Choosing Between Two PhD Offers: A Side-by-Side

Choosing Between Two PhD Offers: A Side-by-Side

When you choose between two PhD offers, compare three lines first: who you are admitted to (an adviser, or rotations and later matching), how many funded years and on what condition, and the department's completion record. At Stanford Computer Science, 59.6% of PhD students graduate in 6 years or less.

Ranking and the stipend headline are the easiest lines to compare. In our view they are the least decisive once both offers pass your budget. A stipend that is $2,000 higher matters less than a funding promise that ends a year before the typical student finishes. This page gives you the comparison table, where to find completion data, and a constructed two-offer case that shows how the table changes the answer.

Every university and Council of Graduate Schools (CGS) page quoted below was checked on October 29, 2026. The Stanford figures come from its public dashboard, which states its own date ranges.

The three lines to compare first

Who admits you: an adviser or a department

At some US programs, including the three below, admission does not come with a named adviser. Carnegie Mellon's Computer Science Department says "Official advisor matching occurs after you are admitted to the program" (CMU CSD doctoral admissions). Stanford CS says "Each PhD student spends the first few months becoming better acquainted with the various faculty and their research areas before aligning with a faculty member to pursue their doctoral research" (Stanford CS PhD admissions FAQ). MIT EECS says that after admission "some interviews and student visits will be arranged in order to determine the level of shared interest between the student and faculty" (MIT EECS admissions FAQ).

So "admitted" can mean "admitted to work with Professor X" or "admitted to find an adviser here". Write down which one each offer is. If it is the second, count how many faculty in your area are taking students next year. Ask them directly during the visit.

Funded years, and the condition

MIT EECS funds students "throughout the duration of their graduate program, assuming satisfactory progress in their research and in their academic PhD milestones." CMU's Computer Science Department funds admitted PhD students "while they are in good academic standing in their respective programs." Those are open-ended promises with a condition. Other offers state a number of years instead. A number is easier to read but has an end date. A condition has no end date but depends on how "progress" is judged.

Completion and time to degree

This is the line a ranking-and-stipend comparison leaves out. Stanford's Institutional Research & Decision Support office publishes doctoral completion and time-to-degree data by program. For Computer Science, the dashboard shows:

• a median time to degree of 5.5 years for the 253 graduates of the CS PhD program from Summer 2020 through Spring 2025
• "59.6% of Computer Science PhD students graduate in 6 years or less", based on 208 students starting between Summer 2017 and Spring 2020

Stanford's page defines time to degree as elapsed time: it "measures elapsed time only, not enrolled time" and "does not stop and start if a student takes a leave of absence."

Our arithmetic: 100% − 59.6% = 40.4% of that group had not graduated within six years. That share includes students who finished later, are still enrolled or left. The dashboard's chart of recent entering cohorts separates completed, current and discontinued students; the six-year figure alone does not.

Our view: put the funding promise next to the completion data. If a letter funds five years and the department's median is 5.5 years, the typical student needs funding the letter does not promise. That gap is a bigger difference between two offers than a few thousand dollars of stipend. The countercase: a department with a short median may admit mostly students who already hold a master's. Stanford notes that "Some Engineering programs require a MS degree prior to matriculation in the PhD, which may contribute to lower overall median TTD." Compare like with like, and ask what happens after the funded years before treating a fixed term as a problem.

The comparison table

Fill one column per offer from the letter, the program's pages and written answers. Leave a cell blank rather than guess, then ask the program to fill it. For reading a single offer letter line by line, use the checklist in what "fully funded" means in a US PhD offer.

Location and cost of living matter, but they are your numbers, not the program's. Put your own monthly budget against the 12-month total.

A constructed comparison

Constructed example, written by us. The programs, people and figures are invented and describe no real university.

Context (invented): Priya has two funded offers in computational biology.

By ranking and stipend alone, X wins: higher ranked, $4,000 more a year, and a 12-month promise.

By the table, the picture changes. X's five-year promise ends before its own published median of 6.1 years; our arithmetic: a median student would need about 1.1 years of funding the letter does not cover. X also has two faculty in her area, one away next year, so her real choice of adviser may be one person. Y names the adviser, ties funding to standing rather than a count of years, and reports a shorter typical time.

