<p>
Correcting myself, I think that should be ad infinitum. Maybe it would be easier to say ad nauseum.</p>
<p>Bovertine, I was referencing Placido240’ calculation method of using one’s test scores to determine whether a school is a safety or a match school. Of course other factors come into play, which is why I suggested that at the top, the ceiling effect fails to distinguish clearly between multiple competing applicants and thus this formula is more useful in the center of the range than at the upper tail.</p>
<p>
That’s sort of what I thought, but then why mention rank? That’s where I got confused.</p>
<p>
</p>
<p>Thinking ahead a step, I guess: rank is also easily quantifiable and should also probably be considered within the formula – those students who had high test scores and low GPAs/class ranks are often looked upon as slackers who may not make a strong effort in college, especially if it’s their first time away from home.</p>
<p>
It is easily quanitifiable within the context of a single high school, but as I pointed out it is not unifiorm. Rank is probably useful to actual adcom that have the time and information to parse it all out. But not all that useful for a universal formula. THat’s why when it comes to GPA/rank determinations for application decisions almost everyone relies on their own school’s Naviance tables. I suppose it would be possible to construct that kind of formula for a given school, in fact I bet that’s sort of what some high school GCs do.</p>
<p>Everyone takes the same SAT/ACT. That’s why so many people seem to love these tests so much. I agree that is a valid consideration.</p>
<p>
The problem is that college admissions at <em>most</em> colleges (and certainly the “top” schools referenced in the OP) – is NOt a linear comparison or contest. The ad coms don’t care that you “tie” or “exceed” anyone. </p>
<p>The app process works more like this:</p>
<p>First there is an initial screening in which the students whose stats means they are clearly unqualified for admission are eliminated. There is probably some score cutoff - or more likely a admission index (AI) formula based on scores combined with GPA – but it is much, much lower than 2200. The AI formula is probably the reason my high-GPA daughter was never out of the running at her reach colleges despite her <25th percentile test scores. However the respective colleges figured that number, her GPA & class rank combined with test scores probably put her closer to the middle of the pack.</p>
<p>After the initial screening is done, everything moves to the holistic end of the process. The numerical AI score (whatever it is) doesn’t matter that much, because by definition everyone remaining in the deck has a “good enough” GPA & test score. So now is the time they look at essays, recs, EC’s, etc – and for each applicant make something of a gut level determination along the lines of “YES!!!”, “yes”, “probably”, “maybe”, “not sure”. At a highly selective college, “maybe” and “not sure” will never make the final cut" – and it among the “yes” and “probably” group the ad coms will probably come back and take a 2nd look at that test score and GPA. </p>
<p>That’s why a very high stat student can end up in the reject pile – they made the first cut, but on the 2nd read their application wasn’t enough to get them past “maybe” – maybe that kid was very introverted with no ECs, teachers didn’t know her well enough to write much of a rec, and the essay didn’t convey much of use. </p>
<p>And that “YES!!” student can very well be someone with rather low stats, but maybe they have everything they need in the other areas – first generation student, compelling story, enthusiastic recs, some very significant accomplishments in high school – and everyone falls in love with the kid and wants him in, and is willing to look past the somewhat weaker numbers, as the application makes it so clear that the kid has had a really tough life. It is that “yes” and “probably group” where you might have one kid chosen over another because of better stats - and that probably depends on which other kids are being considered for admission that day, or that hour – because that’s what the ad coms are thinking about when they make the cut. That is if they’ve got apps from Shelley and John and Marie sitting on the table and they know they have to winnow it down to 1 out of 3, and they’re all about equal, if Marie has the best stats she might win that round. But no one is going to remember or point out that Marie’s stats are not nearly as good as David’s, the kid from New Jersey they decided to waitlist 4 days ago – no one remembers David any more. </p>
<p>Think if it like a deck of cards (an analogy I alluded to above). Let’s say you take a deck of cards and you first go through the deck and eliminate all cards with a face value of 6 or below. Now your deck consists entirely of 7-K. Now you go through the deck and pick cards 3 at a time, and out of each 3 you pick the best – but to complicate things a little, you add +2 for every heart, and subtract -1 for every club. (So a 7 or 8 of clubs will never win). Overall, the odds are that you are going to end up with a lot of face cards, with more hearts than clubs. But the 9 of clubs might make it if somehow or other that card gets selected along with the group including the 7 & 8 of clubs-- and it is going to be hard to beat the jack of hearts whenever it comes up. Since face cards always beat numbers (except for hearts) – your final selection is going to consist mostly of face cards, which may lead most people to believe that you need jacks or better to win – but by the rules of the game, a 7 of hearts also has a reasonable chance of getting in.</p>
