A Statistical Way to Look at College Admissions

Applying to College
1

WHAT DO YOU ALL THINK? PLEASE READ!

I am considering putting together a guide helping students find “reaches,” “matches,” and “safeties” using the same standard deviation patterns that the SAT uses in order to help students evaluate their chances.

Tell me if you like these systematics & if it might be something I should pursue.

As I am sure that makes no sense at all, let me give you an example of Northwestern. For the sake of the data, I will only be using grades & SAT scores (which are a good indicator, although they do not guarantee admission). If there is any way I can quantify EC’s, I will try to do so.

But anyway, Northwestern accepts 30% of their applicants. In other words, you have to be in the 70th percentile of their applicant pool to be accepted. Based on the SAT, 70th perentile is an 1130. Therefore, a qualified candidate for Northwestern will score an 1130 or above for the two categories (rank + SAT’s).

Some of the scaled chart is below (remember, the chart is scaled so the 50th percentile of their applicant pool receives a 500):

Class Rank (Top X%) Scaled
1% 640 - 720
2% 630
3% 580
4% 540
5% 510
6% 500
7% 490
8% 480
9% 470
10% 460
11% 460
12% 450
13% 440
14% 430
15% 430
16% 420
17% 410
18% 400
19% 400
20% 390

SAT Score Scaled
2400 700
2390 700
2380 690
2370 690
2360 680
2350 680
2340 670
2330 670
2320 660
2310 660
2300 650
2290 650
2280 640
2270 640
2260 640
2250 630
2240 630
2230 630
2220 620
2210 620
2200 620
2190 610
2180 610
2170 610
2160 600
2150 600
2140 590
2130 580
2120 570
2110 560
2100 550
2090 540
2080 530
2070 520
2060 510
2050 510
2040 500
2030 500
2020 490
2010 490
2000 480
1990 480
1980 470
1970 470
1960 460
1950 460
1940 450
1930 450
1920 450
1910 440
1900 430
1890 430
1880 420
1870 420
1860 410
1850 410
1840 400
1830 400
1820 390
1810 390
1800 390

EXAMPLES:

Bob w/ a class rank in the top 4% & an SAT score of 2180 would earn 540points for his class rank and 610 for his SAT score. Therefore, he has a 1150 combined score & can consider Northwestern to be a “match” (by match I mean a school that he is qualified for, but may or may not be accepted to).

Jim is the valedictorian in his class & receives 680 for this category. His SAT score was 2250 so he receives 630 for this category. With a score of 1310, he can consider Northwestern to be a “safe match” – as long as his EC’s are not particularly dismal.

Billy was a slacker in high school & is ranked in the top 18% of his class, earning 400 points for class rank. But he scored a 2320 on the SAT, earning 660 points for the SAT’s. His total is 1060, which will make Northwestern a “slight reach” to “reach.”

WHAT DO YOU ALL THINK?

2

<p>Interesting, but how do ECs actually factor into this?</p>

3

<p>In this particular instance (I am still working on it), EC's are not factored in. If there is some way we could statistically figure out what % of applicants have what sort of EC's, it would be somewhat possible.</p>

<p>Nevertheless, the point is not to guarantee the accuracy of your admissions decision. The point is for students who don't have the time & resources to spend on college admissions to be able to quickly add up two concrete numbers & get a fairly accurate perspective on their chances at any given school.</p>

4

<p>and one more thing, the reason the scales do not go up to 800 is that too many of the applicants fall in that range to justify a "perfect score" in either of the categories.</p>

5

<p>bump...for tomorrow :)</p>

6

<p>anyways..stambliark4 noice job. props buddy</p>

7

<p>This makes sense, and you could adjust it for hooks (e.g. by adding points for URM, recruited athletics, extraordinary ECs and other accomplishments). On the other hand, doesn't it have to be recalculated for every school?</p>

8

<p>this is extremely flawed...
you are assuming that the applicant pool to a highly selective school represents an even distribution of the entire range of SAT scores and class rank, while this could not be further from the truth.</p>

9

<p>The SAT scores and class rank are distributed over the SAT scores and rank of the applicant pool - a 2040 is not the median SAT score for the overall population.</p>

10

<p>it isn't extremely flawed. if we had a large enough sampling we could see whether or not acceptance percentages would be proportional to RANK+SAT.</p>

11

<p>The problem as I see it is that the objective criteria (sat scores, gpa, rank) get you over the first hurdle. It is usually the subjective factors (res, ecs, essays) that tend to move applicants past being a series of number. Selective private schools are more wholistic and focused on building a class vs being at a large public school which is more purely numbers driven.</p>

