of the top 15-20 students are virtually the same and may have much to do with which courses were selected than capability. A challenging curriculum is often seen as a must for top colleges, so a kid with a marginally lower GPA but a much tougher schedule will have an edge. If they are both in the top 2%, it become just like the 99th percentile SAT score. Above 2250 or so, there is not much of an advantage.</p>

208

<p>cd,
Sorry if you've answered this before, but what you do define as a "low" SAT score? (resulting in being up the creek without a paddle)</p>

209

<p>epiphany:</p>

<p>Anything below the 25% percentile for any of the SAT scores would qualify as "low scores". For HYPSM that is still in the mid 98th percentile. At that point, average admit rates drop well below 10% and you need a solid hook such as legacy, athletic recruit or URM status to have any type of a shot.</p>

210

<p>cellardweller, I agree that a person with scores below the 25% mark needs a strong application in order to overcome the score weakness, but I know for a fact that the "hook" does not have to be legacy, athletic recruit, or URM -- because I know of specific individuals without those particular "hooks" who have gotten in with scores below that level. </p>

<p>It is the biggest myth in the world that the bottom 25 percentile is entirely made up of people with some admission preference. </p>

<p>25% is a LOT of students. It is one out of every four. If you have a thousand students, then that means 250 will have scores below the 25% mark. </p>

<p>Sorry for this second-grade math lesson, but this is the sort of myth that irks me because the numbers are right there. </p>

<p>Obviously it doesn't help to have lower scores, and I am not indulging in the fantasy that some student with scores 300 points below the 25th percentile mark is going to get in.... but the reality is that the ad coms do look at the whole package.</p>

211

<p><a href="%5Bb%5Dcalmom%5B/b%5D%20wrote:">quote</a> you are the one who continues to "flunk" both logic and critical thinking. You cannot impute a causal relationship from a table showing correlations.

[/quote]
</p>

<p>Instead of repeating buzzwords, could you answer the question? That is, tell us what, exactly, are the A and the B for which A was said to cause B where only a correlation exists.</p>

<p>
[quote]
As presented on the Brown web site, it is mostly a matter p.r. -- and I've already pointed out the admissions practices that would best serve the strategic goals of Brown to maximize yield while at the same time enhance ranking.

[/quote]
</p>

<p>Unfortunately your advice to Brown (accept people with exactly one low SAT subscore) was fallacious. The Brown web page indicates that, if anything, it is <em>total</em> SAT score that influences yield; the yield varies linearly and at the same rate with respect to verbal and math scores as one reduces either of them from 800.</p>

212

<p>
[quote]
It is the biggest myth in the world that the bottom 25 percentile is entirely made up of people with some admission preference. </p>

<p>25% is a LOT of students. It is one out of every four. If you have a thousand students, then that means 250 will have scores below the 25% mark.

[/quote]
</p>

<p>At my alma mater, URMs make up about 13% of each class and legacies another 13%.</p>

<p>That could theoretically account for your entire bottom quarter, even leaving out the athletes (for whom I don't have a percentage).</p>

<p>Of course, not all URMs/athletes/legacies have low statistics. My daughter, who is a legacy, is entering this university with stats at about the 75th percentile level, and one of her friends, a URM, has stats at about the 50th percentile level. But there are certainly plenty of kids in these special categories with low stats as well.</p>

213

<p>I don't think anyone should be singling out a "lone metric" when predicting admissions chances, because colleges don't operate based on that. Yes, the correlations are there, but consider, along with whatever calmom's personally known cases are, my own known cases of two recently admitted students with zero hooks (as large hooks are traditionally defined): one SAT section below the 25th percentile, the other SAT section above the 75th percentile. One got into H (ranked 2nd in class), the other into Y and P (ranked 1st in class). High school type (& program) = extremely rigorous. Academic accomplishments & e.c.'s were quite unusual. Clearly, of all the factors, the percentiles of the SAT's had the least influence on the admissions decisions.</p>

214

<p>Here is some other stuff: About 12% of the NMNF are emrolled in HYPMC.
There is a total enrollment of about 7100 students. The top 25% is 1775. 25% of of 16000 NMF is 1920 so not far off (assuming NMF are near the top 1% of SAT scores. Cutoff vary from 204 to about 220 or so or 2020 to 2200). The next 14 schools have freshman enrollment of about 22000. There are about 18% NMF enrolled or about 2880. The top 25% enrollment is about 4500 students. This would make sense as the top 25% SAT cuts are a bit lower for this group.
So if you are looking at 1/2 empty, and you think that most of the top scorers do in fact apply to HYPSMC then you can say that it is tough to get in just on test scores (probabilities are higher than for low scores I know). On the other hand you can say that there is plenty of room for non NMF to get into elites who have overall great apps, notwithstanding legacies etc. It would be interesting to know (and we do not) How many top kids such as those in the state of Illinois, who are ACT kids, do not even apply to the elites. I think calmom's point is that a straight top scores/top GPA without a geographical or other hook would make a kids chances way lower than 1:2 at HYPS. For the whole population you may be able to make a statement but for any individual by looking at hooks or lack there of, the probabilities can be higher or way lower.
The models shown and the calculations are great. My gut feels is still to think the pool of superscores or ACT equivalents country wide is probably closer to 30000. The biggest problem with the composites on the SAT is that the writing portion is not used by the schools (at least not shown) and is often so much lower. I have often seen my students make significant leaps in one of the other sections on a second take.</p>

