2027 Public Health PhD Research: Designing a Realist Evaluation Across Multiple Sites
Co-Founder and Chief Data Scientist
A multisite public health intervention can produce different outcomes in different communities - even when every site follows the same implementation plan. A realist evaluation investigates why, for whom, and under what conditions those differences arise. For prospective PhD students and researchers planning a dissertation, the challenge is to build a credible cross-site explanation without promising more access, data, or causal certainty than the project can support.
This guide explains the design, the doctoral preparation it demands, and the program questions to resolve before committing to the work.
Key Things You Should Know
- A multisite realist evaluation tests and refines explanations linking context, mechanism, and outcome; it does not simply rank sites by results.
- Dissertation feasibility depends on site permissions, comparable but context-sensitive evidence, ethical review, and an advisor experienced in theory-driven mixed-methods research.
- Compare PhD offers on protected research time, funding conditions, partner access, and supervision - not the degree title or online format alone.
What is a realist evaluation in public health research?
A realist evaluation examines how a program generates outcomes through mechanisms that operate under particular conditions. Its working unit is often a context-mechanism-outcome configuration, or CMO. Context includes features such as staffing, policy, community trust, and service access. A mechanism is the response or reasoning that program resources trigger - for example, whether a trusted referral makes a participant feel that seeking care is worthwhile. An outcome is the resulting change the evaluation investigates. The referral itself is an activity, not necessarily the mechanism.
Suppose a hypothetical community screening initiative has higher follow-up in some locations than others. A provisional theory might be that a personal referral prompts follow-up when patients trust the referring worker and can reach the clinic. Interviews, implementation records, and follow-up data could challenge or refine that theory. This is an illustrative design, not a finding or proof that the referral caused the difference.
Realist evaluation is especially useful when the research question concerns how an intervention travels across settings. If the primary question is an average effect under a specified comparison, a conventional impact design may be a better centerpiece; the approaches can also complement one another.
An MPH or research-oriented MS may provide a route into evaluation work, while a PhD is most relevant when the goal is an original, sustained scholarly contribution. Readers weighing that preparation can compare MPH and related public health graduate pathways.
How does a multisite realist evaluation work in public health?
The design begins with a program theory: a set of testable explanations of how the intervention is expected to work for different people and settings. Sites are then selected for informative variation, not merely convenience. For example, a researcher might compare locations with different staffing models or referral pathways while documenting differences in populations served. Variation helps test explanations, but poorly documented variation makes interpretation harder.
The following table shows what a cross-site comparison should distinguish. Each column answers a different question; an observed outcome alone cannot reveal the mechanism behind it.
| Evidence category | Illustrative screening-program question | Interpretive caution |
| Context | Are clinic access, staffing, and community relationships different? | Do not treat a site label as an explanation. |
| Program resource | Who offers referrals, and how are they delivered? | Delivery may differ from the written protocol. |
| Candidate mechanism | Does the referral prompt confidence, reassurance, or concern? | Infer reasoning from appropriate evidence, not outcomes alone. |
| Outcome | Who completes follow-up, and over what interval? | Check definitions, missing records, and who enters the denominator. |
A feasible dissertation follows a sequence that leaves room for the theory to change:
- Specify the population, intervention, outcome, time window, initial CMOs, and plausible rival explanations.
- Secure partner agreements and confirm what each site can provide before finalizing sampling or measures.
- Collect complementary evidence, such as documents, interviews, observations, and appropriately accessible routine data; document differences in how each source was produced.
- Analyze cases within each site first, then compare where a proposed mechanism appears to operate, fail, or take a different form.
- Refine the theory with disconfirming cases and report what the evidence cannot establish.
Comparable outcome definitions matter, but forcing identical interview questions or delivery practices onto unlike settings can hide meaningful context. A prospective doctoral researcher who mainly wants to help implement or monitor a program may need supervised evaluation experience rather than a dissertation; public health BS programs and career preparation describe an earlier-stage route into that work.
What does a public health PhD prepare graduates to do?
