Re-read my earlier post. The survey found that top students at well known high schools had an average of 17-20 percentage point increase across 28 colleges, some of which were far less selective than Princeton, such as NYU and Boston College. Obviously the 17-20 percentage point increase is lower for most lower-admit rate schools, like Princeton and higher for most higher-admit rate schools. It looks like the top students in the survey had an average admit rate across all listed schools of around 50%, so the example in my earlier post of increasing from 40% to 60% was reasonable. My post said these reasons made it more meaningful to compare to the relative boost of SAT increases.</p>
<p>In the study, EA was worth ~100 points on the combined SAT. Princeton’s admissions stats show that for each 200 points on the combined SAT, admit rate increases by an average of just over 2x. So 100 points would be expected to cause an increase by a little over sqrt(2), which I’ll call 1.5 or a 50% increase in odds of acceptance. I mean by a factor of 1.5, so a 2% chance of acceptance would be expected to increase to 3% (not 50 percentage points to 52%). I stand by my earlier comments that an applicant with typical GPA/SAT would have an extremely low chance of acceptance unless he finds some way to stand out from the >10,000 apps with similar GPA/SAT. It could be ECs/awards, it could be background, it could be excelling in a passion as reflected in multiple sections of the app, etc.</p>
<p>Data10, to clarify (by the way, I agree that I have an extremely low chance), do you believe it would be a waste for me to apply SCEA at Princeton?</p>
<p>I have to wonder about any statement that “EA is worth ~100 points on the combined SAT” if you are looking at Princeton where it’s not clear that 100 points on the combined SAT is worth much of anything. Sure, someone with 1550 (out of 1600) SATs may have a 13% chance of admission, and someone with 1450 a 10% chance, but it would go to 13% if he applied EA. Whoop-de-doo. You are still talking about terrible odds all around. </p>
<p>What’s really making the difference is some other set of qualities – leadership, analytical abilities, intellectualism, ability to write – that are slightly more prevalent among 1550 scorers than among 1450 scorers. The people strong with those qualities – and SATs of >1400 – may have a 50%+ chance of admission, whether they apply EA or RD.</p>
I concede it was old, but page 2128 indicates it wasn’t quite as old as 1998-1999. The rate of EA increases above separated for race, family connections, 4 separate types of alumni children, sibling attended given college, and athletic recruit. This seems like they made a strong effort to remove the effects of hooks.</p>
<p>So… now you’re saying that there is a material boost from applying SCEA, but that boost is nonsignificant? (1.5 x 8% = 12% - essentially negligible) Then there was no point in bringing up the study in the first place. Applying SCEA would result in almost no boost for the OP anyway, in order for you to make your point. Essentially, OP is not looking at a difference between applying SCEA and RD anyway, so it would all come down to whether he/she wants peace of mind and how much he/she wants to be at Princeton. </p>
<p>Also, “standing out” is an inherent characteristic. Each person is unique and each person stands out in his/her own way. You can hardly say that any two people with similar GPA/SAT are indistinguishable in any way. It all comes down to what Princeton is looking for in a given year. Perhaps they want a violinist and the OP happens to be a violinist, versus a cello player with similar stats. My advice is to apply being yourself. That’s what the college wants. If yourself is not what they’re looking for, then you wouldn’t be happy there anyway.</p>
I don’t disagree with this. EA/ED is unlikely to make a difference unless you are near the borderline between acceptance and rejection. In the 2nd half of the paper, the author discussed applying game theory to choose which college(s) to use the EA/ED boost, emphasizing colleges where the EA/ED boost is most like to make a difference, rather than following the conventional advice of choosing a first choice, regardless of how low the chance of acceptance is. The problem with this approach is a large portion of students have a poor idea of what their actually chance of acceptance is, often because they emphasize GPA/SAT over more subjective parts of the app, including the ones you mentioned. For example, an applicant might think he has decent chance of acceptance because his GPA and SAT are near the mid 50% of the class. Instead this group has an overall admit rate in the single digits at Princeton. Most apps in all stat ranges get rejected. The ones who get accepted are often ones whose app has impressive not-stat qualities, including unique ones that helps to stand out from the tens of thousands of apps (not just being a hook).</p>
