Methodology: How WorthDegree Calculates College ROI
Every number on this site traces to a named public source and a formula you can re-run. This page documents both — plus what we deliberately do not do. If anything here is unclear or looks wrong, tell us; corrections ship fast.
1. Data Sources
We use only public, official data — no proprietary feeds, no self-reported lists, no paid rankings inputs. Coverage: 6,273 Title IV institutions.
| Source | What we use it for | Update cycle |
|---|---|---|
| U.S. Dept of Education — College Scorecard (built on IPEDS) | Net price, tuition (in/out-of-state), completion rate, admission rate, median earnings 10 years after enrollment — school level, all 6,273 schools | Annual |
| College Scorecard Field-of-Study data | Program-level (CIP) earnings 1 & 4 years after graduation for bachelor's recipients, national 25th/75th percentiles | Annual |
| BLS — Usual Weekly Earnings (2023) | Education-level income baselines: high school ($46,748), associate ($52,416), bachelor's ($77,636), master's, professional, doctoral | Annual |
| BLS — OEWS / Occupational Outlook Handbook | Occupation median wages, 2023-33 growth projections, typical entry education, job-outlook context | Annual |
| Federal Student Aid (studentaid.gov) | Loan rates: 6.53% undergrad direct, 8.05% grad unsubsidized (2024-25) | Annual (each July) |
| USCIS — FY2026 H-1B cap results | International-student stay probability: ~35% selection rate × employer-sponsor rate | Per lottery year |
| College Board — published price trends | Context for attendance-cost components; not used in core formulas | Annual |
Current dataset generated October 2026; baselines refreshed October 6, 2026. Earnings data is published with a 1-2 year lag — that's the official data cycle, not a site defect, and we say so wherever it matters.
2. How We Calculate ROI
The core idea: a degree's real price is what you pay plus what you stop earning while studying. Its real return is the salary surplus over the baseline you'd have without it.
| Component | Formula / Source |
|---|---|
| Direct cost | Annual net price (Scorecard) × years enrolled. Net price = average cost after grants/scholarships for federal-aid recipients. |
| Opportunity cost | Years × high-school median income ($46,748, BLS 2023) × 0.65 part-time offset — most students work part-time during school. |
| Total (all-in) cost | Direct + opportunity cost. This is "the number no brochure prints." |
| Payback years | All-in cost ÷ (starting salary − high-school baseline). If the surplus is zero or negative, we return "cannot break even" — not a negative number that hides the failure. |
| Loan payment | Standard amortization (PMT) at 6.53% undergrad / 8.05% grad federal rates, 10-year standard term. |
| Starting salary | Bachelor's routes: Scorecard school-level median earnings (10 years after enrollment), cross-checked with BLS occupation medians and field-of-study data. |
| Worst case | ~70% of the median (25th-percentile approximation); payback extended ×1.5. Where field-of-study 25th percentile exists, we quote the federal figure directly. |
The three-route model (bachelor's)
Every report compares the same target career across three doors, because the license is often identical while the bill is not:
- Route A — Four-year bachelor's: full-time, school's net price, high-school-baseline opportunity cost.
- Route B — Community college transfer: ~2 years at community net price (estimated at 35% of the four-year school's) + 2 years at the transfer school. Same degree, lower bill — but transfer guarantees cover general credits, not program seats.
- Route C — Certificate / apprenticeship (2 years): BLS occupation median wage, conservative $3K/yr net cost.
Graduate and international variants
- Graduate routes: opportunity cost is measured against the bachelor's wage you already earn (not high school); Route B models employer-sponsored part-time study (~60% employer coverage, income reduced 30%, +1 year).
- International students: tuition uses out-of-state/international proxies; the model runs a dual-exit view — stay in the U.S. (H-1B probability = selection rate × employer-sponsor rate × OPT attempts) vs. return home (U.S. salary × country-specific return-salary ratio). Country assumptions are labeled with their source tier.
- Regional assumption: state selection adjusts the school list and state cost indices; salary medians remain national unless the source explicitly breaks out geography.
3. What We Do — and Deliberately Don't Do
WorthDegree is an interactive data tool, not a counseling service. The boundary, in plain terms:
| We do | We don't |
|---|---|
| Compare routes with the same government math for every school and major | Tell you which school to pick — the decision stays yours |
| Show the full bill: direct cost, opportunity cost, loan pressure, worst case | Predict admission odds, scholarships you'll win, or your personal salary |
| Quote official medians with their lags and percentiles | Present estimates as guarantees, or hide uncomfortable numbers |
| Update from official sources once a year (or as they publish) | Sell your data, take school money, or run ads/affiliate kickbacks — the Honest-Data Pledge applies to this page too |
We are not a licensed educational counselor, financial advisor, or attorney — nothing here is financial, legal, or immigration advice.
4. Data-Integrity Rules (the anti-hype red lines)
These are hard rules, not aspirations:
- Missing data = not shown. A school or program without published net price or earnings is excluded — never filled with a guess or "similar school" average.
- Outliers are named before they're ranked. Small samples, sub-50% completion rates, and specialized-profession colleges (pharmacy, maritime, nursing) are flagged or filtered — see the footnotes on the 2026 ROI list.
- Every number is recomputable. Rankings and reports are computed live from the same public datasets — no hand-entered figures. Re-run any number in the ROI report or the cost calculator.
- Small samples say so. Where official statistics suppress small cohorts (privacy rules), we say "not published" instead of inventing a number.
- Estimates are labeled as estimates. Model assumptions (part-time offset 0.65, employer share 60%, transfer-price ratio 35%) carry their source tier and are visible in the data legend.
Verify It Yourself
The fastest audit: pick a school in the cost calculator and check the all-in number against its Scorecard net price. Or open a full ROI report — every figure links back to the source table above. If your re-run doesn't match ours, that's a bug report, and we want it: hello@worthdegree.com.