Data Vintage

PlainCollege currently uses institution-level Scorecard data retrieved in July 2026 and field-of-study data retrieved in March 2026. The sitewide “Mar–Jul 2026 retrievals” label records those snapshots; it is not the measurement year of every field. Institution-level entry-cohort earnings are measured 6–10 years after enrollment. The program-level earnings metric displayed here is measured four years after credential completion. Underlying reporting years and inflation reference years remain those documented by the Department in its data dictionary.

Deliberate dual-cohort joins

Several PlainCollege derivations intentionally place metrics from different Scorecard populations side by side. That is a disclosed descriptive join, not a claim that the inputs share one fiscal year or one student cohort.

• Ranking “value score” (earnings − net price) on /rankings/best-value and related state lists subtracts one year of average net price for full-time, first-time Title IV recipients from institution-level median earnings for a federally aided entry cohort measured ten years after enrollment. The two axes are different people in different years; the gap is a framing lens, not a same-cohort ROI.
• Homepage / ranking earnings-vs-cost quadrants plot the same two fields as independent axes. Labels that say “best value” mean high plotted earnings with low plotted net price under that join, not a causal school effect.
• Profile College Value Score (six weighted percentiles) also mixes earnings, net price, completion, debt, repayment, and retention dimensions that Scorecard measures on different cohorts and horizons. Missing dimensions are marked N/A and reweighted; we never invent a shared measurement year across them.

Rollback trigger: if a published page claims these inputs share one cohort, one academic year, or a causal return without naming the join, treat that copy as a data-truth defect and correct it.

Processing Pipeline

1. We download the full College Scorecard dataset from the Department of Education's open data portal.
2. Records with a valid federal school ID are normalized and loaded from the institutions returned by the Scorecard API.
3. Source values are mapped into a relational schema; rates are rendered as percentages and money values are formatted for display without replacing the underlying values.
4. Program-level earnings (field-of-study data) are matched to the correct institution using each school's federal ID.
5. When the government marks a value unavailable or privacy suppressed, we leave it blank rather than estimate it. Scorecard reporting standards vary by measure, and earnings receive additional privacy treatment; the public documentation does not support one universal threshold for every field.
6. The processed data powers every institution profile, state page, program ranking, and comparison tool on PlainCollege, and is refreshed whenever the Department of Education publishes a new release.

Coverage

• Institutions: 6,000+ colleges and universities returned by the Scorecard API
• Programs: 2,200+ fields of study with earnings data at the school level
• Geographies: All 50 states, DC, and U.S. territories
• Institution types: Public, private nonprofit, private for-profit

Earnings Data Limitations

• Earnings figures represent median earnings of former students who received federal financial aid. Students who did not receive aid are excluded, which may affect results at selective schools with fewer aid recipients.
• Unavailable and privacy-suppressed earnings remain blank; the Department also applies perturbative privacy treatment to published earnings counts and medians.
• Institution-level entry-cohort earnings are measured at 6 and 10 years after enrollment, not after graduation, so completers and noncompleters can both be represented. Program-level earnings use completer cohorts measured after credential completion.
• Earnings vary by geography, occupation, and individual circumstances, school-level medians are group statistics, not guarantees.

How the Source Agency Collects Data

The College Scorecard draws from multiple federal data systems. Institutional characteristics and cost data come from IPEDS, which collects annual mandatory surveys from all Title IV institutions. Earnings data is derived from tax records maintained by the IRS and Treasury Department, linked to students who received federal financial aid. Completion rates are computed from cohort tracking in the National Student Loan Data System (NSLDS). All data linkages are performed by the Department of Education using de-identified records to protect student privacy.

The Department of Education suppresses cells that do not meet reporting and privacy standards. Earnings fields receive additional privacy treatment, including noise and fitness-for-use suppression. PlainCollege preserves published nulls and suppression rather than inferring missing values.

College Value Score Methodology

The College Value Score shown on every school profile is a composite of six metrics, each converted to a 0-100 percentile score against every operating, reporting US college in our database, then combined into a single weighted score and letter grade. It is a cost/earnings/completion composite, not a measure of teaching quality, campus life, or fit for any individual student.

Benchmarks were recomputed 2026-08-01 directly from every operating school in our database with a non-null, positive value for that metric (sample sizes range from 2,038 to 5,103 schools depending on the metric), then linearly interpolated between percentile breakpoints to score any individual value 0-100. The completion dimension coalesces the Scorecard's separate 150%-of-normal-time fields for four-year and less-than-four-year institutions; it does not claim every cohort has a six-year window. When a school is missing one or more metrics, that dimension is marked N/A and its weight is redistributed proportionally across reported metrics. A school missing all six metrics shows no score rather than a fabricated zero.

Corrections Policy

PlainCollege preserves published source values when displaying earnings, completion, cost, admissions, and enrollment fields. We do not impute suppressed values. Rankings, percentiles, historical slopes, value scores, and ROI outputs are separate PlainCollege calculations with their formulas and limitations disclosed on the relevant page.

A correction report should identify the page, affected field or sentence, proposed correction, and an official source when one is available. We aim to acknowledge a specific data report within three business days. We then compare the rendered value with the current database, transformation code, source snapshot, and source documentation. If our processing or copy is wrong, we correct it and record the material public change. If the portal accurately reflects an upstream value, we explain that result rather than silently replacing the source.

The public data changelog records confirmed data refreshes, corrections, and material methodology revisions from the date the log began. It is not backfilled with unverifiable entries. Typographical changes that do not alter meaning are not logged. Use the page-context correction link to report an issue with this methodology.

Not Affiliated

PlainCollege is not affiliated with the U.S. Department of Education, the College Scorecard program, or any educational institution. We are an independent data portal presenting public government data in an accessible format.