While the overall acceptance rate is near 8%, the vast majority of apps do not have a ~8% chance of acceptance rate. I’ll use a real world example from the Stanford RD decision thread, which I previously analyzed. Posters in the thread with a high SAT & GPA that also had an impressive on state+ level ECs or awards had a near 100% rate of acceptance. While unhooked posters in the same stat range who had ECs or awards that were not noteworthy on at least a regional level had a near 0% chance of acceptance. Two groups in the same stat range had a completely different chance of acceptance, both no where near the overall acceptance rate for their stats, and what stood out among the ECs/awards was not simply random and/or unknown. </p>
<p>As stated in my earlier post, the apps who are most likely to benefit from EA/ED are the ones with a borderline chance of acceptance. The study implies that this group can have a noteworthy boost in chance of admission. I brought up the study in response to comments implying EA/ED did not boost chance of admission. As stated above, this borderline chance of acceptance not much below 50% is certainly possible when the overall rate of acceptance is 8%. As I implied in my first post, the OPs chance of acceptance is most likely not 8%, even though the overall rate is 8% for his stats. Instead it depends on more subjective parts of his app. The OP didn’t give much detail aside from saying has ECs were weak. That obviously may hurt, but it is not necessarily a dealbreaker. For example, maybe he had a unique background & experiences, noteworthy awards, excelled in a unique passion, showed some amazing qualities on his app, or something else that suggests he is likely to go on to do amazing things and pushes him into the borderline range. It’s difficult to give much detail without knowing more details about the OP.</p>
<p>Okay, so staying on topic and not extrapolating to other students: with regard to the OP, SCEA will not significantly boost his/her chances of acceptance. That’s my point. I have definitely seen people at Princeton who should not be there. Many of these people have no “hooks”. I have no idea how they were admitted - you can only generalize so much from statistics. There are also people who had perfect GPA/SAT who cannot make it here - I’ve seen people with much lower statistics do better at Princeton than others. The point is, there is a range of GPA/SAT scores at Princeton and statistics of the past can only tell you so much about whether you, individually, have a high or low chance of getting in.</p>
Like I said in my earlier post, we only know he has an average GPA/ACT for Princeton apps and weak ECs. Only considering that small portion of his app, odds are that he has a low chance of acceptance, with or without SCEA. However, it’s possible that the rest of the app will contain something that helps him stand out, such as the list in my previous post – “unique background & experiences, noteworthy awards, excelled in a unique passion, showed some amazing qualities on his app, or something else that suggests he is likely to go on to do amazing things.” You could say SCEA is not a significant boost in the same way one might say taking the SAT/ACT again and increasing combined score by 100 points (or ACT equivalent) won’t give a significant boost. Neither SCEA or +100 SAT are likely to change the decision unless the rest of the app contains some combination of details that pushes him to the borderline range. Either could be an important boost for a minority of apps that are borderline. </p>
<p>
In your post above, you implied that we cannot accurately estimate chance of getting in, and as such cannot accurately determine who is a borderline candidate that would have the greatest benefit from a SCEA boost. If you believe this, then it follows that SCEA may have a notable impact on the OP since we cannot accurately determine if he is a borderline app for which SCEA may change the decision.</p>
<p>You do realize that people with average GPA/SAT and weak ECs get into Princeton all the time, right? Even if they’re rich, white kids with no other strong point in their application. Princeton admissions is so much of a crapshoot that nobody really knows the chances of an applicant. It doesn’t matter that students with OP’s stats have had low chances of acceptance - perhaps Princeton is looking for exactly that type of student this year - a kid who is so full of himself that he has room to improve. That’s why all of us come to Princeton in the first place. </p>
<p>Again, with the extrapolations. I did not imply that in my above post - I only explicitly state that in this post. I can believe that PAST statistics say nothing about a certain applicant’s chances of being admitted WITHOUT accepting that SCEA has a “notable impact” on borderline applicants. These are not related events, unless you already accept the premise that SCEA boosts a borderline applicant’s chances of getting in. I question the accuracy of these statistics, which by the best estimate, were obtained more than a decade ago!</p>
