Methodology
TL;DR: Authoritative data transparency
EDsmart Data publishes verified numbers from authoritative federal and public datasets. This page explains how we source data, what we normalize, and how to interpret our charts, tables, and rankings. Each data page also carries metric-specific notes.
Primary data sources
College Scorecard
U.S. Department of Education metrics on student costs, completion, debt, and post-enrollment earnings where published.
IPEDS (NCES)
Integrated Postsecondary Education Data System — enrollment, completions, institutional characteristics, and finance where cited.
Bureau of Labor Statistics
Occupational employment and wages; projections where noted on career and workforce pages.
O*NET
Occupation attributes, skills, and related mappings used to connect majors and career profiles.
Other federal or state sources appear only when explicitly named on a given page. See also our Sources directory.
Processing & definitions
Each data page lists the metrics used, filters (for example public-only institutions or a specific award level), and the reporting year or snapshot date. Medians, means, and percentages are computed after those filters are applied unless stated otherwise.
When we show state or national aggregates, we describe whether the value is a median across institutions, a sum, or a population-weighted figure. Sector splits (for example two-year vs four-year public) follow Scorecard/IPEDS field definitions referenced on the page.
Fair comparisons need explicit filters
Raw federal files mix institution types, award levels, and reporting windows. We document the cohort on every chart so a public flagship is not silently compared to a certificate-predominant campus.
Implication: Prefer the page’s stated metric and filter set over a single headline number when planning or citing.
| Variable | Typical mapping | How we treat it |
|---|---|---|
| Net price | Scorecard / IPEDS aid files | Shown by income band when published; otherwise overall net price with a note. |
| Graduation / completion | Scorecard completion rates | 100% or 150% of normal time as labeled on the page—never mixed without saying so. |
| Median earnings | Scorecard earnings cohorts | Years-after-entry or post-completion window stated on the chart; suppressed cells left blank. |
Limitations
Important data constraints
Federal surveys have nonresponse, imputation, and suppression rules (especially for small groups). We do not invent values for suppressed cells. Earnings cohorts and coverage vary by metric—see each page’s notes.
- Privacy suppressions: Small cells (often under agency thresholds) omit earnings or related fields.
- Reporting lag: Comprehensive federal releases typically lag 12–24 months behind the academic year they describe.
- Self-reported institutional fields: Some IPEDS items rely on campus reporting; we cite the file vintage on the page.
Updates & revisions
Need clarification?
When source agencies release new files, we may refresh extracts and republish charts. Material methodology changes that affect site-wide interpretation are summarized here and on the affected pages.