Education Services · National Center for Education Statistics (NCES)
IPEDS Trend Generator & Data Center institution lookups
Datadory delivers higher-education data covering the IPEDS Trend Generator and Data Center layer: preset trend questions across twelve subjects - institutions, admissions, charges, enrollment, degrees, graduation rates, outcomes, aid, revenues, expenses and staffing - computed over 7,000+ US Title IV institutions, with per-institution profile lookups. Sample rows first; delivered daily, weekly, or hourly.
API, files, or your warehouse. Daily, weekly, or hourly.
- Where it covers
- United States - every Title IV postsecondary institution, aggregable nationally, by state or resolved to a single campus through Data Center lookups.
- How far back
- Multi-decade trends across IPEDS collection years with a per-year label on every point; graduation-rate questions are anchored to Cohort Year rather than publication year.
- How fine
- Institution-level for lookups and institution-mode trends; national and state aggregates for system-level questions; filterable along variables such as sector, level and control.
What does the IPEDS Trend Generator & Data Center include?
It is the analytical front end over IPEDS, the Department of Education's census of every Title IV postsecondary institution in the United States. NCES describes the Trend Generator as "a fast and simple way to view IPEDS data over time", and the structure earns the claim: twelve subjects - Postsecondary Institutions, Admissions, Student Charges, Student Enrollment, Residence and Migration, Degrees and Certificates Awarded, Graduation and Retention Rates, Outcome Measures, Financial Aid, Institutional Revenues, Institutional Expenses, and Employees and Instructional Staff.
Each subject carries preset questions phrased the way analysts actually ask them: "How many people are employed by postsecondary institutions?", "What is the graduation rate within 150% of normal time at 4-year postsecondary institutions?" A BUILD TABLE function reshapes any result with variables as row or column dimensions, and LIMIT RESULTS filters it down before the numbers land.
The Data Center half handles the single-institution case. Look Up an Institution returns a profile together with the institution's own reported survey responses; Statistical Tables compute totals, counts, averages, medians, standard deviations and percentiles; Summary Tables aggregate; and the Data Feedback Report sets an institution beside its peers. One census, two postures - system trends and campus dossiers.
What do rows from the trend answers look like?
An answer row is small on purpose: the question identifiers, the computed value, the year label and the count of institutions behind it. That last field is the one worth dwelling on - the verified employment trend resolves to 7,254 contributing institutions, and the count travels with every answer, so any figure you cite arrives pre-sourced.
Labels behave differently by subject, which matters more than it sounds. Employment and enrollment questions carry ordinary collection-year labels; graduation-rate questions are anchored to Cohort Year, because the rate describes a group of entering students tracked to completion rather than a snapshot. Treating those two as interchangeable is how bad higher-ed analysis gets written.
Filter variables arrive documented rather than guessed: sector resolves to "Sector of institution" with a source table and a full description attached. Sample rows above show the attributes observed this session against live trend computations; complete answer series ship with your sample.
Which fields does the field dictionary define?
The definition side: subjectId, subjectText, questionId and questionText pin exactly which of the dozens of preset trends an answer belongs to - the reproducibility contract. The value side: answer, schoolCount, year and dataYear carry the number, who produced it and when.
The third side is filtering. Dictionary variables expose variableName and variableTitle, tableName, variableDescription - the metadata that turns a flat trend into a cut by sector, level or control without reverse-engineering anything.
Because definitions were captured live rather than from a layout document, they reflect what the tool actually returns today, including quirks like the "(old categories)" annotation still attached to some staffing questions. Per-subject question inventories get pinned when your sample is cut so nothing shifts between evaluation and production.
Where does coverage run, and at what grain?
Geographically, the frame is the United States and its entire Title IV universe - every institution eligible for federal student aid, aggregable nationally, by state, or resolvable down to one campus through a Data Center lookup. If your comparison needs other countries, harmonised international tables slot in above this record rather than inside it.