<p>OP and others are missing how competitive the athletes make it on a 2200 kid or 2030 kid applying to the top schools (top 20 schools)…gpa/SAT standards fall off fast at UMich, Cal etc (top 30 schools)</p>
<p>The 2200/4.0 kid wanting to go to MIT is up against some athletes who are 2400/4.0 and in the 70% range of their sport. Now go over to HYPS and other top schools that value athletics and you’ll find athletes with 2300/3.9 and at the top of their sport.</p>
<p>And these athletes can make up 20% of the student body. This is a big reason why at 2200/4.0 kid with no sport or outstanding ec gets rejected.</p>
<p>As to tk’s formula (#152) – I think that is useful for targeting and is pretty close to what I did years ago with my son. (He was NMF with top quartile stats for most colleges, mid-range for Ivies). I just had an ad hoc rule that went something like this:</p>
<p>*top quartile stats + admission rate over 35% = match
*top quartile stats + admission rate over 75% = safety</p>
<p>I didn’t engage in the same analysis with my daughter because she didn’t have “stats” that fit so neatly. So I had to really take a more individualized look at stats at various colleges. For example, I identified UC Berkeley as a “match” (not a “reach”) because even though her test scores were weak for that campus, she had “ELC” status (at that time afforded to all students identified as being in the top 4% of their high school class at the end of junior year) – and there were published stats showing that Berkeley admitted roughly 70% of all ELC students. (The less selective UC campuses would admit 95%+, making them safeties). </p>
<p>Outside of the public system I had to look to more holistic factors – but it is possible to identify some. I’d note that there are some holistic factor that really can’t be quantified in any way at all (such as “wonderful essay”) – but it is often possible to get data to give an indication as to whether or not other qualities may be a tip factor for admissions. For example – if an LAC is known for its strength in arts and humanities, but they’ve just finished construction on a brand new science center – it’s a pretty good bet that they will look kindly on an applicant who participated in Intel, is interested in biochemistry, and writes about that new science center in that “why this college” part of the application.</p>
<p>
</p>
<p>Quite an interesting analogy there, Calmom! It certainly appears probable that some sort of sub-group comparison takes place, because quite a lot of students find themselves accepted to schools with lower admission rates and yet rejected or waitlisted at schools with much higher acceptance rates. And hardly anyone, except the occasional athlete or URM, gets in across the board.</p>
<p>It could also turn out that some lessor-but-nearly-equal kids are ultimately let in because their area rep is underperforming in his or her usual acceptance ratios. One must keep the hired help happy! Who would want to advocate for a whole region and then see 98-99% of “their” applicants rejected?</p>
<p>@calmom</p>
<p>We defined a safety as one where S’s SATs/ACTs were above the 75% percentile for all students </p>
<p>“That’s not an appropriate way of identifying a safety. Test scores do NOT play that big of a role.”</p>
<p>In reality I was highlighting and simplifying what we actually did. My son developed ten criteria which were: excellent academically; very small size (which emphasized student-faculty ratio as well); liberal arts; open curriculum; diverse (racially, socially, economically, etc.) student body; collaborative as opposed to competitive environment; isolated campus but with reasonable access to a city (but not in a big city); well developed study abroad program; reasonably close to home (within 5 hours - we are lucky since we live in the northeast and could impose such a criteria); and offered four specific majors (math, computer science, physics and Spanish). We considered every college starting with these criteria. So a college like Franklin & Marshall that we might have looked at otherwise didn’t even make our list to consider since they have no computer science major and a minimal CS department at this time. We also looked at every ranking and rating we could fine including all the quantitative ones as well as subjective assessments. With respect to quantitative rankings, we looked at how consistent the rankings were between different lists (e.g. US News compared with Forbes). For example, Williams or Amherst will have very similar rankings or ratings no matter what quantitative index you use. Conversely, Trinity College or Hamilton will have much more variation. In addition, we absolutely looked at acceptance rates. We established a rough taxonomy where schools with acceptance rates: below 30% were treated as reach; 30 to 39% were regarded as targets; and those with rates of 40% and above were considered safeties.</p>
<p>No I will return to my original point. Using standardized test scores, there is a generally consistent and linear relationship as I originally pointed out. The probability of acceptance increases as the difference between the college’s average scores and the applicant’s scores increases (positively). There are, as I originally stated, exceptions all along this line. It is clearly not the only criteria but it seems to work reasonably well. It certainly is an appropriate way of developing some useful information. Of course, everything else can come into play and it has become much more diffficult to reliably classify potential colleges into neat categories.</p>
<p>I see a lot of talk about the threshold for the ivies being 2200. Where does this place unhooked applicants in the mid 2100s? Do they still have a decent shot?</p>
<p>@pizzagirl</p>
<p>Quote:
