The Common Data Set is the most useful document in US admissions that Indian families never read. Every year most selective American colleges fill in the same standard questionnaire and publish it: how many applied and were admitted, what admissions says it weighs, the score ranges of the students who enrolled, and how much aid went to international students. It is the college's own data, in the same format at every college, and it answers questions no ranking or brochure will.
It is also around forty pages of tables, which is why almost nobody opens it. Five items do most of the work for an Indian applicant. Below is what each one says and how to read it, using the 2025-26 Common Data Sets of MIT, Cornell, Northwestern and Boston University, read on 30 October 2026.
What is the Common Data Set, and where to find it
The Common Data Set is a set of standard questions and definitions agreed between colleges and the publishers who survey them, the College Board, Peterson's and U.S. News among them. Because every college answers the same items in the same order, item C1 at Cornell means exactly what C1 means at Boston University, and you can compare them line by line.
To find one, search for the college's name and "common data set". It is usually on the institutional research office's website, as a PDF or a web page. The 2025-26 edition describes the class that entered in autumn 2025 and appears during 2026, so it is the most recent complete year.
It runs in ten sections, A to J: general information, enrolment and persistence, first-year admission, transfer admission, academic offerings, student life, annual expenses, financial aid, faculty and class size, and degrees conferred. For a first-year applicant, section C (admission) and section H (financial aid) hold almost everything that matters.
C1: the international student acceptance rate, if the college publishes it
C1 gives the number of first-year applicants, admits and enrolled students. Divide admits by applicants and you have the overall admit rate: MIT admitted 1,334 of 29,281, or 4.6 percent; Northwestern 3,933 of 52,676, or 7.5 percent; Cornell 6,077 of 72,523, or 8.4 percent.
The overall rate is not your rate. In the current edition, C1 also has an optional table that splits applicants, admits and enrolled students by residency: in-state, out-of-state and international. Where a college fills it in, you can see the number that matters most to an Indian applicant.
| 2025-26 | International applicants | International admits | International admit rate | Admit rate for US applicants |
|---|---|---|---|---|
| Cornell | 18,566 | 664 | 3.6% | 10.0% |
| Boston University | 15,794 | 2,521 | 16.0% | 11.9% |
The two cases point in opposite directions, and that is the lesson. Cornell, which is need-aware for internationals, admitted international applicants at about a third of the rate it admitted Americans. Boston University, which gives internationals no need-based aid, admitted them at a higher rate than Americans. The policy behind the number is in section H, and it is worth reading the two together.
MIT and Northwestern left the residency table blank, which many colleges do. Where it is missing, assume the international rate at a selective private university is well below the overall rate.
C2: the waitlist in numbers
C2 shows how many applicants were offered a place on the waiting list, how many accepted it, and how many were finally admitted from it. Cornell offered 9,720 places; 6,598 accepted and 254 were admitted, under 4 percent. MIT offered 561; 468 accepted and 10 were admitted.
Read this before deciding how much hope to put in a waitlist offer. It is usually a small chance, and a family should commit to another college by the deposit deadline. What to do while waiting is covered in deferred or waitlisted as an international student.
Common Data Set C7: what the college says it weighs
C7 lists eighteen factors and asks the college to mark each one very important, important, considered or not considered. It is the nearest thing to an admissions office stating its priorities in writing.
| Factor | MIT | Northwestern |
|---|---|---|
| Rigour of secondary school record | Important | Very important |
| Academic GPA | Important | Very important |
| Standardised test scores | Important | Considered |
| Application essay | Important | Very important |
| Recommendations | Important | Very important |
| Interview | Important | Not considered |
| Extracurricular activities | Important | Very important |
| Character and personal qualities | Very important | Very important |
| Level of applicant's interest | Not considered | Considered |
Three things to take from a table like this. First, the interview: Northwestern does not count it, so it is not worth flying for, while MIT rates it important. Second, demonstrated interest: where the level of applicant's interest is considered, campus visits, information sessions and a specific supplement register; where it is not, that time belongs on the essays. Third, character and personal qualities is marked very important at both, which is the item most Indian applications underplay, because it is carried by recommendations and essays rather than marks.
