§ Methodology

Admission strategy

The headline admit rate in most college guides is a weighted average across application rounds. This page explains what we surface from the Common Data Set, what those numbers mean, and what they do not prove.

What we use

§ C.1

Total applicants and admits

C.1 gives the denominator for the overall admit rate and for residual calculations. The card uses applied and admitted counts from the latest 2024-25 or newer primary CDS row.
§ C.21

Early Decision counts

C.21 reports whether the school offers Early Decision and, when present, the number of ED applicants and ED admits. We compute ED admit rate as admitted divided by applicants. If the school reports a second binding deadline, the displayed ED rate is still the combined CDS figure.
§ C.22

Early Action flags

C.22 reports whether Early Action is offered and whether it is restrictive. It does not publish EA applicant or admit counts, so we do not display an EA admit rate. Any EA rate requires a non-CDS source, usually a school press release or independent reporting.
§ C.2

Wait-list behavior

C.2 reports whether the school has a wait list and, when present, how many applicants were offered a spot, accepted the spot, and were admitted. The conditional rate, admitted off the wait list divided by accepted wait-list spots, is the actionable number; it is still noisy year to year.
§ C.7

Admission-factor importance

We surface factors marked Important or Very Important for first-generation status, legacy relation, geography, state residency, and demonstrated interest. Demonstrated interest is shown beside yield because the two signals should be read together.
§ C.13

Application fee and waiver

C.13 supplies the application fee and whether fee waivers are available. This is an operational planning input, not an admissions-strength signal.

Why "non-early residual"

We compute the residual as total admits minus ED admits, divided by total applicants minus ED applicants. We do not call this the regular-decision admit rate because deferred ED applicants can be reviewed later, and schools with EA fold applicants into the total without publishing EA counts in CDS.

Read ED rates carefully

Published ED rates include recruited athletes, legacy applicants, and other institutional-priority applicants. We do not estimate a general-pool ED rate in v1 because CDS does not separate those applicant groups.

For context, the NBER paper Legacy and Athlete Preferences at Harvard documents the admissions weight attached to recruited-athlete and legacy preferences in one highly selective setting. That research is not used as school-level data on this site; it is methodological context for why the ED caveat is displayed inline.

Yield and wait-list interpretation

Yield is the share of admitted students who enrolled. High yield can mean a school is a top-choice destination, or it can mean the school predicts interest carefully. When C.7 marks demonstrated interest as Important or Very Important, the card makes that connection explicit.

Wait-list admits vary heavily by year. A high conditional admit rate in one CDS year should be read as a rough historical signal, not a promise about the next cycle.

Legal context

Early Decision is presented by schools as a binding commitment. We take no legal position on ED. Readers who want legal context can inspect the public docket for D'Amico et al v. Consortium on Financing Higher Education et al, filed August 8, 2025 in the District of Massachusetts.

What we do not capture

  • EA admit rates, because CDS does not publish EA applicant or admit counts.
  • ED-1 versus ED-2 count splits, because CDS reports one combined ED figure.
  • Likely letters, athletic pre-reads, or informal recruiting channels.
  • Major-level, program-level, or residency-stratified admit rates.
  • Applicant recommendations. The card surfaces public data; it does not tell anyone where to apply ED.

Sources and audit trail

Every school card links to the archived CDS source for the displayed year. The public serving table is school_browser_rows; PRD 016B adds columns to that resource, not a new API endpoint.

curl 'https://api.collegedata.fyi/rest/v1/school_browser_rows?school_id=eq.bowdoin&select=school_id,school_name,canonical_year,applied,admitted,ed_offered,ed_applicants,ed_admitted,ea_offered,ea_restrictive,wait_list_offered,wait_list_accepted,wait_list_admitted,admission_strategy_card_quality' \
  -H 'apikey: eyJhbGciOiJIUzI1NiIsInR5...' \
  -H 'Authorization: Bearer eyJhbGciOiJIUzI1NiIsInR5...'

Phase 0 measurement results are preserved in the repository's PRD findings note; the API page documents the public columns exposed by this card.