A teaching degree ranking that changes its formula at every level.
Hakia Education ranks teaching programs across 83 specialties and four degree levels. The interesting decision is that the four levels are not scored the same way, because the federal data supports different signals at each one. Grading a master's on undergraduate graduation rates would produce a confident number that means nothing.
So we published four formulas instead of one, and explained why each measure was left in or out.
"I want to become an elementary teacher in Texas" is the example query, and it is a real one the search handles. Click to enlarge.
288 of the ranking pages are master's, 134 bachelor's, 61 doctoral, and 57 associate's, plus state and facet views on the specialties that earn them.
The Challenge
Teacher preparation is unusually hard to rank honestly. The federal government does not publish per-school earnings for education graduates, so the single number most rankings lean on does not exist here. The tempting move is to use the field-level salary anyway, which is identical for every school and therefore separates nothing while looking rigorous.
The levels differ just as much. An associate's in elementary education is a transfer degree from an open-admission community college, so admit rate is meaningless. A master's is taken by working teachers, so undergraduate graduation rate describes the wrong population. One formula stretched across all four levels would be wrong three times.
And the search problem is the same one every prospective teacher has. They know what they want to do, not what the degree is called. "Become a principal" is not a program name.
Each level is scored on what its own data can actually support.
A program is only ever compared to other programs in the same specialty at the same level. The weights below are published on the site, along with the reasoning for every measure that was left out.
Associate's
Graduation rate 55% · Affordability 45%Selectivity is not used, because an associate's in elementary education is a two-year transfer degree offered almost entirely by open-admission community colleges. Ranking on admit rate would score every program identically. Earnings are excluded too, since they depend on transferring and finishing a bachelor's.
Bachelor's
Graduation rate 50% · Selectivity 30% · Affordability 20%Earnings are deliberately left out: federal per-school earnings are not published for education programs, and the field-level salary figure is identical for every school, so ranking on it would separate nothing. The main list caps at two programs per state so one public-university system cannot take it over.
Master's
Program strength 40% · Degrees awarded 30% · Local pay 30%The undergraduate stats do not describe a master's, so the formula changes entirely. Program strength leads on the institution's Carnegie research classification. Degrees awarded caps out around fifty a year, so a high-volume online degree factory earns nothing for sheer size.
The methodology page is written to be read, not to be linked and ignored. Each formula gets a paragraph explaining the choice, including why a high-volume online degree factory earns no advantage from awarding thousands of degrees.
Four decisions that keep a plausible number from becoming a wrong one.
The wage is a place, not a promise
Local pay uses the BLS metro-area figure where the school sits, or the state figure otherwise, adjusted for cost of living through the Bureau of Economic Analysis Regional Price Parities. Every card that shows it says plainly that it is a location figure, not that program's graduate earnings.
Cost of living gets subtracted
A high nominal salary in an expensive metro does not get to outrank a lower salary that goes further. Without the adjustment, the ranking would simply list expensive cities.
Missing data is neutral, not zero
Scores are percentile-based within a specialty, and a program missing a field gets the neutral median for it. It is neither rewarded nor punished for a gap in the federal record.
State approval is not a demerit
A national accreditor like CAEP earns a bonus but is not required, because several states approve educator preparation through their own boards. A strong flagship should not be marked down for its state's accreditation choice.
The caveat sits directly under the number it qualifies, in the same card, rather than in a footer nobody reads. That sentence is the difference between a useful figure and a misleading one.
Built to answer the whole question, not just rank the schools.
A ranking page opens with the size of the pool and the numbers that frame it, then works through career outlook, pay by state, what it costs, formats and paths, how the degree actually works, a state breakdown, an FAQ, and its sources. Someone landing from search should not need a second tab.
Elementary education master's: 302 programs analyzed, 84 with an online or hybrid option. The sponsored unit underneath carries an AD label and sits outside the ranked list entirely.
A few of the 83 specialties
The Technical Edge
683 static pages in front of one live search API. The application work is in the pipeline that generates them and the index they search.
Semantic search over the catalog
Queries are embedded and matched against a vector index of roughly 154,000 degree programs, scoped to the education cluster. "Become a principal" finds administration degrees without matching a single word.
Three federal datasets, joined
IPEDS for graduation, admissions, and tuition. BLS Occupational Employment and Wage Statistics for area pay. BEA Regional Price Parities to adjust that pay for local cost of living.
Field-first URLs, demand-gated
Every path leads with the specialty, with level, facet, and state nested underneath. Deep combinations get built only where search demand justifies them, so the site never ships a thin cross-product.
Every page prerendered
Astro builds the whole site to static HTML served from Cloudflare's edge. The search box is the only thing that talks to a server at read time.
Sources on the page
Each ranking ends with a data and sources section, and per-figure footnotes name the exact release, down to the BLS May 2025 wage data.
Shared with the other verticals
The search component, API client, and theme layer live in one internal package used across every Hakia subdomain, so improvements land everywhere at once.
The shortcut in education rankings is to pick one formula, apply it everywhere, and let the missing data quietly become a zero. It produces a clean-looking list and a number that falls apart the moment somebody checks it.
Hakia Education takes the longer route: four formulas for four levels, every excluded measure explained, the wage figure captioned with what it is not, and 643 ranking pages that each cite the federal release behind them.
Sibling sites: Hakia Business and Hakia, which covers the root domain and the nursing vertical.
Have data nobody has turned into pages yet?
Public datasets become authority sites when someone does the joining, the scoring, and the honest captioning. That is the part we like.
Start Your Project