What we use
AdmitScale is built on public federal data every U.S. college reports to the U.S. Department of Education: IPEDS and the College Scorecard, supplemented by college-published admissions detail where we hold it. Between them these sources describe roughly 6,000 U.S. institutions. Coverage is uneven, and where a college reports nothing we say so rather than filling the gap.
- 01
IPEDS
Integrated Postsecondary Education Data System, published by the National Center for Education Statistics. Source of: institutional characteristics, size, applicants and admits, admit rate, published cost of attendance, grant aid distribution, retention, and six-year graduation rate.
- 02
College Scorecard
The Department of Education's institution-level file. Source of: degree levels offered, average net price, net price by family income band (the NPT4 income-band fields), and institution-wide median earnings ten years after entry. Earnings are institution-wide, not by major.
- 03
College-published admissions detail
Where a college publishes a GPA range or distribution, or an SAT/ACT middle 50%, we use it as reported and mark it as reported. Many colleges publish none of these. Those schools stay eligible and carry a lower confidence level.
Each figure in your college list is tagged to its source, or marked "not publicly reported," so you can audit it.
How we synthesize
Each source is public, but they don't talk to each other. IPEDS uses one school identifier system, the College Scorecard another schema, and college-published admissions detail arrives in whatever form the college chose. Making them describe the same institution requires:
- 01
Normalization
We map each dataset to a common institution identifier so "University of Pittsburgh," "Pitt," and IPEDS unit ID 215293 resolve to the same school across every source.
- 02
Reconciliation
When two sources disagree, we prefer the more recent and more specific reported value, and we never let a modeled value overwrite a reported one.
- 03
Eligibility
We keep active institutions that award bachelor's degrees and enroll first-year students. Optional data - admissions detail, cost, outcomes - never removes an otherwise eligible school. It shows as not reported and lowers confidence instead.
- 04
Profile matching
Once a unified institution record exists, we score it against your inputs: academics, intended major, budget, geography, and preferences. Colleges you name are assessed with the same data and combined with new matches we surface.
How fit scores are calculated
Each school in your college list receives a Weighted Fit Score from 0 to 100. It is not an admission probability. It measures how well the school matches the profile you gave us across six components, each with a fixed weight:
- 25%Program fit
Whether the college awards degrees in the intended major and at what depth, based on reported completions. Where completions are not reported, the school is not credited with strength it has not shown.
- 20%Academic fit
How the student's GPA and test scores compare to the ranges the college reports. Where a college reports no range, that college is scored on what it does report and the confidence level drops. Modeled ranges are never scored as if they were reported.
- 20%Geography fit
Match against your stated states, regions, distance from home, and how strictly you want those honored.
- 15%Affordability fit
How the estimated cost compares to your stated family budget, using published cost of attendance and College Scorecard net price for your income band, where reported.
- 10%Size fit
Match against your stated enrollment size and campus setting preferences.
- 10%Outcomes fit
Six-year graduation and institution-wide median earnings ten years after entry, where the college reports them.
How well the school matches the profile and preferences you gave us. Not an admission probability.
A separate admissions calculation based on what the college reports about its admitted class.
High, Moderate, or Limited, reflecting how much admissions data the college publishes.
A flag where the major admits separately from the institution, such as Nursing, Engineering, Architecture, or Art.
Weights are fixed across users. We don't adjust them based on what you might want to hear. A high Fit Score never implies a high chance of admission, admissions tiering is calculated separately.
Reach, Match, and Likely
Every college on the list is classified as Reach, Match, or Likely, and every classification carries the evidence behind it plus a confidence level of High, Moderate, or Limited.
- Reach: admission would be a stretch on the figures the college reports for its admitted class.
- Match: the student's profile sits inside the reported range for admitted students.
- Likely: the reported admit rate and ranges put this student well inside the admitted profile.
- Confidence reflects how much the college publishes. Thin reporting lowers confidence rather than changing the tier.
- Institutional admission is not the same as admission to a competitive major. Programs such as Nursing, Engineering, Architecture, and Art often admit separately, and the report flags that where it applies.
