Data sources & methodology

Historical file population and revisions

Historical NCES files are matched by exact UNITID to the connected institution directory. Rows outside that population are preserved in the complete original-language source snapshot, with documented institution-ID exclusions and directory evidence, rather than being assigned invented school matches. Connected field totals are not complete national totals. The source file year and cohort labels remain explicit. When an official ZIP includes more than one CSV vintage, the configured member name and archive/member checksums are recorded; the collector does not guess a version or combine vintages.

Institution directory

Connected NCES IPEDS directory records provide institution names, locations, types, levels and official website links. We connect active institutions open to the public and preserve historical records that no longer appear in that filtered collection. The directory includes universities, colleges and other postsecondary institutions, including US territories and other IPEDS reporting jurisdictions. Records in Micronesia, the Marshall Islands and Palau use their own country codes (FM, MH and PW); they are not classified as US institutions.

Published undergraduate charges

Connected IC and COST1 files provide full-time, first-time undergraduate academic-year prices. Approved reporting periods and official files are listed below. CHG2AY0–3 are published in-state tuition and fees; CHG3AY0–3 are out-of-state tuition and fees; CHG4AY0–3 are books and supplies. Prices are nominal US dollars. Housing, food, insurance and travel are not included. Out-of-state is not a verified international-student fee category.

Public institution average net price

The connected COST2 NPIST and GIST fields cover full-time, first-time degree/certificate-seeking undergraduates at public institutions with standard academic calendars who paid in-state or in-district tuition and received federal, state/local or institutional grant or scholarship aid. Other grant sources are excluded. COST2_2024 indices 0, 1 and 2 correspond to 2021–22, 2022–23 and 2023–24. Average net price includes a population cost of attendance less qualifying grant aid, rather than tuition alone. Negative reported values are retained and do not promise money to an applicant. These cohort averages are not international prices or personal aid offers. Private/largest-program and income-band fields are outside this adapter.

Japan provider-listed programmes

The JASSO Study in Japan filtered English CSV is stored separately from IPEDS completion facts. It includes multiple degree levels and two instruction labels: full English and English with supplementary Japanese. Exact duplicate rows are consolidated, and URL variants of a source record are retained together. CollegeVerse derives stable record keys; they are not official national institution codes. The export does not supply a reporting year, per-row update date or institution location. The retrieval date is not a programme revision date. Missing listings can become inactive while keeping their old evidence. Confirm current intake, curriculum and admissions with the institution.

England provider register

Selected fields from the Office for Students workbook Register sheet retain legal and trading names, UKPRN, registration category, degree-awarding power labels and university-title labels. They are independent of school completion facts and current courses. This source covers English providers rather than all UK institutions; deregistration and penalty sheets are outside the connected directory. Retrieval date is not a regulatory event date. Registration fact changes remain held for specific review, and absent observations retain history without being called closed. Contains Office for Students information licensed under the Open Government Licence v3.0; CollegeVerse selects fields and normalizes whitespace.

Ontario programme enrolment

The Government of Ontario open workbook Program sheet supplies historical fall full-time headcounts by institution-defined programme and Undergraduate/Graduate grouping. Fiscal year is a source funding-year label aligned with April 1–March 31, not a verified current entry cycle. Institution reporting labels are preserved and are not national identifiers. Undergraduate is not inferred to be bachelor-only, Graduate is not inferred to be master-only, and names are not joined to US CIP codes. Asterisks remain suppressed, zero remains a reported zero, and unavailable numbers are not reconstructed. Other demographic worksheets are outside the connected dataset. Definitions from the associated older-named dictionary were checked against the selected 2024–25 fields; new fiscal-year scope requires review. Catalogue/licence changes can hold publication. Contains information licensed under the Open Government Licence – Ontario. These observations do not establish current courses, tuition, teaching language, DLI or permit eligibility.

Ontario college programme enrolment

A separate Government of Ontario College enrolment Program worksheet retains publicly funded college labels, institution-defined programme names, source fiscal year and fall full-time headcounts. It has no programme-level credential, campus, study-year or teaching-language join. Fully identical source rows are consolidated without summing; unnamed zero rows are excluded with diagnostics. Internal programme-label spacing remains intact because similarly spelled names can carry different source counts. The 2024–25 dictionary internally references 2023–24 procedures; selected scope is explicitly verified. Suppression is retained, other worksheets are outside this connection, and current courses or permit eligibility are not inferred. Contains information licensed under the Open Government Licence – Ontario.