C9: middle 50 percent SAT scores, and who submitted them
C9 gives the 25th, 50th and 75th percentile scores of enrolled students, and the share who submitted each test.
| 2025-26 enrolled first-years | SAT 25th | SAT 50th | SAT 75th | Share submitting SAT |
|---|---|---|---|---|
| MIT | 1520 | 1550 | 1570 | Test required |
| Northwestern | 1520 | 1540 | 1560 | 48% |
| Cornell | 1490 | 1530 | 1550 | 1,639 students; no percentage given |
| Boston University | 1420 | 1470 | 1510 | 36% |
Two cautions. The middle 50 percent describes students who enrolled, not those admitted, and at a test-optional college it describes only those who chose to send a score, who skew high. And for an international applicant a score at the 25th percentile is not a safe score; aim for the 75th. What that means in practice for Indian students is in what SAT score you need for the Ivy League.
C21: the early decision acceptance rate
C21 reports how many applied early decision and how many were admitted through it. C22 covers early action, but asks only for dates, not numbers, so early action admit rates are not in the Common Data Set.
| 2025-26 | ED applicants | ED admits | ED admit rate | Rate for everyone else |
|---|---|---|---|---|
| Cornell | 10,057 | 1,889 | 18.8% | about 6.7% |
| Northwestern | 5,939 | 1,212 | 20.4% | about 5.8% |
The gap is real but smaller than it looks, because recruited athletes and other priority applicants sit in the early pool. And early decision is binding: if the aid offer falls short, the way out is narrow. Whether it suits an Indian family, and how to check the price before committing, is in early decision vs early action for international students.
H6: how many international students get financial aid
H6 is the item Indian families most need and least read. It shows whether international undergraduates can receive need-based aid, merit aid or neither, then how many received aid, the average award and the total. Set the number aided against the international undergraduates in item B2 and you have the share.
| 2025-26 | International undergraduates (B2) | Receiving aid (H6) | Share | Average award |
|---|---|---|---|---|
| MIT | 529 | 411 | about 78% | USD 81,991 |
| Northwestern | 1,107 | 297 | about 27% | USD 74,186 |
| Cornell | 1,523 | 318 | about 21% | USD 88,263 |
| Boston University | Not in the sections read | 231 | Merit only | USD 56,639 |
This one table explains more about aid for internationals than any brochure. MIT is need-blind, so it admits internationals without regard to need and most of them receive aid. Cornell and Northwestern are need-aware, so the admitted international class is weighted towards families who can pay, and roughly one in four or five receives aid, though those who do receive large awards. Boston University ticks non-need-based aid only: its 231 awards are merit scholarships. Read alongside C1, it is probably part of why Boston University can admit internationals at a higher rate than Americans: no international is admitted on need-based aid it would have to fund.
Which universities fall in which group is set out in need-blind universities for international students. H6 is how you check the claim against the college's own numbers.
The other sections worth five minutes
- B: enrolment and persistence. B2 gives the number of international undergraduates, and the graduation and retention items show how many students finish.
- G: annual expenses. Tuition, fees, housing and food for the year, the base for any budget.
- I: faculty and class size. The student-to-faculty ratio and how many classes have fewer than 20 students, which says more about teaching than any ranking.
- J: degrees conferred. The share of graduates in each discipline, which shows what a college is really built around. A university that graduates few students in your intended major is telling you something about the department.
If you are building the list this month
Open the Common Data Set for every college on the list and fill in five numbers for each: the international admit rate from C1 if it is there, the SAT 75th percentile from C9, the early decision rate from C21, the aid policy and share from H6, and whether C7 counts demonstrated interest and the interview. Label each college reach, target or likely from those numbers, as set out in how to build a college list, and then decide where to apply early.
Then put it beside the major. Which universities are strongest for the subject your child wants to study, and how far that differs from the overall rankings, is in the top 50 universities in the US, rebuilt for your major. If you would like help reading the numbers against a real profile, the free profile review under admissions counselling is where we start.
Where these numbers stop
The figures are from each university's 2025-26 Common Data Set, read on 30 October 2026: MIT's web edition, and the PDFs published by Cornell, Northwestern and Boston University. Admit rates are calculated from the published counts and rounded. The H6 shares compare an aid count against a single autumn enrolment figure, so treat them as approximate. Every college fills in its own Common Data Set and some leave items blank, so a missing figure means not reported, not zero. The 2026-27 editions will describe the class entering in autumn 2026 and will change every number here.