A public health PhD develops the ability to formulate research questions, select defensible methods, manage research ethics, analyze evidence, publish findings, and build an independent line of inquiry. For multisite realist work, that means explaining not only what happened at participating sites but also how confidently the evidence supports a transferable explanation - not a promise that another site will achieve the same result.
A practice-focused DrPH may be a better fit for someone whose chief objective is to lead improvement within an organization. A PhD is generally the stronger fit for an applicant who wants to develop and test theory, produce a dissertation, and compete for research-intensive roles. Neither degree substitutes for partner relationships or experience conducting interviews and interpreting program data.
If the immediate goal is advancement in practice rather than independent research, review MPH compensation and career advancement before assuming another doctorate is necessary.
Coursework varies by department and concentration. A candidate planning a realist evaluation should look beyond a general "research methods" requirement for training in evaluation theory, qualitative interviewing and analysis, epidemiology, quantitative analysis, mixed-methods integration, research ethics, and reproducible documentation. Independent study or cross-department courses may fill gaps, but their availability and approval should be confirmed before enrollment.
Methodological depth matters more than collecting course titles. The student should be able to explain why particular sites and participants were sampled, how interview accounts will be tested against other evidence, what outcome data can and cannot show, and how contradictory cases will alter the initial theory. Quantitative trends may describe differences across sites without identifying their causes; qualitative accounts may illuminate mechanisms without establishing how often they operate.
If the main gap is introductory evaluation or data analysis rather than readiness for original research, a focused course, supervised project, or inexpensive online public health master's option may be a more proportionate investment. Check whether that route actually offers the methods and supervision the project requires.
Requirements differ by university and discipline. Programs commonly assess academic preparation, research interests, writing, recommendations, and evidence of methodological readiness; a particular MPH, publication, GRE score, or prospective advisor's prior commitment should not be assumed to be universal. Applicants from clinical, social science, quantitative, community, and other backgrounds can be competitive when they can connect their preparation to a feasible research question.
For a multisite realist proposal, an applicant can demonstrate readiness with a concise initial program theory, a justified reason to compare settings, relevant methods experience, and a candid account of access still to be negotiated. Experience with one partner is valuable but is not permission to recruit across its network or use its records. Ask each program whether faculty must agree to supervise applicants before admission and how advisor matching works after entry.
If missing introductory public health or statistics preparation is the main barrier, examine course-level prerequisites rather than assuming another full degree is mandatory. Some prospective students may first explore public health master's conditional admission options; those policies do not establish eligibility for a PhD and must be checked separately.
Some doctoral study is offered online or in hybrid formats, but delivery mode does not determine whether a multisite dissertation is workable. Live seminars, campus visits, committee meetings, assistantship duties, and fieldwork can constrain a working student's schedule. Interviews across sites may require travel or evening appointments even when coursework is asynchronous.
Request a term-by-term schedule and ask how often students must attend in person, whether required courses run during working hours, and whether remote students receive equivalent access to methods mentoring and research groups. Confirm who pays for site visits and whether employment or caregiving commitments can be accommodated under actual leave and progression policies. For a broader comparison of formats and entry rules, see comparing online doctorate admission routes in public health; an accessible admission route is not necessarily a strong research fit.
Completion time depends on coursework, examinations or other advancement milestones, dissertation design, partner approvals, data collection, and the availability of faculty and sites. Multisite work adds coordination risk: one delayed agreement can affect sampling, travel, and analysis. Ask programs for their published completion policies and cohort outcomes rather than relying on an advertised minimum duration.
Compare offers using a written funding schedule that separates tuition coverage, stipend, health insurance, required work, summer support, research expenses, and continuation fees. Identify which years are guaranteed and which depend on grants, satisfactory progress, or renewed appointments. Budget separately for transcription, secure storage, and site travel; do not assume an advisor's grant will cover them. Applicants considering epidemiology-based departments may find additional funding questions in epidemiology doctorate scholarships and grants, while verifying every award against their own offer.