If you’ve read my posts in this thread, you would know the answer to that. In the majority of my posts in this thread, I’ve said that the OP needs some way to stand out from the 10,000+ other apps with similar GPA and test scores, but it doesn’t need to be ECs. I’ve given lists of other possibilities besides ECs in several other posts, including the one you just replied to. The data published by Princeton indicates about an ~8% acceptance rate for students with the OPs GPA and ACT score. Obviously having weak ECs is going to drop the overall admit rate for this GPA/ACT/weak EC/unhooked group to significantly below 8%.</p>
<p>
I mentioned analyzing the Stanford RD thread earlier. I was able to predict admission decision correctly for about 90% of the posters in the thread, with just the limited information in a single post; and the few that I missed, I usually predicted to be borderlines. If I had access to essays, LORs, school history, and the rest of the app; I expect I could increase well above 90%. Admissions decisions for selective colleges are not just a random crapshoot where one has no idea whether their chances are good or bad.</p>
<p>
The study was published in December 2010, so the authors who were professors at Harvard and Stanford, certainly did not think the older data no longer applied to recent college admissions, and most selective colleges no longer give a boost to EA/ED. While certain aspects of admissions have changed in recent years, the core criteria colleges care about has not had great changes. Colleges still care about GPA, SAT, ECs, LORs, etc. And many still care about their yield rate and whether they are a first choice vs backup/safety, both of which can be enhanced through an early decision program, particularly SCEA or ED.</p>
<p>And as I said, standing out can be an inherent characteristic of personality - something that is unalterable and innate, something that does not “stand out” to you, but does to admissions in a given year. Weak ECs do not necessarily drop the acceptance rate. Perhaps Princeton could be looking for intellectuals in a given year - those who spend their time excelling at school and, in their time outside of school, do something that isn’t really considered an EC, like read a lot. In this view, this applicant has high grades (but average for Princeton) but weak ECs. If Princeton was looking for a future Comp Lit academic, then we have a winner here! That’s why, like I have previously said, I see so many unhooked, academically average students who, in your view, have very low chances of being accepted! </p>
<p>Forgive me if I’m skeptical about your ability to so accurately predict an applicant’s chances of being admitted (why one would spend so much time analyzing an RD thread to get a (reliable) accuracy rating is beyond me), but if you can predict that with any accuracy, you should become an admissions officer! Because then you already know what a college is looking for! Though your incoming class might be a little lopsided - perhaps 90% engineers, all with great GPA/SAT and outstanding ECs, because you overlooked the average (for Princeton) GPA applicants with “weak” ECs. Or, perhaps you would fail to have a compete orchestra because you failed to include any of the “average” violinists. </p>
<p>The fact that the professors teach at Harvard and Stanford does not make their analysis right or applicable. That’s just like saying Karl Marx says socialism is the right form of economy and so we should orchestrate a government takeover of all large businesses because Marx is in any way authoritative. I am challenging the very basis of the study. It was conducted even before Harvard and Princeton reinstated their early action processes.</p>
I’m not going to waste the time to do an analysis to show something as obvious that having weaker ECs than most apps decreases average chance of admissions at selective colleges. Instead I’ll just point to the Princeton CDS in which they rate ECs as “important.” In recent years, they’ve always rated ECs as either “important” or “very important”.</p>
<p>
Out of curiosity I entered the following info into a spreadsheet for each poster – GPA, test scores, course rigor (1-5 scale), ECs/Awards (1-5 scale), and race. I then tried to make up a set of rules/weightings to predict the decisions. My results are not difficult to replicate. With just the following 3 rules, one can predict the vast majority of posters accurately.</p>
<ol>
<li> Posters who have a less than a 3.8 GPA (ignoring freshman year) or less than a 2000 SAT(or ACT equivalent) get rejected. </li>
<li> Among the remaining, posters whose EC/Awards rate 0-2 (below regional level) get rejected.</li>
<li> All remaining posters get accepted.</li>
</ol>
