03 · What You Need to Know
A research protocol explains the study before the evidence exists
A research paper explains what was done after a study has been conducted. A protocol works in the opposite temporal direction. It describes the intended study before or during its implementation.
That distinction gives the protocol several functions.
It helps researchers determine whether the different parts of the study fit together. It provides a common reference for people conducting the research. Where formal review is required, it gives reviewers enough information to evaluate the proposed study. It can also preserve a record of what was intended before results were available and provide a basis for documenting subsequent changes.
The World Health Organization's recommended protocol format illustrates this breadth. Its guidance includes the rationale and background, objectives, study design, methodology, participant safety and follow-up where applicable, data management and analysis, quality assurance, expected outcomes, dissemination, project duration, anticipated problems, project management, and ethical considerations.
A protocol is more than a methods section
A methods section and a protocol overlap substantially, but they serve different purposes.
A journal article's methods section tells readers what was done and provides the information needed to understand and evaluate the resulting study. A protocol must support implementation prospectively. It may therefore contain operational details that would be unnecessary or handled differently in a published article.
For example, WHO recommends that protocols describe not only the study design and methodology but also project duration, anticipated implementation problems, team responsibilities, data management, quality assurance, ethical considerations, and dissemination. For human-participant studies, instruments and informed-consent materials may also form part of the protocol package submitted for review.
A protocol should therefore answer not merely:
What method are we using?
but also:
Exactly how will this study operate?
What does a research protocol usually contain?
There is no single protocol structure required for all research. The appropriate contents depend on the methodology and on any institutional or regulatory requirements.
Still, several elements commonly need explicit treatment.
| Protocol element |
What it clarifies |
| Background and rationale |
Why the study is needed and how it relates to existing knowledge |
| Research question and objectives |
What the study is intended to answer or accomplish |
| Study design |
The overall methodological structure connecting the question to evidence |
| Population, cases, or evidence source |
Who or what will provide the evidence and under what eligibility or selection rules |
| Sampling |
How participants, cases, observations, documents, records, or other units will be selected |
| Procedures |
What will happen, in what sequence, by whom, and under what conditions |
| Instruments and measures |
How relevant information will be collected, measured, elicited, or recorded |
| Data management |
How research data will be organized, coded, protected, monitored, stored, and documented |
| Analysis |
How the evidence will be transformed into findings capable of answering the research question |
| Quality procedures |
How consistency, accuracy, adherence to procedures, or other quality requirements will be maintained |
| Ethical considerations |
Risks, protections, privacy, confidentiality, consent, and other relevant ethical issues |
| Timeline and responsibilities |
When major activities will occur and who is responsible for them |
| Dissemination |
How findings are expected to be communicated or used |
WHO's recommended format includes these kinds of elements and specifically asks investigators to describe the study population, eligibility, methodology, instruments, data handling, analysis, quality assurance, timeline, anticipated difficulties, responsibilities, and ethical considerations.
Do not interpret the table as a mandatory universal template. A secondary-data study, laboratory experiment, qualitative field study, and clinical trial may need quite different protocol structures.
Different studies may need different protocol templates
The existence of multiple protocol formats is a useful reminder that “research protocol” is a category of document rather than one universal form.
The NIH Intramural Research Program, for example, currently provides separate protocol templates for behavioral and social science research, interventional drug and device trials, natural history and observational studies, prospective data collection, repositories, retrospective data or biospecimen review, and secondary research using existing data, biospecimens, or medical records.
This matters because researchers sometimes obtain a protocol template from an unrelated study type and assume every heading must be completed. A template designed for an interventional clinical trial may contain requirements that make little sense for an archival study. Conversely, a lightweight generic template may omit information required for a regulated intervention.
Use the protocol format required by the responsible institution when one exists. When no formal template is mandated, adapt the structure to the methodological and operational needs of the study.
A research proposal and a research protocol overlap, but they are not necessarily the same document
The distinction varies among institutions, so it should not be treated as universal terminology.
