Guide
How the Compass Score and Value work
College Compass Guide publishes two scores for each college, built from federal data. The Compass Score measures outcomes, the entering class, cost and teaching together. The Value score measures what a college costs after aid and what its graduates earn and owe. This guide explains both. The full tables are on the Methodology page.
The Compass Score
Eight parts make up the score. Each part has a weight, and each one is placed on a fixed scale between two marks: the first mark earns no credit, and the second earns full credit. Most parts use a straight line. Admission rate and earnings use a logarithmic scale, so each step counts for the same proportion.
| Part | Weight | No credit at | Full credit at |
|---|---|---|---|
| Graduation rate (six years) | 25 | 40% | 90% |
| First-year retention | 10 | 60% | 95% |
| Entering students’ SAT or ACT | 20 | SAT 900 | SAT 1450 |
| Admission rate | 20 | 80% | 15% |
| Median earnings ten years after entry | 5 | $30,000 | $110,000 |
| Median federal debt at graduation | 10 | $35,000 | $5,000 |
| Net price after aid | 5 | $45,000 | $5,000 |
| Students per faculty member | 5 | 25 | 4 |
- Each part’s credit is averaged by its weight, and the average becomes the index: 100 times the average, so the index runs from 0 to 100. The index appears once, in the “How this score is built” box on each college page. The grade does not use it directly.
- The average covers only the parts a college reports. A missing part is left out rather than counted as zero, but the parts that remain must carry at least 60% of the total weight, and graduation must be among them. Otherwise the college gets no Compass Score.
- Letters are set on a curve by rank among all rated colleges: A+ is the top 3%, A the top 10%, A− the top 20%, B+ the top 35%, B the top 55%, B− the top 75%, and C+ the rest. A letter shows where a college stands among all U.S. four-year colleges we can rate. No college gets a D or an F.
- Compass Recommended is the top 35% (grades A+ to B+). Honor Roll is the top 10% (grades A+ and A).
- Each college’s page and card lead with its best true fact: its rank in its state (when the state has 8 or more rated colleges), a top-quarter figure for earnings, graduation, debt or net price, or its letter.
- Specialized schools, such as music conservatories and medical schools, get a grade and their own page, but are left out of the overall rankings and the Honor Roll.
The Value score
Value uses three parts, averaged the same way and set on the same curve as the Compass grade. Net price after aid counts 40%: a net price of $10,000 or less earns full credit, and $40,000 or more earns none. Median earnings ten years after entry count 35%, on a logarithmic scale, from $30,000 (no credit) to $75,000 (full credit). Median federal debt at graduation counts 25%: $10,000 or less earns full credit, and $35,000 or more earns none. A college needs all three figures to get a Value score.
Graduation is not part of Value, because the Compass Score and Academic Excellence already count it. The scales are fixed, so one college’s Value never changes because another college’s figures did.
How to use the two scores
- A high Compass Score means strong outcomes and a strong entering class, with cost and teaching counted as well.
- A high Value score means a college costs less after aid and its graduates earn more relative to what they owe.
- Neither score measures fit. Location, size, campus life and the program you want matter too, and no score captures them.
What the scores do not measure
Reputation surveys, alumni giving and magazine opinions are not used. Campus life is not scored here. The scores rely on federal data, which lags a year or two and covers only some students, so a college that has changed recently may look different in the next data release.
Next steps
- Read the Compass Score section of the Methodology and the Value section.
- See how the scores are used in the rankings.
- Read Reading earnings and debt numbers to understand two of the inputs.
Updated Oct. 20, 2026.