That is one reason why grads of such places go to such lengths to convince anyone who will listen how great the place is. You got to reel in a few more suckers or else you have less than 2,000 total students. It gets embarrassing at some point. </p>
<p>“Really? The top schools have to scare up applications, because no one is otherwise willing to pay full freight and go? I don’t believe it. In fact, I believe that if, tomorrow, all the top schools banded together and said that they were no longer offering financial aid, sticker price or don’t show up, they’d still be able to fill their freshman classes with relatively little impact to their overall stats.”</p>
<p>I completely agree with you.</p>
<p>It could also turn out that some lessor-but-nearly-equal kids are ultimately let in because their area rep is underperforming in his or her usual acceptance ratios. One must keep the hired help happy! Who would want to advocate for a whole region and then see 98-99% of “their” applicants rejected? </p>
<p>Are we talking about Alums or adcoms here? I understand alums get pretty annoyed too when none of the people it their region get picked after they spend so much time interviewing them.</p>
<p>
Lesser but nearly equal on what basis? SATACT score again? A kid who wins the IMO and scores a 2350 is certainly not “lesser but nearly equal” to a kid who scores a 2400 and has virtually no other ECs or accomplishments. And IMO is just one example of a mulititude of possible evaluative criteria. You seem unable or unwilling to grasp the notion of holistic admissions other than as some forced and inequitable handicapping system for “lesser” applicants.</p>
<p>To clarify - if by “lesser but nearly equal” you merely mean “with somewhat lower but close standardized test acores” then I have nothing to disagree with. But “lesser”, at least to me, implies something beyond that.</p>
<p>A score of 2200 means a lot more in the Early admissions round when 25% to 40% of the kids get in at these elite schools (only Georgetown, Columbia , Yale and Stanford and maybe MIT are under 20% in the ealry round; all the others are higher). Suddenly, in April all these 2200’s become leppers in the 6% to 20% regular decision round. What I take from this is any counselor who doesn’t urge a senior to apply early is committing malpractice.</p>
<p>I’d posit that class rank is not necessarily a good discriminator.</p>
<p>Schools that have selective admit programs within a regular school may have 25% of the class selected from among the top 1-3% in the school system, with dozens of these kids making NMSF, average SATs in the group well over 2100, etc. If these kids went to their local schools, they’d be superstars. At these programs, they do very well, but are one of many.</p>
<p>When the school system doesn’t rank, but instead gives bands of GPA ranges which include all students in the school, you are then comparing the kids who never took an honors class and have a 3.8 to the ones who did full IB plus 8-10 APs and have a 3.8. It’s apples and oranges, and the GC saying the IB kids had “toughest schedule” on the letter doesn’t make a whole lotta difference – unless one can suss out from Naviance, CC, etc. which colleges like kids from your kid’s program.</p>
<p>We had a pretty good idea where the rejections would fall with both kids because some of the schools on their lists wanted to see that mega-GPA no matter what.</p>
<p>
</p>
<p>Sure, but let’s take two applicants with roughly the same ECs and accomplishments, one has a 2400 and the other a 2350 or even 2300. By a pure technical match-up, the 2400 should win, but the difference overall is not all that huge. The 2350/2300 could well get the nod simply because his/her area rep is underperforming and needs to get a fair share of advocate “wins” too, to remain enthusiastic. One might call this a win in favor of geographic diversity when, in fact, it’s more a matter of keeping peace in the office.</p>
<p>tk - I love your general approach, but I’m much more conservative than you.
We had schools where my kids were in the 75%th percentile, overall acceptance rate was ~ 34%, and I still thought of those as low reaches or high matches, not safeties. If anything, I started adding in more safeties to S’s list, which turned out to be useless when he got in a reach school … but wow, I think I’m way more conservative!</p>
<p>I also wouldn’t consider any school a “safety”, no matter how strong the stats, unless the school clearly accepts a strong majority of applicants. For me that’s at least 70% - I wouldn’t quibble with 60%… but if they are turning away more than half, that probably means they are turning away quite a few who have impressive stats. </p>
<p>I think the part that many CC’ers miss is that in addition to holistic practices, many schools are wary of accepting students that they don’t believe are likely to attend. That college that accepts 40% of its applicants is not simply grabbing those with the high end stats – they are trying to protect their yield.</p>
<p>Misjudging a “safety” is a problem, because when a parent or kid takes a school for granted, they are less likely to focus the attention needed on the application that may be the difference between acceptance and waitlist. (My guess is that most yield-protecting decisions are waitlists rather than outright rejection, with the ad com figuring that if the kid really wants to come, he’ll call).</p>
<p>“one has a 2400 and the other a 2350 or even 2300. By a pure technical match-up, the 2400 should win . . .”</p>
<p>Not so, most schools look to see if the SAT/ACT is above a certain threshold - a 2350 and 2400 are both above it. Adcomm does not care by how much - it is pass/fail. They move on to other non-statistical factors.</p>