Temporally, the value is depth: multi-decade trend series across IPEDS collection years, one observation per institution per question per year, with cohort anchoring on completion measures. Few education sources let you watch a sector's revenue mix change across three decades without stitching files together first.
Grain follows the question. System-level trends want aggregates; competitive benchmarking wants the institution rows underneath them. Because both come from the same census and share UNITID keys, moving between them is a join, not a re-collection.
Field dictionary
Every field below is documented against real records. The full dictionary ships with the sample.
| field | type | definition | example |
|---|---|---|---|
subjectId | integer | Numeric identifier of a Trend Generator subject area. | 5 |
subjectText | string | Display name of the subject area. | Employees and Instructional Staff |
questionId | integer | Numeric identifier of a preset trend question within a subject. | 11 |
questionText | string | Natural-language question the trend answers. | How many people are employed by postsecondary institutions? (old categories) |
answer | number | Computed trend value for the selected year and filters. | |
schoolCount | integer | Number of institutions contributing to the answer. | 7254 |
year / dataYear | integer | IPEDS collection year a row refers to, with underlying data-year labels; graduation-rate questions use Cohort Year. | 2011 |
variableName / variableTitle | string | Filter variable code and title from the question dictionary. | sector / Sector of institution |
tableName | string | Underlying IPEDS table the filter variable derives from. | hd2011 |
variableDescription | string | Full metadata definition of a filter variable as documented in IPEDS. | |
UNITID | integer | IPEDS institution identifier keying every Data Center profile lookup and joining across all IPEDS products. |
Questions buyers ask
What does the IPEDS Trend Generator & Data Center contain?
Two complementary tools over the US postsecondary census. The Trend Generator answers preset questions across twelve subjects - institutions, admissions, charges, enrollment, migration, degrees, graduation rates, outcome measures, financial aid, revenues, expenses and staffing. The Data Center adds per-institution profiles, statistical and summary tables, and peer-comparison feedback reports.
How many institutions sit behind an IPEDS trend?
Institution-mode trends run to 7,254 contributing institutions on the verified employment question, and every answer carries its schoolCount so the denominator is never a mystery. Aggregated national and state views draw on the same Title IV universe.
Can I break a trend down by institution type?
Yes. Question dictionaries expose filter variables such as sector, level and control, each arriving with a code, a title, a source table and a full description. That is what turns a single national line into a public-versus-private comparison without rebuilding the query from scratch.
Does this cover universities outside the United States?
No - the frame is every Title IV postsecondary institution in the United States, which is precisely why the figures are census-complete rather than sampled. For cross-country work, harmonised international education datasets slot in above this record and Datadory delivers those too.
How far back do IPEDS trends go?
Multiple decades. Trends run across IPEDS collection years with a per-year label on every point, and completion measures anchor to Cohort Year because a graduation rate describes an entering class tracked to the end. Longitudinal questions stop needing file archaeology.
Can I evaluate records before committing to a feed?
That is what the sample is for. Name the subjects, questions and years you care about and we cut answer series toward them, so you can validate the counts, the cohort labels and the filter behaviour against your own pipeline before anything recurring starts.
Datasets that pair with this one
- IPEDS Use the Data portal The bulk-file side of the same census - component surveys and directory data keyed on UNITID, the natural companion when a trend needs the rows underneath it.
- NCES Condition of Education (CoE) NCES's curated indicator tables - pre-interpreted context to set beside the raw trend lines this record carries.
- National Teacher & Principal Survey (NTPS) Staffing detail the trend subjects summarise - go here when the faculty-composition question needs survey depth rather than a decade-long line.
- Peterson's college search listings A commercial postsecondary listing layer - useful cross-check when your segment definitions reach beyond the Title IV universe.
- World Bank education statistics (EdStats) Harmonised indicators across countries - the above-the-frame complement once US-only trends stop being enough.
- OECD Education at a Glance OECD-member attainment, spending and earnings tables - put the American census beside its peers on the questions that cross borders.
See the rows before you pay anything.
Name this dataset and we send real records from it — scoped to the fields you asked for.