Judging a Department, Not a Logo: How to Evaluate a Graduate Program
A university's overall ranking or R1 status describes the whole campus, not the specific department that will fund, advise, and place you. Evaluating a graduate program means checking twelve things at department level: faculty availability, a written funding guarantee, the stipend-to-rent ratio, placement outcomes, and your advisor's own record, before you accept anything.
How this page works. Score each item inline as you read; the running total lives in the chip in the corner. Use the tabs above to hold up to three departments at once and switch between them before you answer each item, and use the PhD/Master's toggle to match what you're applying for. Nothing you enter is saved or sent anywhere: it lives only in this browser tab until you close or reload it.
Why doesn't my university's overall ranking tell me anything about my department?
A university's research designation is computed once, for the entire campus, across every field it teaches. As of the February 2025 Research Activity Designations from the Carnegie Foundation and the American Council on Education, Research 1 (R1) status requires at least $50 million in total annual research spending and at least 70 research doctorates awarded annually, both counted institution-wide, not department by department. 187 institutions held R1 status under this methodology in 2025. A hospital-anchored medical center can carry an entire campus past that $50 million line while a three-person humanities department two buildings away has no lab, no grant pipeline, and no cohort of funded students to speak of. The label says nothing about that department, because it was never measuring that department.
The same logic runs the other way. A mid-sized state university with an unremarkable overall reputation can house one of the strongest programs in your specific subfield, fully funded and well-placed, precisely because the rest of the campus isn't competing for the same research dollars. See what R1 and R2 status actually mean for applicants and how proprietary rankings are built in how grad school rankings are constructed before you let either one order your applications. Neither tells you what follows: whether the people, the money, and the outcomes in your specific department hold up. The twelve checks below do, and you run the same twelve whether the shell says R1 or nothing at all.
Who would actually advise me here?
A department's faculty page is marketing copy. It lists everyone who has ever supervised a student in your area, including people who retired their lab five years ago, took an administrative post, or haven't accepted a new student since a grant lapsed. Count only the faculty currently active in your specific subfield, then confirm, individually, that each one is taking students for your entry year. One perfect-sounding advisor is not a plan; if they leave, go on leave, lose funding, or simply say no, you have no department left to work with.
Find this artefact
Filter the department's own faculty list by research area, cross-check each name against their lab website or recent publications for signs of activity in the last two years, then send a short, specific email asking whether they are accepting new students for your entry year. A template for that first message is in the cold email to a professor guide; keep it short and specific to their work, not a form letter.
What does the funding letter actually promise?
A PhD funding guarantee is a specific written commitment, not a general impression. It names a number of years, a dollar stipend, and exactly what tuition waiver, mandatory fee coverage, summer support, and health insurance premium coverage means in practice, since a "tuition waiver" that leaves you owing $2,000 a year in mandatory fees is not full funding. "Students are typically funded" describes what has usually happened so far, not what happens to you if a grant lapses or your advisor's lab loses a line of funding in year three.
Find this artefact
Read the offer letter itself, line by line, plus the graduate school's own funding policy page, not the department's admissions marketing page. Look specifically for the number of guaranteed years, the funding mechanism for each year (fellowship, teaching assistantship, research assistantship), the stated stipend amount, and a sentence about fees and health insurance. Run the numbers through the funding offer comparison calculator to see the real Year 1 cash value once fees are subtracted.
Does the stipend actually cover rent where this program is?
A stipend is not a number in isolation; it is a number divided by twelve, set against the median one-bedroom rent in that specific city. A $34,800 stipend in a college town with $1,450 median rent leaves room to live; the same $34,800 in a coastal city with $2,900 median rent does not, no matter how prestigious the department. As a rule of thumb, treat a ratio of monthly stipend to median rent below 1.4 as a hardship regardless of the department's reputation, since rent alone would eat most of a month's income before food, insurance, or a flight home. Do not estimate a national median stipend for your field from memory; NSF's Survey of Graduate Students and Postdoctorates in Science and Engineering publishes stipend and support data by field each cycle, so look up your own field's own current figures there rather than a number a search engine hands you.
Find this artefact
Get the rent figure from a current student, the university's own cost-of-living page, or a local rental listing site, not a national average, since rent varies enormously by city. Enter both numbers below; the calculator does the division for you.
Where did the program's recent students actually end up?
A real placement list names students, graduation years, and first destinations, covering several recent cohorts, not a paragraph claiming "our graduates go on to excellent careers." For a national benchmark on how long a doctorate should take, NSF's Survey of Earned Doctorates is a near-census of U.S. research doctorate recipients each year and tracks time to degree from both bachelor's completion and graduate school entry; a department's own reported median time to degree is worth comparing against that. A published completion or attrition rate tells you what happened to the students who didn't finish, which a placement list of graduates alone will never show you. If a program will not produce any of these three things on request, that refusal is itself the data point.