What Priya should ask before choosing, not after:

1. X: "What happens to funding after year five for students who are making progress?" A written "the department continues support" closes most of the gap.
2. Y: "When does Dr. Menon's current grant end, and how is summer paid if it is not renewed?" If the grant ends in year two with no fallback, Y's summer line is weaker than it looks.

If X answers question 1 well and Y's grant ends soon, X is the better offer. If not, Y is. Either way, the decision comes from rows that the ranking and the stipend headline do not show.

Where to find completion data

1. The university's institutional research office. Stanford publishes program-level dashboards. Search the university's site for "time to degree" or "doctoral completion".
2. The department's own pages. Some graduate handbooks or program reviews state a typical time to degree.
3. The graduate coordinator. If nothing is published, ask in writing: "What is the median time to degree for recent PhD graduates, and what share of an entering cohort completes?" A program that cannot answer has told you something too.

Do not substitute a national average or a forum estimate for a department's own figure.

A decision procedure

Run these in order. Stop when one offer clearly wins a step.

1. Budget. Does the 12-month total cover your costs after fees? An offer that fails this is out, whatever its rank.
2. Funding against time. Does the funding promise, by years or by condition, cover the department's median time to degree? Pass: yes, or a written answer about support after the funded years.
3. Adviser certainty. Do you have a named adviser, or at least two faculty in your area who say they are taking students? Pass: yes.
4. Adviser record. Do the adviser's former students finish? Pass: you can find them with degrees.
5. Everything else. Ranking, location, lab culture, partner and family. These decide between offers that pass steps 1 to 4.

Our view is that ranking belongs in step 5. The countercase: if you want a faculty job at a research university, ask both departments for their placement lists. If the lists differ sharply in the jobs you want, that can outweigh a step 2 or 3 difference.

Deadlines while you decide

For a funded fall offer at a program that follows the CGS April 15 Resolution, you do not have to accept or decline before April 15; for Fall 2027 entry that is April 15, 2027. Whether your offer is covered is answered in whether April 15 applies to your grad offer.

• Waiting on one offer while the other has a check-in date: how to ask for more time on a PhD offer.
• One row could be fixed by the program: negotiating a PhD funding offer.
• Decided: how to decline a PhD offer. Decline the other offer as soon as you know, so its funding can go to someone on the waitlist.

Questions about choosing between PhD offers

Should I choose the higher-ranked PhD program?

Only after both offers pass on budget, funding years against time to degree, and adviser certainty. Between two offers that both pass, ranking is a fair tie-breaker.

How do I compare PhD offers with different stipends?

Compare the same 12-month window and the same costs. A 10-month stipend plus uncertain summer pay is not the same as a 12-month stipend. Then compare how many years each promise lasts.

Is it better to be admitted to a specific adviser or to do rotations?

Neither is better in general. A named adviser removes uncertainty about who you will work with. Rotations give you a backup if your first choice falls through. The risk is a rotation program where only one or two faculty work in your area.

Where can I find a PhD program's completion rate?

Start with the university's institutional research office. Stanford's dashboard is one example. If nothing is published, ask the graduate coordinator for the median time to degree and the share of a cohort that completes. Still waiting on a waitlist at a third program? Send one short letter of continued interest for grad school before you decide.

A new paper after applying can still matter at the interview stage even where the file is closed.

Get a free review of your PhD statement of purpose

Get a free Quick Review of your PhD statement of purpose against GradPilot's adviser-fit rubric if you are applying again or adding programs. It reads whether your statement names a direction specific enough for a faculty reader to match to a person or group. The general PhD statement of purpose rubric reads the whole statement; the PhD statement of purpose review page has both.

More PhD offer guides: the PhD essays hub collects PhD statement, proposal and strategy guides.

Sources

Retrieved October 29, 2026. Pages state no cycle year unless noted.

• Stanford Institutional Research & Decision Support, Doctoral Completion & Time-to-degree: Computer Science median time to degree (Summer 2020 to Spring 2025 graduates), six-year completion share (students starting Summer 2017 to Spring 2020), definitions. Figures read from the page's embedded dashboard.
• Carnegie Mellon Computer Science Department, Doctoral Admissions: adviser matching after admission; funding while in good academic standing.
• Stanford Computer Science, PhD Admissions FAQ: aligning with a faculty member after arrival.
• MIT EECS, Graduate Admissions FAQs: funding condition; post-admission visits.
• Council of Graduate Schools, April 15 Resolution: no obligation to decide before April 15.