<p>Obviously these rules would not apply well to all apps since the posters on CC are a unique subgroup that tends to be high-stat, well-informed students who attended quality high schools and took rigorous courses. There is also likely a bias towards posting accepted results and not posting rejected results. For example, most posters in the thread were accepted, while Stanford’s overall acceptance rate was under 6%. Nevertheless, the point remains that the admissions decisions did not appear anything close to random. Instead they appeared to be highly predictable among the unique CC poster subgroup.</p>
<p>
It doesn’t make them correct, but I’d consider the opinion of a joint effort between Harvard and Stanford professors that is published in a peer reviewed journal, which is read and criticized by other college professors, to be more reliable than the opinion of a random, new poster on CC with the beliefs described earlier about ECs.</p>
<p>You mentioned the data was before Princeton’s recent SCEA program. The 2012-13 Princeton CDS indicates that Princeton considers “level of applicant’s interest”. Applying SCEA indicates strong interest since SCEA is a single choice program where the student can only apply early to one school, which is usually his top pick. Princeton’s website specifically mentions SCEA being for students whose first choice is Princeton.</p>
<p>Actually, the SCEA programs have numerous loopholes, allowing the applicant to apply early to other schools which are:</p>
<ul>
<li>Public schools</li>
<li>Rolling admissions schools</li>
<li>Schools with an early deadline for scholarships</li>
<li>Schools outside the US</li>
</ul>
<p>However, these are schools which HYPS generally believe are not likely to win a cross-admit competition with HYPS, so the SCEA restriction fulfills its intention to cause the applicant to reveal his/her first choice among schools that HYPS believe may win a cross-admit competition with them (including each other).</p>
<p>If you’re going to take anything Princeton writes at face value, Princeton also writes the following:“All applicants to Princeton, whether they apply early action or regular decision, receive the same comprehensive, holistic review. Those who apply early gain no strategic advantage; the only advantage is one of convenience. If you know that Princeton is your first choice, then it may make sense for you to apply early.”</p>
<p>So then we’re done and your point is moot, right? See, if this were just an opinion, yes, I would credit the Harvard and Stanford professors. However, mine is empirical evidence gathered over the period of a year and a summer with dozens of actual Princeton students. Yes, you can crunch the numbers all you want, but you can’t beat actual case studies. The Harvard and Stanford professors also have little to no insight in the admissions process. Professors are hardly included in the admissions process. Therefore, their argument is reduced to mere numbers, which can always be beat.</p>
<p>FYI, my roommate had a 3.75 GPA and 1900-range SAT. ECs were standard for Princeton. My friend had the grades, but only one or two ECs in which he showed any sort of dedication. By your system, both would have been rejected. A reduction to numbers is not what college admissions is about or what it needs - only by a qualitative approach can you construct a vibrant class. Have you, out of interest, tried your method with Princeton admissions? That might be informative.</p>
True, but the point remains that SCEA usually indicates first choice, as I wrote.</p>
<p>
I’d be very surprised to see a selective college explicitly say that they favor early apps over regular, in a similar way to how colleges whose stats exhibit clear “Tufts syndrome,” with rejecting more qualified apps who are using the school as a safety, say that they have never practiced “Tufts syndrome.”</p>
<p>
Lets compare the two data sources:</p>
<p>Source 1 : A peer reviewed study involving a survey of ~2400 applicants, calculating rate of acceptance after factoring out strength of apps, many types of hooks, and more than a dozen factors in total</p>
<p>Source 2 : Knowing some students and presumably talking about their acceptances and or college app with some of them in such a way that you can tell if they had better chance of admission by applying with SCEA than they would with RD</p>
<p>Source #1 sounds more reliable to me. I also question whether talking to dozens of students gives you more insight in the admission process than Harvard/Stanford professors would have by researching admissions in multiple studies and writing a well known book on the subject. And of course, they also talked to students. I do accept that it’s possible that Princeton differs from most schools in the degree of early boost. However, I very much doubt that knowing a small subgroup of students would provide enough information to draw any conclusions. </p>
<p>