We don't force a fixed count into each tier. If the eligible pool doesn't support a full set of Likely schools for a profile, the report says so instead of reclassifying schools to fill a quota.
How geography works
Geography is an input you control, not a guess we make. You name the states and regions you'll consider, an acceptable distance from home, and how strict that should be.
- Strict: only colleges inside the stated states, regions, or distance appear.
- Preferred: the list stays inside your area, with at most a few genuinely exceptional out-of-area colleges, each one explained.
- Distance from home is an AdmitScale estimate based on the college's reported location, and is approximate.
- Every out-of-area recommendation is labeled as such, with the reason it earned a place.
How net-cost estimates work
"Sticker price" may differ substantially from what a family ultimately pays. Net cost is closer. We estimate it using:
- The college's published cost of attendance - tuition, fees, housing, and books - as reported to IPEDS
- The College Scorecard average net price, and net price for your family's income band where that field is reported
- Your stated family income range and budget, from your assessment
- A coverage check: where a college reports no net price for your band, the report says so and lowers the cost-confidence level
Estimates only, and calculated by AdmitScale. The only authoritative net price is the official Net Price Calculator on each college's website, which we link to in your college list. We help you narrow which calculators are worth running.
How aid-context signals work
An aid-context signal doesn't mean you'll get a scholarship. It is a planning indicator based on institution-level grant-aid patterns, the difference between published cost and reported net price, and the student's position relative to available academic ranges. It is not an estimate of individual scholarship eligibility or an expected award. We derive it from:
- The share of students receiving grant aid, and average grant aid, as reported to IPEDS
- The gap between published cost of attendance and reported net price
- Where the student's reported academics sit relative to the college's reported admitted range
Signal, not guarantee, and not a published award table. We don't display merit grids, need-met percentages, or debt figures, because we don't ingest them reliably enough to show them as reported facts.
Why this is hard to find elsewhere
- 01
Several sources, one view
IPEDS, the College Scorecard, and college-published admissions detail each tell you something different - selectivity, cost, outcomes. Reading them separately takes hours per school. We unify them.
- 02
Monthly synthesis layer
The federal data refreshes on its own schedule. Our synthesis is updated monthly, so the picture stays current without waiting for the next annual federal release.
- 03
Profile-matched, not generic
Most public data is presented as one-size-fits-all averages. We evaluate each eligible college against your profile, intended major, and financial goals. The dataset is public; the personalized lens is not.
What we don't claim
Plain English about our limits, because hiding them is what makes other tools feel like sales pitches.
- 01
Essay quality
We can't evaluate your writing. A great essay can move you across a school's mid-50%. We don't see it.
- 02
Recommendations
We can't see your teachers' or counselors' letters. They matter.
- 03
Demonstrated interest
Some schools track campus visits, email opens, and event attendance. We can't.
- 04
Demographic context
Admissions readers consider context (first-gen, underrepresented backgrounds, school context). We don't model this.
- 05
Institutional priorities
Each year, every school has shifting internal priorities (geographic balance, donor relationships, athletic recruits). They're invisible to outsiders.
- 06
Human judgment
Admissions decisions are made by people using information and institutional context that no automated planning tool can fully observe.
Estimates only. Not a guarantee of admission.
When the data is wrong
Federal data has lag. A college's true admit rate this year may differ from what was last reported, and college-published figures are self-reported. We mitigate this by:
- Refreshing monthly so we catch updates as they publish
- Preferring the more recent reported value when two sources disagree
- Marking figures a college doesn't report as not publicly reported, and lowering the confidence level
- Linking directly to each college's official source so you can verify
If you spot data that looks wrong, email counselor@admitscale.com - we review reported data discrepancies promptly.
Refresh cadence
Our synthesis layer updates monthly. Underlying federal data follows its own schedule:
- IPEDS: annually, typically published in spring
- College Scorecard: annually, usually winter
- College-published admissions detail: whenever the college posts it, typically late fall
Your college list reflects the most recent data we have on the day you build it.