How the Hakia Score Works

If you're choosing where to spend four years and tens of thousands of dollars, you deserve to know how the rankings you're reading were built. Every ranking site makes choices about what to measure and how to weight it, and those choices shape which schools appear at the top. We think you should be able to see ours and decide for yourself whether they match what matters to you.

Every accredited program in our database gets a Hakia Score from the same algorithm applied to the same federal datasets, so a score means the same thing whether you're looking at computer science or nursing. We don't accept payments from schools to adjust scores, and no member of our editorial team manually overrides the algorithm.

The Hakia Score is a weighted composite on a 0-100 scale. Before anything is weighted, each program is placed in a peer group: programs in the same field at the same degree level. A computer science bachelor's is scored against other computer science bachelor's, never against a nursing master's. Within that peer group, a program's value on each metric becomes a percentile, and those percentiles combine into four weighted categories. A score of 80 means a program beats roughly 80% of comparable programs on the things that matter, so the number means the same in every field. Here are the four categories and why they carry the weight they do.

Our heaviest category, because what happens after you graduate matters more than anything else. It splits evenly between two measures. The first is graduation rate: the share of students who actually finish their degree, a direct signal of whether a school carries its students through to completion. The second is earnings: the median salary for the occupations a program feeds into, so a program is rewarded for opening doors to well-paid work.

Graduation rate comes from IPEDS. Salary comes from the Bureau of Labor Statistics, mapped to each program through the federal CIP-to-SOC crosswalk, so a cybersecurity program is measured against what security analysts actually earn, not a generic average.

What you pay weighed against what you get back. Half of this category is straight affordability, measured by in-state tuition, where a lower price scores higher. The other half is return on investment: median earnings in the field relative to that tuition. A program that costs little and leads to solid pay scores well here even if it isn't the most prestigious name on the list.

This is the category that surfaces the programs most ranking sites bury: affordable public universities and community colleges that quietly turn out strong earners at a fraction of the cost.

Source: tuition from IPEDS, earnings from BLS.

Two signals of the academic environment. The first is selectivity: we invert the admission rate, so a school that admits 10% of applicants scores higher than one that admits 70%. Not because selectivity causes quality, but because it correlates with the resources, faculty, and peer group that do. The second is incoming preparation, measured by the SAT midpoint of enrolled students.

Open-admission schools, including most community colleges, don't report an admission rate or test scores. They aren't penalized for it: when a metric is missing, its weight is redistributed across the metrics we do have, so an open-access school is judged on its outcomes, value, and scale, not on a number it has no reason to report.

Source: IPEDS Admissions survey (ADM component).

Whether a school has a real, established program in the field, not just a line in the course catalog. We measure two things: total enrollment, which reflects institutional resources, and program completions, the number of students who finish the program each year, identified by CIP code.

A department that graduates hundreds of students a year has faculty depth, staying power, and an alumni network a five-graduate program doesn't. This category keeps a school from ranking highly in a field where it barely has a program, the kind of thing that happens when a single course gets dressed up as a major.

Source: IPEDS Completions and enrollment data, filtered by CIP code.

The same 4-factor model with the same weights applies across all degree levels: associate's, bachelor's, master's, and doctoral. We rank each level separately. A bachelor's program competes against other bachelor's programs in the same state and field, never against associate's or master's programs.

The main difference by level is in the graduation rate window. Associate's programs use a 3-year completion rate (150% of the expected 2-year timeline). Bachelor's programs use a 6-year rate. Graduate programs use the institutional graduate completion rate where available.

We also apply a minimum threshold: a degree level only appears in our rankings for a state if at least 2 schools offer programs at that level. A state with one doctoral program in cybersecurity doesn't get a doctoral ranking because there's nothing meaningful to compare.

Our online rankings use the identical methodology: same four factors, same weights. The only difference is the pool of schools: we filter to institutions flagged in IPEDS as offering distance education in the relevant program field. This includes fully online programs, hybrid programs, and schools where the majority of coursework can be completed remotely.

Online programs often show lower graduation rates than their on-campus counterparts. This isn't necessarily a sign of lower quality. Online students are more likely to be working adults, part-time students, or career changers with different completion patterns. We don't adjust for this. The algorithm treats an online 45% graduation rate the same as an on-campus 45%. Prospective students can weigh that context themselves.

Every data point in our rankings traces back to one of three federal sources. We don't survey schools, we don't accept self-reported data, and we don't use third-party aggregators. Here's what we pull from each source and when.

ABET accreditation is the gold standard for computing and engineering programs. It means a program's curriculum has been reviewed by industry and academic experts and meets established quality standards. For fields like computer science, software engineering, and computer engineering, ABET accreditation matters to employers, and some won't consider candidates from non-ABET programs.

We display ABET accreditation status on school profiles where applicable, but it is not a factor in our composite score. The reason: ABET accreditation is binary (you have it or you don't), while our scoring model needs continuous variables that differentiate across a spectrum. Including it as a scoring factor would create a cliff where every ABET school gets a bonus and every non-ABET school gets penalized, which doesn't reflect the gradual differences in program quality we're trying to capture.

We believe ABET accreditation should inform your decision, which is why we surface it prominently. But it shouldn't be the only thing that matters, and baking it into the algorithm would overweight a single credential at the expense of the graduation rates, program size, and career outcomes that also signal quality.

Hakia earns revenue through affiliate partnerships with education platforms. When you click certain links or widgets on our site and enroll in a program, we may receive a commission. This is how we fund the data work, research, and infrastructure behind these rankings.

Affiliate relationships do not influence rankings. Our scoring algorithm has no input for "is this school an affiliate partner" because that input doesn't exist. Schools cannot pay to improve their rank, appear higher in results, or be featured in our top program cards. The affiliate widgets on our pages promote education platforms broadly, not specific ranked schools.

We disclose this because transparency about revenue models is part of what makes a ranking trustworthy. If you see an affiliate widget on a ranking page, know that it exists alongside the rankings, not because of them.

No ranking system is complete, and we'd rather be upfront about the gaps than pretend they don't exist.

• Teaching quality: IPEDS doesn't measure how well professors teach or how engaging the coursework is. Graduation rates are a distant proxy at best.
• Student experience: Campus culture, student organizations, research opportunities for undergrads, and the day-to-day experience of being a student aren't captured in federal data.
• Program-specific graduation rates: IPEDS reports institution-wide graduation rates, not by major. A school's CS program might retain students better or worse than the school average.
• Employer perception: Some employers strongly prefer graduates from specific programs. This reputation factor isn't in our data.
• Cost of living: A $145,770 salary in San Francisco has different purchasing power than the same salary in Austin. We report raw salary data without cost-of-living adjustment.
• Recent changes: IPEDS data lags by about two years. A program that hired five new faculty last year or launched a new AI specialization won't show those improvements until the next data cycle.

Rankings are a starting point, not a verdict. Use ours to narrow your list, then visit campuses, talk to current students, and evaluate the things data can't measure.

We refresh scores when new IPEDS and BLS datasets are published, typically once per year. The current scores use IPEDS 2024 and BLS OEWS May 2025. When a new dataset is released, we re-run the entire scoring pipeline across every program and publish updated scores within 30 days.

Between major data releases, we update editorial content (career guides, program descriptions, enhanced research on top schools) on a rolling basis. Each page displays its last-verified date so you can see how recently the content was reviewed.