In general, a research proposal is often designed to justify a project. It explains why the study matters, what it intends to accomplish, why the proposed approach is appropriate, whether it is feasible, and sometimes why it deserves approval or funding.
A research protocol is usually more implementation-oriented. It specifies how the approved or planned study will actually be conducted.
Research proposal
Often emphasizes why the project should be undertaken, approved, supported, or funded.
Research protocol
Emphasizes the detailed plan governing how the study will be conducted.
The documents can overlap extensively. A sufficiently detailed proposal may already contain much of the eventual protocol. Some institutions may even use the terms differently or combine the functions into one document.
For that reason, follow the terminology and requirements of the institution, ethics committee, funder, sponsor, registry, or program responsible for your project rather than insisting on a universal semantic boundary.
A protocol is also different from a project timeline
A timeline answers when activities are expected to happen. The protocol explains the scientific and operational logic governing those activities.
For example, a timeline might say:
June–August: conduct interviews.
The protocol should explain who will be interviewed, why they are eligible, how they will be recruited, how consent will work where applicable, who will conduct the interviews, what guide will be used, how recordings and notes will be handled, and how the resulting material will be analyzed.
The timeline is therefore one component of the broader plan, not a substitute for it.
The protocol can expose contradictions before the study begins
One of the most useful reasons to write a protocol is that writing forces decisions that can remain conveniently vague in conversation.
Imagine trying to complete these sentences:
- The primary objective of the study is...
- Eligible participants are...
- Participants will be recruited by...
- The primary outcome is...
- The information required for the main analysis will be collected using...
- The sample size is based on...
- The data will be analyzed by...
- Identifiable information will be accessible to...
If several cannot yet be completed, the protocol has identified unresolved planning work.
NIH's protocol-development guidance similarly notes that protocol and consent templates help researchers organize the study and meet applicable regulatory requirements, while its available templates are intended to be adapted to the particular research rather than used mechanically.
In this sense, a protocol is not merely a record of decisions. Writing it can help produce better decisions.
Create the protocol after the broad design exists, not before you know what study you are doing
There is little value in producing a highly detailed protocol for a research idea whose basic question, evidence requirements, and design remain unstable.
Before detailed protocol development, you should generally have enough clarity about the research question, purpose, intended evidence, broad design, population or evidence source, and feasibility to begin making operational decisions.
If those elements remain uncertain, first resolve the decisions that follow from settling the research question.
This does not mean waiting until every detail is final before opening a protocol document. Drafting can begin early and evolve with the study. The distinction is between using a protocol as a working structure and prematurely treating an unstable plan as final.
Create it before important decisions need to become operational
The other boundary is equally important. Do not wait until halfway through data collection to document how the study was supposed to work.
A protocol becomes particularly valuable before:
- submitting the study for ethics or regulatory review where applicable;
- seeking formal site or institutional authorization when a study plan is required;
- registering a study where protocol information must be supplied;
- training multiple researchers to implement common procedures;
- beginning recruitment or data collection;
- implementing interventions;
- collecting information that will later be difficult or impossible to recover; or
- making expensive commitments based on the planned design.
For WHO-supported research involving human participants, the protocol is part of the documentation submitted for ethics review along with study instruments, informed-consent forms, and other associated materials. WHO's review process checks that the necessary documentation is present before a proposal proceeds to ERC review.
Your own institution may have different submission requirements, but the principle remains useful: the protocol should exist before a reviewer or research team needs a stable account of what the study will do.
For human-participant research, the protocol and ethics application are closely related but not identical
The protocol describes the study. Ethics review evaluates the ethical acceptability of that study under the applicable framework.
WHO's protocol guidance therefore includes scientific and methodological details alongside safety, informed consent, privacy-related considerations, and other ethical issues. Its ethics-review guidance requires the research protocol and associated study documents for covered WHO-supported research involving human participants.
A weak scientific design can itself create an ethical problem because participants may be exposed to burden or risk without a study capable of producing meaningful evidence. WHO's ethics guidance explicitly links scientific design and methodology to the ethical integrity of research.