First identify the university's recognized institutional accreditor and confirm the status through the accreditor or the U.S. Department of Education's accreditation information. Then check the Council on Education for Public Health's published listing for the exact school, unit, or program and its scope. A university's institutional accreditation does not mean every public health degree holds separate CEPH accreditation, and a school's CEPH status should not be described as an individual PhD designation without checking its scope.
Ask the university which unit awards the PhD and request clarification if the website describes accreditation ambiguously. For comparison, reviewing accredited online MPH pathways concerns professional master's options; an MPH listing is not evidence about a particular PhD's accreditation or dissertation support.
Relevant paths include university research and teaching, research institutes, government evaluation, health-system research, and senior evaluation roles in nonprofits or consulting. The duties differ: a faculty applicant may need publications and grant experience, while an evaluation lead may need to negotiate with partners, manage teams, and turn findings into usable recommendations. A PhD alone does not confer a faculty appointment or leadership authority.
There is no single meaningful "public health PhD salary." The U.S. Bureau of Labor Statistics reports wages by occupation, not by this degree or by realist-evaluation specialization. When estimating earnings, compare the latest occupational data for the specific jobs you would pursue with actual vacancies, location, sector, and required experience; an occupation-wide median is not a graduate offer.
If the desired job is primarily program coordination or routine evaluation support, efficient public health degree pathways and earnings may be worth comparing with the time and opportunity cost of doctoral study.
What should you look for in a multisite realist evaluation PhD program?
The strongest fit is a program where the proposed question can survive contact with its supervision, funding, and partner arrangements. Look for faculty who can critique theory-driven evaluation, committee members who cover the needed qualitative and quantitative methods, and a realistic route to the sites. A well-known department without available mentorship or accessible data may be a weaker choice than a less prominent program with both.
Before accepting an offer, put these questions to the prospective advisor and doctoral director:
- Who can supervise realist evaluation, and who will advise on cross-site sampling, qualitative rigor, and outcome measurement?
- Which partner permissions, data-use agreements, community approvals, and institutional review determinations are needed before recruitment or record access?
- What happens to supervision, funding, and dissertation scope if an advisor leaves or a site withdraws?
- Who controls publication review, site identification, and authorship, and how are disagreements with partners handled?
- Which milestones and funding conditions apply if approvals or fieldwork take longer than planned?
Compare the answers in writing alongside each offer's funding terms and dissertation policies. Doctoral public health study costs and career directions can provide a broader comparison, but for this project the deciding evidence is whether a feasible research design and durable mentoring arrangement exist.
Other Things You Should Know About Public Health
Sometimes, but anonymity is difficult when a site's location, program, or distinctive characteristics make it recognizable. Agree on disclosure and publication review terms early, and do not promise anonymity that the dissertation cannot protect.
Follow the agreement and applicable review requirements, then consult the advisor about whether the remaining sites can still test the research questions. A contingency plan should address recruitment, permitted use of previously collected data, and changes to dissertation scope.
Records may document delivery and outcomes, but they seldom establish participants' reasoning by themselves. Whether interviews are necessary depends on the proposed mechanisms and what other ethically accessible evidence can test them.
Yes, when their contributions meet the relevant journal's authorship criteria and all authors can take the required responsibility for the work. Discuss contributions, review timelines, confidentiality, and disagreements before drafting.
- Supplementary Guide: Realist evaluation (HTML) | GOV. UK https://www.gov.uk/government/publications/the-magenta-book/supplementary-guide-realist-evaluation-html
- Ethics Review of Multijurisdictional Research https://ethics.gc.ca/eng/consultations_2021_multi_research-recherche.html
- Methodological approaches to study context in intervention implementation studies: an evidence gap map - PMC https://pmc.ncbi.nlm.nih.gov/articles/PMC9749183/
- Health Services Research and Implementation Science | ucsd.edu https://ph.ucsd.edu/phd/public-health/index.html
- PhD in Health Services | UW School of Public Health https://sph.washington.edu/program/phd-health-services
- Realist evaluation | betterevaluation.org https://www.betterevaluation.org/methods-approaches/approaches/realist-evaluation
- Realist Evaluation | undp.org https://erc.undp.org/methods-center/methods/evaluation-methods/realist-evaluation