<p>Conclusions. Well, it is better to do well and have a hook than not to. Duh. There is a lot of room at the top 20 schools for top kids(not just top scores). There are lot of good students who are enrolled at top state schools and LAC's. (See the number of NMF at OU, tex,ASU, the Cal schools, LACS etc). Going into the college game thinking rank, and test scores are the way to the elites without taking into account distinguishing factors, geography, etc is foolhardy. Knowing a bit about where you are from, your high scools characteristics, what others have done to distinguish themselves etc can make the college app game much lest of a "crapshoot" than it appears. An honest assessment of "fit" is appropriate,</p>

215

<p>Calmom:</p>

<p>I did not say you positively had to have a hook to get in with SAT scores in the 25th percentile. Your chances are just much lower, as I said a single SAT score in that range drops your chances to below 10% on average. It is also true that many of the URMs (but not all) and many of the athletic recruits are in that quartile. Legacies are all over the range. This does mean though that statistically, and anecdotal evidence aside, that chances for unhooked kids in the lower quartile are very low, in the order of 3-5% at best.</p>

216

<p>I haven't been able to look at this thread in a while, and I want to make certain to thank cellardweller for making his analysis more clear.</p>

<p>I still have a feeling it proves too much, since it would lead to a conclusion that high-scoring kids have about the same chance at HYPSM and Brown or Dartmouth, and that doesn't correspond to observed reality at all. (It also makes me feel really bad for my kids, who must have either written terrible applications -- but I don't think so -- or had terrible luck to blow such high chances.) In any event, the differences in analysis are of degree, not kind. Like cellardweller, I have seen precious few kids with high SATs and high grades shut out of the top [20] [30] [whatever] colleges unless they really made a special effort to "achieve" that. Many choose not to play that game, or to play another game ("Let's go to college for free!"), but the ones who want that get it (although not HYPS).</p>

217

<p>oldolddad:</p>

<p>I would not use NMF status as a metric for admission. The PSAT is taken much earlier than the SAT and although the numbers often track it is not used directly in the admissions process. Clearly NMF status can offer a slight boost to a candidate over another another one with the same SAT score but won't compensate for a low SAT score or rank.</p>

<p>Now that most schools are standardizing on the new SAT and using the full scores, it is clearly the better metric. In general, two SAT II scores will be added to the three part SAT score for a combined test score. </p>

<p>There is nothing magical about using a 99th percentile with about 20,000 students versus a larger pool with 30,000 or 50,000 students. The problem is that admission drops rapidly with scores in the 98th percentile. I have found through trial and error that using the 99th percentile cutoff provides the optimum discriminant ability for the SAT score metric. A lower cutoff and you loose to much specificity (i.e too many admits are outside the cutoff value), a higher cutoff and you loose too much sensitivity (you are not capturing enough of the admits inside the cutoff).</p>

<p>Also, I never claimed that somebody in that group would have a 50% chance of admission to HYPSM. I actually estimated it at 25% and at 33% for the next tier of schools. To get to 50%, you pretty much need a perfect score of 2400.</p>

<p>The point is also valid that many schools below the top 20 can also prepare candidates well for graduate or professional schools.</p>

<p>In regards to professional schools, though, there is very susbstantial differences in admission rates to top schools between the top colleges and the next tier. A disproportionate amount of the students admitted to top medical, law and business schools (over 50%) come from a very small set of colleges (and a few LACs). Sure you can get into HMS from State U., but the odds are very much against you.</p>

218

<p>cd,
i.m.o., the problem with your statements is that they dwell on "averages" (statistics). That's all that the CDS's are: they're a report of ranges, based on a bunch of data points, without correlating those data points to the variations of "other factors" (not hooks, btw) in admissions. </p>

<p>Statements like "overall," or "generally" tell us very little, relative to the subset of already highly achieving students (the only ones credibly competing for upper-tier admissions).</p>

<p>The below-25th-%ile students even applying to HYP are not the same variety of below-25th-%ile applying to non-HYP schools, let alone to schools in the 2nd & 3rd tiers. My evidence for that is not just the statements of HYP reps, who say that "80-90% of our applicants are qualified to attend." It is also the CDS's & other statistical reports of non-HYP schools across the country, which reveal that many high-scorers apply to, and are admitted to, many non-Ivy schools.</p>

<p>Student X,with a very high score and highly inflated GPA or a non-challenging high school program, is not even in the running for an Elite, if not hooked. Whereas a non-hooked student such as those I described in my previous post above is very much in the running -- as evidenced by results. The latter's chances were never 3-5%. However, Student X's chances may very well be 3-5%.</p>

219

<p>QUOTE: "Sure you can get into HMS from State U., but the odds are very much against you."</p>

<p>But you can get into outstanding public law schools and med schools from mid-level State U's and from non-illustrious privates. Happens all the time.</p>

220

<p>JHS:</p>

<p>I appreciate all your comments and some of them have helped me refine the analysis further.</p>

<p>In regards to odds of admission, I did not claim that they were the same at HYPSM or the next tier including Brown, Dartmouth, Columbia etc.. </p>

<p>In the first group, I found that a 99th percentile score of 2200+ gave you approximately a 25% chance at HYPSM (slightly less at MIT which is heavily biased with high math scores). A perfect score would give you a 50% chance. </p>

<p>I found that chances rose to around 33% at the next tier of colleges with close to 75% chance of admission with a perfect score, a substantial difference. </p>

<p>I fundamentally subscribe to the conclusion that kids with top scores/grades get into top universities or LACs unless they make a choice to go elsewhere or fail to apply to enough schools. </p>

<p>The counterpoint is also important. Unhooked students will low SAT scores (lower than 25% percentile for a given school) will have a very low probability (5% or less) of gaining admission to top universities or LACs.</p>