Last reviewed August 17, 2026

What IPEDS is, and what it cannot replace

NCES already has a statistical system of record. It is a stack of survey tables with dictionaries, not a PDF, and it is not this year’s school-authored Common Data Set.

NCES and UNITID

The National Center for Education Statistics runs IPEDS, the Integrated Postsecondary Education Data System. About 6,400 Title IV institutions file interrelated annual surveys. The join key this site actually uses is UNITID — Harvey Mudd is 115409, Virginia Tech is 233921.

The navigational query “IPEDS” belongs to nces.ed.gov/ipeds. This page is the counselor-facing explanation of what those tables are, how they show up on a school page here, and what they cannot stand in for.

How data gets in

Components include HD (directory), IC (characteristics), ADM (admissions), EF (fall enrollment), GR (graduation rates), SFA (student financial aid), and F (finance), among others. NCES releases provisional files after quality control, including imputations for nonrespondents, then final files with institutional revisions. Recent years still arrive as Microsoft Access databases plus dictionaries. That is the silo on the federal side: not a 47-page PDF, a stack of tables.

IPEDS described as survey components HD, IC, ADM, EF, GR, SFA, and F, released first as provisional then final Access databases. — IPEDS survey-component stack. Official documentation: nces.ed.gov/ipeds. UNITID is the join key on this site. See Virginia Tech’s federal baseline table.

Worked examples on a live school page

The school-page federal baseline table (the IPEDS coverage layer) keeps source_table, source_variable, release type, and definition-alignment visible. Typical alignments:

Example IPEDS facts and how they align to CDS
Fact	Alignment	Why it matters
Admission rate	Near CDS	ADM totals can match C1 in spirit and still differ in population, gender treatment, and vintage.
Endowment	Not CDS-equivalent	Finance (F) is additive context. CDS never had a standard endowment table.
Locale	Direct	HD locale is a federal classification, not a school-authored CDS cell.

Federal baseline table with source column showing ADM admission rate as near CDS, finance endowment as not CDS-equivalent, and HD locale as direct, with source_table.source_variable labels. — Virginia Tech school page, NCES/IPEDS baseline table. Source column is source_table.source_variable plus release status. Not a substitute for the 2025-26 CDS extract. Open the Virginia Tech hub.

What IPEDS cannot replace

In public language, the coverage layer’s non-goals are: IPEDS is not a drop-in Common Data Set. It does not give you current-year wait-list behavior, Early Decision versus other rounds, H2A merit counts, or the school-authored PDF as accountability. CDS files often publish months before the matching IPEDS release. We do not silently blend CDS and IPEDS into one unlabeled number.

For those CDS-native slices, use the year page: Virginia Tech 2025-26 and Harvey Mudd 2025-26.

Why we load it

Directory-scale coverage when no public CDS exists — the search result is still a school page, not a dead end.
Historical series that outlast a school’s current IR listing.
Finance (endowment draw, Part H) that the CDS template never had. See the endowment draw-rate recipe.

How to tell CDS from IPEDS on a school page

CDS lives on the year page and in the document ledger: a downloadable file, field IDs, an academic year in the URL. IPEDS lives in the “source-labeled federal facts” table: UNITID, source_table, source_variable, provisional or final. Scorecard is a third layer, with its own vintage note. If a row does not name its source, do not cite it.

Annotated split of a school page: Common Data Set extract with field IDs on one side, IPEDS source_table.source_variable on the other, both circled as separate source layers. — Virginia Tech school page, CDS extract versus IPEDS baseline. Source labels are the product. Harvey Mudd uses the same two-layer layout. Open Virginia Tech.