12

<p>Lets say some one is ina gifted school. Had almost all A but was ranked 25/50. Gets a 2400 on SAT...etc etc you get my point right?</p>

13

<p>This is pretty accurate for the majority of schools probably - b/c the SAT scale is based on a normal (gaussian) distribution which is describes most applicant pools where the majority of applicants fall in the middle. However for some self selecting pools at schools such as MIT, UChicago, etc, this would not be as accurate</p>

14

<p>GPA and rank mean wildly different things at different schools. There are private schools where over 30 percent of each class are admitted to ivies and ivy-like schools and public schools where the top 5% average 1100 on SATs.</p>

15

<p>That would not work for my school where 110 out of the 600 graduating seniors had GPA of 4.0 or higher.</p>

16

<p>^^^^^^
holy *****</p>

<p>Good work, but I think something like this would only be effective for state schools- most of them.</p>

17

<p>70% percentile of Northwestern's and 70% percentile of SAT is very different</p>

<p>Everyone takes the SAT.. mental retards, slackers, stupid people, careless people (their parents made them take it), etc.</p>

<p>Northwestern's pool is usually the top of the class...</p>

<p>So its 70th percentile for the 20% of your class..</p>

18

<p>Before I say anything, I want to remind you that this is not designed to serve as an absolute indicator of your decision in April. It is to define general guidelines that help people that are not on CC assess in a matter of seconds whether a school is a reach, match or safety.</p>

<p>Alright let's see if I can address what everyone has said.</p>

<p>Cevonia -- yes, it does have to be recalculated for every school, but it doesn't take very long (as long as you have a good set of data similar to what northwestern provides).</p>

<p>"you are assuming that the applicant pool to a highly selective school represents an even distribution of the entire range of SAT scores and class rank, while this could not be further from the truth."</p>

<p>It actually does not reflect that at all. If you are familiar with percentage patterns on the SAT, you will realize that it reflects a somewhat exponential curve. Also, if you look at the class rank data, it is clearly shown that there is not an equal of number of applicants at each category. Based on the chart, 50% of the applicants have a class rank of top 6% or higher.</p>

<p>"The problem as I see it is that the objective criteria (sat scores, gpa, rank) get you over the first hurdle. It is usually the subjective factors (res, ecs, essays) that tend to move applicants past being a series of number. Selective private schools are more wholistic and focused on building a class vs being at a large public school which is more purely numbers driven."</p>

<p>If I were to create a systematic admissions calculator (so-to-speak), yes, these many factors would have to be included. But as far as looking at a general interpretation if a school is a reach, match, safety for the avg-nonCCer, my hope is that it is a decent indicator. If we were able to scale EC's in a similar fashion, that would be great -- but it would take much more data than what is available on a college's web site.</p>

<p>"This is pretty accurate for the majority of schools probably - b/c the SAT scale is based on a normal (gaussian) distribution which is describes most applicant pools where the majority of applicants fall in the middle. However for some self selecting pools at schools such as MIT, UChicago, etc, this would not be as accurate"</p>

<p>At a self-selecting school such as MIT or Chicago, the average applicant would have much higher statistics. Therefore, a median applicant's SAT Score at Northwestern would be much lower than the median SAT Score at MIT. </p>

<p>"GPA and rank mean wildly different things at different schools. There are private schools where over 30 percent of each class are admitted to ivies and ivy-like schools and public schools where the top 5% average 1100 on SATs."</p>

<p>Yes, at a top-notch private or public school (such as the school you describe above) this would be flawed. But for most of us that attend schools that do not send 30% to Ivies, my hope is that it is fairly accurate.</p>

<p>"That would not work for my school where 110 out of the 600 graduating seniors had GPA of 4.0 or higher."</p>

<p>This could mean that your school is highly competitve, or it could also indicate grade inflation. Maybe we could take into account the competitiveness of the school when scaling the scores?</p>

<p>"Good work, but I think something like this would only be effective for state schools- most of them."</p>

<p>Do you not think that it still serves as a rough reach, match, safety indicator? Is it completely flawed or just not entirely accurate?</p>

19

<p>I think using GPA would probably be more effective than class rank</p>

20

<p>Instead of making this stupid inaccurate charge, you can be studying and raising your class rank and SAT...</p>

<p>On your college app, are you gonna put "I didnt study much but I did make a chart on how I can get in it your college.. I'm amazing... According to it, I am a match.. So accept me. And consider the chart my ECs.. :) "</p>