Public Health PhD Statement of Purpose

Public Health PhD Statement of Purpose

A public health PhD statement of purpose is a research document. Harvard's PhD in Population Health Sciences caps it at 1,000 words and asks for "a self-assessment of any past research experience and analytical skills" and "a suggestion of up to three faculty members whose work most closely matches your research interests".

That changes what the statement is for. An MPH statement can spend a paragraph on why public health matters to you. A PhD statement that does the same spends part of a 1,000-word limit on something the prompt asks for in one point out of four. The reader wants to know whether you can run a study: what you have done with data, what you cannot do yet, and who on the faculty you would work with. When your draft is ready, choose the general PhD statement of purpose rubric for a review; it reads research direction, preparation and faculty fit.

Every Harvard, Boston University and UC Berkeley page quoted below was checked on October 31, 2026, for Fall 2027 entry where the page names a deadline. Recheck each one before you submit. More doctoral guides are in the PhD essays hub and the graduate school essays hub.

What a PhD prompt asks that an MPH prompt does not

Harvard's four points, verbatim:

your motivation in seeking doctoral-level training in population health sciences in your chosen field of study at Harvard a self-assessment of any past research experience and analytical skills a description of your current research interests and a suggestion of up to three faculty members whose work most closely matches your research interests your career aspirations.

BU's three points ask the applicant to "address why the applicant wishes to obtain a doctoral degree in epidemiology and the reasons for pursuing the degree at BUSPH", "comment on the applicants plans to use the education and training acquired at SPH" and "note any personal qualities, characteristics, and skills the applicant believes will enable them to be a successful epidemiologist."

Read together, the three pages point the same way: research and work experience carry the weight. Berkeley's admissions statistics list an average of six years of relevant professional or research experience, BU requires one to two years in a research setting, and Harvard gives preference to work experience or graduate work. Motivation is one of Harvard's four points; the self-assessment and the research interests are two more.

Our view: spend no more than one paragraph on why public health matters to you. Spend the rest on what you have done with data, what you want to study next, and who you would study it with.

The countercase: Harvard does not require a master's degree, so an applicant straight from a strong undergraduate degree has less research to assess. For that reader, the self-assessment is shorter and the coursework carries more weight: name the statistics and quantitative courses Harvard expects ("coursework in introductory statistics or quantitative methods"), and the thesis or project where you used them.

PhD, DrPH or MPH: which statement you are writing

An MPH statement is a third document, closer to a personal statement; the MPH personal statement examples cover it. If your strongest evidence is a program you ran rather than a study you analysed, you may be writing a DrPH statement. The DrPH statement of purpose guide covers that document. If you are choosing between master's options first, see MPH vs MS in epidemiology.

The self-assessment paragraph

Harvard's word is "self-assessment". It asks you to judge your own skills, not list them. A good one has three parts: what you did with data, with which tools; what you could not do; and what the PhD would teach you.

Constructed example, written by us, not a real applicant statement. Context (invented): an MPH graduate who has worked with routine health data.

Before: During my MPH and my work as a data officer, I gained strong analytical skills in Stata and R and experience with large health datasets. I am passionate about using data to improve maternal health and I am confident I can succeed in doctoral research.

After: For my MPH thesis I linked two years of routine maternal referral records from four districts and estimated, in Stata, how travel time to the referral hospital related to the interval between referral and arrival. I cleaned the records myself and documented every exclusion. The analysis was descriptive: I could not separate the effect of distance from the effect of which clinic referred the woman, because clinics differ in staffing and I had no staffing data. I can fit and interpret regression models in Stata and R. I have not yet used methods for clustered data or causal inference, and that is the training I am applying for.

The "before" says "strong" and "confident" and shows nothing. The "after" names the data, the tool, one result, one limit and the method that would fix it. It uses 108 words.

Home-country data as a research question, with feasibility

If your question uses data from your own country, show the reader it can be done:

Working with routine programme data in agriculture rather than health? See the agricultural economics master's statement of purpose.

• Access. Who holds the data, and do you have permission or a route to it? Say so in one sentence.
• Ethics. If the data identify people, say which approval it would need. You do not need the approval yet.
• Supervision. Name the faculty member whose methods fit, not only whose region fits.

You are not writing a proposal; Harvard asks for "a description of your current research interests", not a study protocol. But one sentence each on access and ethics turns an interest into something a reader can picture as a dissertation. The PhD proposal data access guide goes further if your program also asks for a proposal.

Faculty fit: up to three names, each with a reason

Harvard asks for "up to three faculty members whose work most closely matches your research interests." Three is a ceiling, not a target.

• Read two or three recent papers by each person before you name them.
• Give each name one reason tied to your question: a method, a dataset, a population.
• Name the field you are applying to first. Harvard says "Applicants are required to select a field of study in their application and indicate an area of specialization."
• Do not name someone only because they work on your region.

A sentence that works: "Professor [name]'s work on travel-time models for emergency obstetric care is the method I would use to separate distance from clinic effects." A sentence that does not: "I am inspired by Professor [name]'s outstanding research in global health."

A constructed maternal-referral statement, annotated

Constructed example, written by us, not a real applicant statement. Context (invented): Abena M., an MPH graduate from a Ghanaian university, three years as a regional health data officer, applying to the Epidemiology field of a public health PhD with a 1,000-word limit. Faculty names are left as placeholders on purpose. The excerpt is 493 words.