Find this artefact
Look for a "placements," "alumni," or "outcomes" page, often a PDF, sorted by year. If it doesn't exist publicly, email the graduate coordinator directly and ask for the last five years by name, year, and destination, plus the median time to degree and the completion rate for the same cohorts.
Has this specific advisor actually graduated people recently?
A department's overall placement record can hide a single advisor with no completed students in years, an overloaded roster of students stuck past year six, or a habit of pushing people out before they finish. You need the advisor's own recent graduation count, and you should get it without simply asking them, since almost everyone describes their own record favorably. See how to choose a thesis advisor for the fuller set of fit questions beyond graduation counts.
Find this artefact
Search the university's own thesis and dissertation repository (most have one, searchable by advisor or department) for completed dissertations under this advisor's name in the last five years. Cross-check with Google Scholar for recent papers co-authored with a student who is now listed elsewhere as faculty or a postdoc, and check the lab's own current roster for students well past a normal time to degree. A separate but related check: whether the department has a written policy for changing advisors at all, since its absence is itself informative about how a bad match gets handled; see how to choose a thesis committee for how committees factor into that safety net.
Is there real money behind research in this specific field here?
A university's total research spending is dominated by whichever field spends the most, which at most large universities is medical or biomedical research. That total tells you nothing about funding in your field specifically. NSF's Higher Education Research and Development (HERD) Survey, whose FY 2024 cycle was released in December 2025, collects R&D expenditure by field of research and source of funds for every surveyed institution, and that field-level breakdown, not the campus total, is the correct unit for judging a department's funding environment. A mid-sized computer science or engineering department in a field with heavy federal and industry investment can outspend a "more famous" university's version of the same department, if that university's research money is concentrated somewhere else on campus.
Find this artefact
Look up the institution's HERD data by field (available through NCSES's public data tables) or ask the department directly what its annual research expenditure or active grant total is in your specific field. A department that can't or won't answer this is telling you something too.
What do current students actually say when you ask them directly?
Faculty and admissions staff describe the program they wish they had. Current students describe the one they're actually living in. Contact them directly through the department's own student directory, not only through the advisor who might forward your email selectively, and ask questions specific enough that a vague, upbeat non-answer becomes obvious for what it is.
Ask these four questions
How many years did your most recent graduating cohort actually take? Has anyone left the program, and do you know why? If your funding ran out today, what would actually happen? Knowing what you know now, would you choose this advisor and this department again?
Silence is data. One non-reply might mean a busy week; three non-replies from three different students, or replies that dodge every specific question, usually means either genuine dissatisfaction or a culture where students have learned not to say anything candid to a stranger. See grad school interview questions for how to frame these conversations if you get one during a visit or interview weekend.
How do the twelve answers turn into a score?
Each item carries a weight of 1, 2, or 3 based on how directly it predicts whether you'll finish funded, on a reasonable timeline, with a real placement outcome. A Yes counts for its full weight, a Partly for half, and a No or an Unknown both count for zero, on purpose: the two are stored separately below because they mean different things. A No is a confirmed gap. An Unknown is a question you haven't asked yet, which is why the summary lists your Unknown items as "questions to ask" rather than lumping them in with confirmed problems.
In PhD mode, the twelve weights sum to 24 and the score is round(100 × weighted sum ÷ 24). In Master's mode, three doctoral-specific items drop out (the five-year funding guarantee, published time to degree, and the advisor's PhD graduation record, worth 3 + 2 + 2 = 7 points) and the denominator becomes 24 − 7 = 17, shown next to the score so a master's score is never silently compared on the PhD scale. A score of 75 to 100 reads as strong, proceed. A score of 50 to 74 reads as workable, but get clear answers on every item that isn't a full Yes before you commit. A score below 50 reads as investigate before accepting, because that many unknowns at once is rarely an accident.
Live scores for every department you've added. Answer items above, or switch the active tab, to update this table.
| Department | Mode | Score | Band |
|---|
Questions to ask · active department
Confirmed gaps · active department
How do I bring this into a campus visit or interview?
Print this page's checklist as a one-page form. Tick the boxes by hand during a visit, phone call, or email exchange, add up the weighted score afterward, and compare it against the bands above.
