Detailed admit rate analysis for Stanford admissions

Colleges & Universities Stanford University
1

<p><a href=“http://www.stanford.edu/dept/uga/pdf/CounselorsNL_Web.pdf[/url]”>http://www.stanford.edu/dept/uga/pdf/CounselorsNL_Web.pdf&lt;/a&gt;&lt;/p&gt;

<p>Based on the statistics provided in this Stanford press release, I decided to do some analysis. Let us say there are 3 students (X, Y and Z) who applied to Stanford with slightly varying performance metrics. Based on my analysis below, Student X had a 54% chance of getting selected for admission, Y had a 42% chance and Z had a 32% chance.</p>

<p>Curious what you guys think ?</p>

<p>Student X:
GPA: 4.0, SAT-M 700+, SAT-V 700+, Class Rank - Top 10%
According to that press release the respective admit rates were as follows:
17%, 17%, 21%, 16%</p>

<p>What will be the cummulative admit rate? Based on my probability analysis, the cummalative admit rate of student X will be:</p>

<p>1 - {(83/100)<em>(83/100)</em>(79/100)*(84/100)}
1- .4572
54.28%
There was 54.28% chance of student X getting admission in Stanford that year.</p>

<p>Student Y:
GPA: 3.8, SAT-M 700+, SAT-V 690+, Class Rank - Top 20%
According to that press release the respective admit rates were as follows:
10%, 17%, 9%, 15%</p>

<p>1- {(90/100)<em>(83/100)</em>(91/100)*(85/100)}
1-.5778
42.22%
There was 42.22 % chance of student Y getting admission in Stanford that year.</p>

<p>Student Z:
GPA: 3.5, SAT-M 650, SAT-V 650, Class Rank - Top 20%
3%, 10%, 9%, 15%
1- {(97/100)<em>(90/100)</em>(91/100)*(85/100)}
32.47%
There was 32.47% chance of student Z getting admission in Stanford that year.</p>

<p>Math doesn’t lie. Any thoughts?</p>

2

<p>It’s interesting.</p>

<p>But there’s absolutely no doubt that it’s a drastic oversimplification.</p>

3

<p>I would argue the cummulative admit rate presented here is more comprehensive than the individual admit rate listed in that press release by Stanford.</p>

4

<p>I’m not arguing about statistical validity. I haven’t even looked at the math and numbers, really. I argue that it’s an oversimplification.</p>

<p>There’s far too many other variables, both within (ECs) and without (AA) the applicant’s control, that come into play.</p>

<p>Again, it’s interesting, but it’s nothing to base predictions or decisions around.</p>

5

<p>I agree one should not entirely base their chances on this statistics. The other subjective variables include ECs, essay, APs etc. I keeping wondering why Stanford provides admit rate stats for individual things like SAT-M, SAT-V, GPA, School Rank etc. if they feel they look at every applicant so individually.</p>

<p>Stanford would never admit this but I wonder if there is a formula that can capture their selection process. Wouldn’t that be great, some of us don’t have to waste our time trying. Hey, that is just me…</p>

6

<p>Although it would probably be correct to say “students with GPA: 4.0, SAT-M 700+, SAT-V 700+, Class Rank - Top 10%” have a higher chance than most people ignoring ECs and essays, I agree the analysis may be oversimplified.
Math doesn’t lie, but statistics done wrong does… I really forgot probability :wink: you should ask a Stanford professor!
yeah, I sorta want to know what life as an admissions officer is like…ooooiiiioooo.</p>

7

<p>Wouldn’t what you term as a “cumulative admit rate” be actually an average? Because according to your calculations, if I got a 700+ on my verbal SATs, I’d have a 21% chance of getting in, even if I had a 2.5 GPA. Just my two cents.</p>

8

<p>Yah… thats extremely poor statistics right there. There is no foundation for that…</p>

<p>A better way would be to take an average of sorts…</p>

<p>Speaking of poor statistics, anyone else think that AP stats test rough?</p>

9

<p>Your math does lie. The issue is that there is an association between the variables. For example, the 4.0 student that is accepted at 17% ON AVERAGE has a 700M, 700V, and top 10% class rank. Consequently, the odds of this person being accepted are likely below 17%.</p>

<p>If the factors really were independent and a high GPA had nothing in common with a high SAT, then you would have something. You might be able to make an argument among variables that are less correlated. For example, if you were to say that student X had a 40% chance of acceptance based on stats and a 30% chance of being acepted based on EC’s, it may be conceivable that these attributes are fairly independent and their chance of acceptance is (1 - .6*.7) = 58%. Even this is exaggerated, as there is SOME association b/w academic and EC quality.</p>

<p>The best measure is to take the average applicant, adjust for hooked applicants, and give this applicant an AVERAGE chance (i.e. a percent chance equal to the acceptance rate, adjusted for hooked applicants). If we say, arbitrarily, that:
*the average applicant has a 3.8uw GPA & a 2050 SAT
*the average unhooked applicant has a 7% chance of acceptance</p>

<p>With this, it is fairly reasonable that an applicant with a 3.8uw GPA & a 2050 SAT has a 7% chance of being accepted.</p>

10

<p>I took ap stats last year and it was easier than the barrons book i studied from! calc may have pwned me today though…it’s sad, i know.</p>

11

<p>I didn’t take Stats this year despite being in the class - I got into Stanford EA, and Stanford doesn’t offer AP Stats credit - but several of my friends took the AP test. All of them said it was unexpectedly hard.</p>

12

<p>really, all the kids at my school (including my little freshman brother) thought the AP test was remarkably simple.</p>

13

<p>math doesn’t lie, but it can easily mislead</p>

14

<p>Who cares if someone has 40% chance of getting in. At the end he’s either in or out.</p>

15

<p>Math doesn’t lie–but it does not always tell the whole truth. I agree, with a previous post, your analysis oversimplifies–and to a fault.</p>

<p>1) You’re assuming that the overall chances are averages of admit rates between different factors–but these factors aren’t weighted equally (such as GPA vs SAT) For example, if 100% with 2400s got in, and 50% with 3.8s got in, and a person has a 2400/3.8, that does not make their chances 75% because there is a difference between the weight of the two factors.</p>

<p>2) The mere numbers can’t count for anything subjective, such as a URM or a first gen. student.</p>

<p>3) The majority of the kids admitted on the low end of the applicant pool had something very outstanding about them. 100 kids, 12 admitted. All 100 have GPAs between 3.6 and 3.79, however, the 12 were admitted because they were hooked with high SATs or something like that. The others had no real reason to be admitted, because nothing stood out except their lower GPAs.</p>

<p>4) The overall admittance rate of a particular group doesn’t speak for individual chances. For example, 100 kids. 50 get in. The 100 all had 2250+ SATs. 50 get in because their GPAs were really high, the other 50 rejected because they had abysmal GPAs (this is a simplification, but bear with me). Therefore, an applicant with 2250 does not necessarily have a 50% chance, because it depends on those other things that might have helped some kids get rejected, and some accepted.</p>

16

<p>Why does it say Gpa 4.0 or higher? Does that include weighted grades? I have above a 4.0W but a much lower one uw.</p>