Was your roommate a poster in the Stanford RD thread on CC? Note that the post you quoted stated, “Obviously these rules would not apply well to all apps since the posters on CC are a unique subgroup that tends to be high-stat, well-informed students who attended quality high schools and took rigorous courses. There is also likely a bias towards posting accepted results and not posting rejected results.” The rules are also not expected to be the same for all CC RD threads. Stanford is particularly big on showing a passion and excelling in it, which correlates a higher weighting than most colleges on ECs/awards. In contrast Princeton has a reputation for placing a greater emphasis on stats than Stanford and a lower emphasis on ECs than Stanford (not zero weighting on ECs).</p>
<p>So we’re not going to accept Princeton’s statement that it gives no advantage to SCEA applicants but we are going to accept its claim that it gives “important” to “very important” weight to ECs? So I can go through something and pick out the points I like and use those and just ignore the rest? Great way to write a paper or a thesis, by the way. </p>
<p>Actually, only 355 applications to Princeton were surveyed, with 70 of those being Early Action. Again, the data is over a decade old, when Princeton’s overall admit rate was 10.8% and Harvard’s was 11.3%. I do not claim that I have greater insight into the admissions process - I am simply saying that I have empirical evidence gathered from something that is more than just statistics. Viewing students by their numbers is the most repugnant form of dehumanization in college admissions. I can humanize those numbers that the study crunched and, from what I have seen so far, there is no reason to suggest that SCEA gives any significant advantage to students - if they got in SCEA, they were going to get in RD anyway. </p>
<p>I frankly do not care which source sounds more reliable to you. To me, a set of numbers on a sheet of paper is magnitudes less powerful than a series of faces, each with a story to tell. I can also spot when statistics simply no longer hold or are just purely incorrect. I see no reason to believe that the study you cite holds any longer, based upon pure observation. Now, that study is the only one of its kind to report those findings and none in recent years have confirmed it. If I cited data from over a decade ago in my field of research, I would be laughed out of the room. College admissions is a rapidly changing arena and it is much harder to get into college now than thirty, twenty, or even ten years ago. These changes reflect paradigm shifts that may make data from even a decade ago obsolete. </p>
<p>How applicable do you think your method is to Princeton applicants? If it does not apply to Princeton, then why mention it in a discussion about Princeton SCEA? (That’s a rhetorical question, by the way)</p>
In your earlier post, you wrote that the Harvard and Stanford professors have little to no insight in the admissions process. If you still think they have “little to no insight”, then it shouldn’t take much insight to have greater than little to none.</p>
<p>
We are not talking about football players being accepted with stats at the bottom of the class that are obviously weaker academically than the vast majority of the class. Instead we are talking about borderline candidates that almost, but not quite meet the acceptance threshold. Do you really think that you can tell in a casual day-to-day conversation whether they are a borderline candidate that would have a greater chance via SCEA or RD, especially since you called admissions a random crapshoot where you don’t know who is likely and unlikely to be admitted?</p>
<p>
The study was published in December 2010. The Harvard and Stanford professors publishing the study obviously thought the data was still applicable and were not laughed out of the room. Instead the study has It has been cited in dozens of other studies since then, some in recent months.</p>
<p>
I’ll quote a portion of the earlier post – “Nevertheless, the point remains that the admissions decisions did not appear anything close to random. Instead they appeared to be highly predictable among the unique CC poster subgroup.” The point was that admissions did not appear to be a random crapshoot, as you previously stated. Instead using a simple set of rules one could accurately predict the vast majority of the posted results. Sure it’s theoretically possible that Princeton makes their admissions decisions completely randomly among posters with similar stats, such as reviewers throwing the apps in the air and giving high ratings to the ones that fall on top, but I think it’s far more likely that the vast majority of weighting in Princeton admissions decisions are not random and instead applicant strength can be roughly estimated by looking a a combination of stat and non-stat qualities, just as Stanford decisions did not appear to be random and instead could be predicted well by looking at a combination of qualities (among CC thread posters).</p>