Admissions statistics

Connected ADM datasets describe first-time degree/certificate-seeking undergraduate admissions in their labeled fall reporting years, at reporting institutions without an open-admission policy. Admission rate is admitted divided by applicants when both are present and applicants are positive. It is not a prediction for an individual or a particular program. SAT ranges describe enrolled score submitters, not minimum entry requirements.

First-year retention

Connected EF part-D percentages concern prior-fall first-time degree/certificate-seeking cohorts assessed a year later. EF2024D uses fall 2023 entry cohorts and fall 2024 assessment. NCES describes persistence or completion; four-year institutions use bachelor-seeking cohorts. Adjusted cohorts apply defined exclusions and inclusions, and grouping uses entry full-time or part-time status. The site retains source-published percentages and flags rather than recomputing them. These are not programme graduation rates or individual predictions. Total entering-class and student-to-faculty fields are outside this adapter.

Bachelor entry-cohort graduation

Connected DRVGR2024 GBA4RTT, GBA5RTT and GBA6RTT fields are NCES-derived cumulative percentages for full-time, first-time bachelor or equivalent students entering in 2018, using the adjusted entry subcohort with permitted exclusions. Four, five and six years correspond to 100%, 125% and 150% of normal time in the definitions. Source-file year and entry-cohort year remain separate. Source percentages are retained, including zero; CollegeVerse does not reconstruct raw denominators. This connection does not supply per-cell imputation flags or campus parent-allocation status, and the dictionary does not distinguish provisional from final release. These observations are independent of first-year retention, annual awards, all-student graduation and current programme availability.

Fields and completions

CIP 2020 describes fields of study. Connected annual completion files report awards conferred during July through June, labeled by the ending year. For example, year ending 2025 describes July 2024 through June 2025. We retain six-digit fields, first major and positive completion counts at supported award levels. Records with zero completions are excluded. A historical graduate record is not a current course catalog, and award counts are not always counts of unique students.

Taiwan annual institution directory

Taiwan MOE annual directory observations retain four-digit school codes, source academic years, original Traditional Chinese names, ownership/system/county labels and directory contacts. They are separate from IPEDS schools and award records. Common character variants are normalized for search only. No English institution names, courses, fees, eligibility or cross-code merger inference are supplied. Academic-year calendar labels are displayed beside original source years; semester and entry dates are not derived. Historical records remain when omitted by a later download and do not establish current presence or closure. Source website text is retained; nonstandard historical values are not automatically linked. Ministry of Education, Department of Statistics (Taiwan), 大專校院名錄, released under the Open Government Data License, version 1.0; attribution and official resource links appear on directory pages.

Japan MEXT school-code directory

Separate Japanese-language observations retain the 13-character MEXT school code, exact reference date and provisional/final release label. The connected CSV files cover universities, junior colleges and colleges of technology, including source flags for abolished records. Default browsing uses the latest retained date and the source main flag; optional filters and historical profiles are not evidence of current admission availability. Attribute setting dates are not institution founding dates, and transition pointers are displayed without inferring a merger. Codes were reassigned in 2021 according to the official catalogue. Original names, postal codes, addresses and raw date strings are retained. No English name, course, telephone or website is invented; MEXT and JASSO observations are not joined by name. Source: Ministry of Education, Culture, Sports, Science and Technology (Japan), school-code catalogue. CollegeVerse parses and formats the CSV values and does not imply official endorsement.

Budget scenarios

Single-year and multi-year calculators use reader inputs, a selected currency and explicit study-period assumptions. The multi-year tool compounds separate cost-group assumptions and rounds supported currency units per year. Entered grants apply only in the chosen years; excess amounts are not transferred between years. The tool does not supply program prices, forecast growth, convert currencies or determine funding eligibility. Scenario CSV files retain assumptions and amounts without creating a saved budget profile.

Missing values and source flags

Blank and official missing/sentinel values become unavailable, never zero. Reported zeros remain zero. Source flags are stored alongside figures, including reported and imputed values. Dataset coverage varies; missing data does not establish a school’s policy or program availability.

Collection and review

Original source language, URL, time and content hash are retained. Schema changes, invalid figures, unusual coverage drops and access failures enter exception logs. Changes retain staged evidence before publication. Validated remote official tables can publish under source-specific quality rules; unusual values and quality risks remain held for review. Collected reference text enters editorial drafts, and field dictionaries monitor definition changes. Source refreshes check configured files; they do not guarantee discovery of a newer annual release.

Approved dataset coverage

These counts describe retained records, rather than a complete current program catalog. Different datasets cover different populations and reporting years. A newer source check does not change the reporting period.

Older directory records can be retained when absent from the latest active/public-facing collection; that absence does not determine the institution's current status. Completion records include only positive first-major counts with connected institution IDs and supported award levels.