I want to train as an epidemiologist to answer one question about maternal referrals: when a health centre refers a woman in labour to a district hospital, how much of the delay before she arrives is distance, and how much is the system around her?

I began asking this as a regional health data officer. Each quarter I compiled referral records from district reports. Two districts with similar road distances to their hospitals reported very different referral-to-arrival times, and the difference did not shrink when I removed incomplete records. Our quarterly report could describe the gap but not explain it. That is when I realised I needed research training, not more reports.

For my MPH thesis I linked two years of referral records from four districts with travel-time estimates from a road network file. Using Stata, I estimated how travel time related to the referral interval and found a strong association, but I could not separate distance from the clinic that referred the woman. Clinics differ in staffing, transport arrangements and how they record times, and I had none of those variables. I wrote the limitation into my thesis and presented the results to the regional health directorate.

My analytical skills are real but bounded. I can clean and link administrative records, fit and interpret linear and logistic regression in Stata and R, and write code another analyst can rerun. I have not used multilevel models, I have no formal training in causal inference, and I have never designed a primary data collection. The doctoral coursework in epidemiologic methods and biostatistics is the training I need to move from describing delay to explaining it.

My current research interest is the structure of referral delay. I want to study whether the referral interval is better explained by facility characteristics, such as staffing and transport arrangements, than by distance alone, using multilevel models on routine data. The regional health directorate has allowed me to use anonymised referral records for my MPH work, and I would seek its permission and ethics approval again for doctoral research. The data are routine, so the question is feasible at low cost; the challenge is method, which is why I am applying.

Three faculty members' work matches this question. Professor [name A] develops methods for analysing clustered health-facility data, which is the structure of my records. Professor [name B] studies access to emergency obstetric care using travel-time models, and her recent work on road networks is close to my thesis method. Professor [name C] works on measurement error in routine health information systems, which bears directly on my concern that clinics record times differently.

My career aim is to lead an epidemiology unit that works with a ministry of health, so that routine data are analysed well enough to change how referrals are organised. I want to be able to tell a regional director not only that a gap exists, but what is causing it and what would close it.

What the annotations show.

• Motivation in one paragraph (paragraph 1). One question, stated before any biography. Harvard's first point is covered in 45 words.
• Self-assessment (paragraphs 3–4). Data, tool, result, limit; then what she can and cannot do. This is Harvard's second point, and it takes the most space.
• Research interest with feasibility (paragraph 5). Access, ethics and the method gap in four sentences.
• Faculty (paragraph 6). Three placeholders, each with a reason tied to her records or method. In your own statement, each name must be a current faculty member whose work you have read.
• Career (paragraph 7). Short and tied to the question.

Questions about public health PhD statements

I have no publications. Is that a problem?

Not by itself. None of the three program pages we checked asks for publications. Harvard asks for a writing sample, which may be "a term paper, senior thesis, master’s thesis, first authored report or manuscript, research report, or similar work." The PhD SOP guide for applicants with no publications covers how to write the research paragraph without them.

Do public health PhD programs require a master's?

It depends on the program. Harvard says "a previous graduate degree is not required." BU's Epidemiology PhD asks for a master's "in a related discipline". Berkeley's Epidemiology PhD says a master's "is preferable."

Is Harvard's personal statement a separate document?

No. Harvard's Population Health Sciences page lists the personal statement as "Required" in addition to the statement of purpose. Write the research case in the statement of purpose and keep the personal statement for the other document.

Do Harvard and BU want the GRE?

At Harvard's PHS program, "GRE General: Required", and "MCAT: Can be submitted in lieu of GRE". At BU's School of Public Health, "A GRE score is not required for application to BU School of Public Health and will not be considered if submitted."

Does BU waive the English test for Ghana or Nigeria?

BU's Epidemiology page lists Ghana, Nigeria and Uganda among the countries where a previous degree means "SPH Admissions will automatically waive the TOEFL/IELTS requirement". Kenya is not on that list. Read the full rule on BU's page before you rely on it. For master's applicants from East Africa, see the English-test rules for Kenyan, Ugandan and Ethiopian degrees.

Not ready for a doctorate yet? Doctors starting with a master's can use our guide to the SOP for an MPH after MBBS.

Get a free review of your statement of purpose

Get a free Quick Review of your statement of purpose against GradPilot's PhD statement of purpose rubric. It reads for research trajectory, research experience, research questions, faculty alignment and academic preparation. For the faculty paragraph alone, the PhD SOP adviser-fit rubric checks whether each name comes with a reason a faculty reader could verify. More on PhD SOP review is on the PhD lander.

Applying to a master's first? The epidemiology master's statement of purpose examples cover the MS and MSc version of this document.

Sources

Retrieved October 31, 2026.

• Harvard Kenneth C. Griffin Graduate School of Arts and Sciences, Population Health Sciences (deadline Dec 01, 2026): statement prompt and limit, four points, field selection, prior degree, statistics coursework, writing sample, personal statement, GRE and MCAT.
• Boston University School of Public Health, PhD in Epidemiology application requirements (deadlines November 1 and December 1, 2026): master's and experience requirements, statement points, GRE, English waiver list.
• UC Berkeley School of Public Health, Epidemiology PhD (no year stated): qualifications, admissions statistics.
• UC Berkeley School of Public Health, Doctor of Public Health (no year stated): DrPH purpose and experience requirement.