- 1. Three or more faculty in your subfield, all confirmed taking students (w3)YPN/U
- 2. Funding guaranteed in writing, 5+ years (w3, PhD only)YPN/U
- 3. Program publishes a placement list, recent cohorts (w3)YPN/U
- 4. Median time to degree published, 6 years or fewer (w2, PhD only)YPN/U
- 5. Completion or attrition rate published (w2)YPN/U
- 6. Stipend / 12 ≥ 1.8× local median 1BR rent (w2)YPN/U
- 7. Health insurance premium fully covered (w2)YPN/U
- 8. Advisor graduated 2+ students in last 5 years (w2, PhD only)YPN/U
- 9. At least one current student replied candidly (w2)YPN/U
- 10. Research expenditure in your field disclosed or findable (w1)YPN/U
- 11. Teaching load stated in offer, not "as assigned" (w1)YPN/U
- 12. Written policy exists for changing advisors (w1)YPN/U
Y = Yes (full weight), P = Partly (half weight), N/U = No or Unknown (zero weight, but write which one it was; an Unknown is a question you can still go ask).
What does a high score not tell me?
A department can pass every one of these twelve checks and still be the wrong place for you. None of this measures whether the advising style fits how you work, whether the location supports your health or your family situation, or whether you can live in that city, that climate, or that degree of isolation for five or more years. A funding guarantee, a strong placement record, and a responsive advisor describe a department that can support a student. They don't guarantee that student is you.
Weigh a strong score against fit the way you'd weigh a strong resume against an interview: necessary, not sufficient. A good department you can actually survive for the length of the program beats a great one you cannot, every time, regardless of what either one scores here.
Frequently asked questions
How do I evaluate a graduate program if the university's overall ranking is strong?
Check the department, not the campus. A university-wide ranking or a Carnegie research designation is computed across every field the institution teaches; as of the February 2025 Research Activity Designations, Research 1 status only requires $50 million in total annual research spending and 70 research doctorates counted institution-wide, so one well-funded school or medical center can carry the whole label. Run the twelve checks in this guide against the specific department and the specific advisor, and treat the university-level ranking as background information, not evidence.
What are the biggest PhD program red flags?
A funding letter that says students are "typically funded" instead of stating a guaranteed number of years in writing, a faculty page listing people who haven't taken a student in years, no public placement list for recent graduates, an advisor with no completed dissertations in the last five years, and current students who won't reply to a polite email or who reply only in vague positives. Any one of these is worth a direct question before you accept; two or more together are worth walking away from.
What questions should I ask a PhD program before accepting an offer?
Ask for the number of guaranteed funding years in writing, whether the health insurance premium is fully covered, the median time to degree and completion rate for recent cohorts, a placement list naming where recent graduates went, and how many students your prospective advisor has graduated in the last five years. Then ask a current student the same questions yourself; a program's own answers and its students' answers should match.
What does a real PhD funding guarantee look like?
A real guarantee states a specific number of years (five or more for most PhDs), a dollar stipend amount, and exactly what is waived: tuition, mandatory fees, and the health insurance premium. It names the funding mechanism for each year, such as a fellowship in year one and teaching or research assistantships after. "Students are typically funded" or "funding is available" is not a guarantee; it is a description of what has usually happened so far.
Can a lower-ranked university have a better graduate program for me?
Yes. Department strength and university ranking measure different things: a ranking or Carnegie designation is a campus-wide number, while your funding, your advisor's availability, and your field's placement record are set at the department level. A mid-sized university with unremarkable name recognition can outspend and outplace a famous university's version of your specific department, especially in a field where the famous university's research money sits somewhere else on campus.
Does this scoring tool work for a master's program too?
Yes. Switch the applicant type toggle to master's mode and the three doctoral-only items, the five-year funding guarantee, published median time to degree, and the advisor's PhD graduation record, drop out of the scoring. The tool then scores the remaining nine items out of 17 points instead of 24, and it will not compare a master's score directly against a PhD score on the same scale.
Sources
- Carnegie Foundation and American Council on Education, 2025 Research Activity Designations Fact Sheet (released 13 February 2025). Confirms the R1 threshold of $50 million total annual research spending and 70 research doctorates, counted institution-wide across all fields, and the count of 187 R1 institutions in 2025.
- NSF NCSES, Higher Education Research and Development (HERD) Survey, FY 2024 cycle (results released 23 December 2025). Confirms that HERD collects R&D expenditures broken out by field of research, the correct unit for judging a department's funding environment rather than a campus total.
- NSF NCSES, Survey of Earned Doctorates (SED), 2024 cycle. A census of U.S. research doctorate recipients that reports time to degree from bachelor's completion and from graduate school entry, the national benchmark referenced in Step 4.
- NSF NCSES, Survey of Graduate Students and Postdoctorates in Science and Engineering (GSS). Publishes financial support and stipend data by field and institution; use it to check a current, field-specific stipend figure rather than relying on a remembered national average.