Innovative College Prep
My ICP

Data transparency · Research methodology

Know what the evidence can tell you.

Every useful statistic needs its source, population, reporting period and limitations. ICP preserves those distinctions instead of implying that every field is complete or current.

Five questions before comparing a number

What does it measure? Which students does it describe? What year is it from? Is it a published fact, a historical observation or your own estimate? What can it not tell you? Open “About this number” for a plain explanation, then “Research record” if you want its original source.

Understand the five kinds of college costs
Tuition & fees
Instruction and required college fees; living costs are additional unless explicitly included.
Attendance budget
College charges plus estimated living, books, travel and other included expenses, before aid.
Average annual cost after grants & scholarships
A historical average after federal, state/local and college grant aid for a specified federal-aid population. Outside private scholarships are not automatically deducted; loans and work-study are not discounts.
Your net cost
Your entered attendance cost minus selected grants and scholarships. Loans and work-study remain funding sources.
Borrowing
Debt to repay, not a discount. Work-study is earned over time, not an upfront bill reduction.
Full research methodology & selection rules

Two coverage levels

The original 201 institutions remain intact: 179 have partial source-reviewed CDS coverage; 22 have federal alternatives and no accepted CDS observations. Nationwide discovery adds 2,293 federal-only reporting units, for 2,494 total. Additional institutions do not inherit a CDS, current policy review or another campus’s data.

How records are selected

The reproducible offline pipeline checks pinned archive hashes, manifests and campus identities. Nationwide selection uses active HD2025 four-year, undergraduate, degree-granting institutions in U.S. states, DC and territories. A federal-only record requires at least one eligible positive indicator. institution identifiers stay strings; successor and branch campuses are not substituted.

Institutional and federal evidence

CDS observations describe institution-published populations and academic cohorts. Federal IPEDS and College Scorecard fields have their own definitions and reporting lags. These remain separate evidence layers. Federal-only does not mean every observation was individually reviewed against a college website.

Verification, conflicts and missingness

Only accepted display observations are published. Missing or suppressed values remain unavailable. Ambiguous, extracted-unverified and flagged observations are not promoted into facts. Conflicting sources remain separate; their dates and populations may explain a difference. Some CDS sources are blocked or unavailable.

Statistics are not application rules

Historical acceptance rates describe applicant pools, not individual admission chances. SAT/ACT percentiles usually describe enrolled score submitters, not universal entry requirements. Reported GPA may have unknown weighting or reporting populations. Current testing policies and deadlines require cycle-specific official evidence; historical CDS is insufficient.

Programs and career exploration

CIP 2020 classifications organize reported first-major bachelor degree completions in 2023–24 and 2024–25. A completion record does not verify that a program is currently offered. Career examples are editorial questions linked to the BLS handbook, not a formal CIP-SOC crosswalk, a placement claim or an employment guarantee.

Outcome restrictions

Program debt and one-year earnings use the source’s federal-aid reporting groups for broad fields and bachelor-level credentials, which can span campuses. They are not individual-specialty or individual-campus salary estimates. Suppression, cohorts, inflation units and missingness remain explicit. Institution four-year earnings affected by the documented October 15, 2026 errata are quarantined; flagged rows are conservatively withheld. No earnings rankings are created.

Historical trends

The supplied archive supports comparable two-year completion counts for the same institution, specialty, bachelor award and first-major definition. Charts require both reported observations. No interpolation occurs. Comparable annual admissions and cost series are unavailable; differently defined single-year observations are not joined into a trend.

Explainable preference matching

Each active preference has importance 0–3. A college receives that weight only when its known field satisfies the preference. Unknown weight is shown separately; missing data is never a match. Results sort by matched weight, then less unknown weight, name and institution identifier. No admission probability or student eligibility is inferred. Size uses all-undergraduate enrollment; cost uses historical average net price, never a predicted family award.

Financial and academic calculations

GPA uses user-selected scales, credits and weighting and remains a student estimate. Financial tools keep attendance costs, average net price, gift aid, borrowing and work-study distinct. Scenario projections use explicit user assumptions, not predicted eligibility or actual aid awards. Student Aid Index is not a bill. Formula explanations accompany each tool.

Research dates and freshness

The source archive snapshots are October 16–19, 2026. Each observation retains its own reporting period, evidence ID and source review date where recorded. Federal publication dates do not make historical costs current. Website URLs are institution-reported and are not all independently retested. There is no single “everything updated today” date.

Private planning and public sharing

Student notes, essays, academic inputs, budget estimates and saved preferences stay in the existing browser-local workspace. No account or automatic synchronization exists. Local storage is not encrypted. Generic school activity links contain only approved public settings; they never include student workspace contents. Exported student backups must be kept private.

Original sources & attribution

Federal records are attributed to their federal providers; institutional CDS and external resource content remain subject to their owners’ terms. ICP does not redistribute commercial rankings or claim a proprietary ranking dataset.

Original 201-college selection notes and detailed limitations