Do not therefore think of the protocol as the scientific document and the ethics application as an unrelated administrative document. They address different functions but should describe the same study.
The protocol should match what participants will actually experience
For human-participant research, participant-facing documents should be consistent with the protocol.
If the protocol says interviews will last approximately 45 minutes but participant information says two hours, something is wrong. If the protocol includes audio recording but the consent materials do not explain it where disclosure is required, the documents are misaligned. If the recruitment materials promise anonymity while the protocol requires retaining identifiable contact information, the terminology and procedures need clarification.
WHO's informed-consent guidance treats the information sheet and consent certificate as parts of the consent documentation and emphasizes adapting them to the particular study.
Consistency across the protocol, instruments, recruitment materials, consent documents, data-management arrangements, and actual procedures is therefore part of study readiness.
A protocol can coordinate a research team
For a one-person project, informal knowledge can sometimes conceal missing documentation. The researcher knows what they meant even when the written plan does not quite say it.
That becomes dangerous as the team grows.
If three interviewers use different interpretations of eligibility, two laboratories process specimens differently, or several sites follow different versions of the procedure, the study may accumulate unwanted variation.
WHO's recommended format specifically notes that when multiple sites are engaged in a specified protocol, the methodology should be standardized and clearly defined. It also recommends documenting project-management roles and responsibilities.
A protocol therefore provides a shared reference point. Depending on study complexity, it may be supplemented by manuals of operations, standard operating procedures, data dictionaries, analysis plans, or other specialized documents.
The protocol should say enough about analysis to test whether the data will be adequate
Analysis planning is not merely something to write because the protocol template contains an analysis heading.
Thinking about analysis before data collection tests whether the study will generate the information required to answer its question.
WHO recommends that protocols explain how data will be managed and analyzed, including the proposed statistical methods, sample-size reasoning, and approaches to missing or spurious data where relevant. For qualitative research, it calls for sufficient detail about how the data will be analyzed.
The required detail differs across methodologies. A confirmatory clinical trial may require extensive advance specification. An exploratory qualitative study may preserve legitimate analytic flexibility. What matters is that the protocol's analysis section is consistent with the research question, design, and data being collected.
The protocol should identify foreseeable implementation problems
Research protocols are sometimes written as though the study will operate in a frictionless environment.
WHO takes a more practical view. Its recommended format includes a section on anticipated problems and asks investigators to discuss difficulties they expect in completing the project within the proposed timeframe and funding, together with possible solutions.
This is useful even when your institution does not require such a heading.
Ask what could realistically threaten implementation:
- slow recruitment;
- limited site access;
- equipment availability;
- participant attrition;
- language barriers;
- data-access restrictions;
- staff turnover;
- technical failures;
- seasonal constraints; or
- analytical expertise that is not yet secured.
The protocol need not predict every inconvenience. It should show that obvious threats to feasibility have not been ignored.
A protocol can remain a controlled living document
Creating a protocol does not imply that the study can never change.
Piloting may reveal a procedural problem. Recruitment may prove slower than expected. New safety information may emerge. A technical procedure may become unavailable. Qualitative research may include forms of iteration that are part of the methodological design.
The important issue is how changes are governed.
A protocol should have a version and date so that researchers can identify which plan was in effect. WHO's submission guidance explicitly requires protocol version information for its ethics-review process and asks investigators responding to review to submit a revised protocol incorporating requested amendments.
Formal amendment requirements depend on the governing institution, ethics body, regulator, funder, or registry. For your own project, determine which changes can be handled through internal documentation and which require review or approval before implementation.
If an established plan changes, follow the principles for responding when the research plan changes rather than silently editing the protocol until it appears that the revised procedure had always been intended.
A protocol is not automatically a preregistration
Both documents can specify research plans prospectively, but they are not interchangeable concepts.
A protocol is the operational plan for the study. Preregistration generally involves creating a time-stamped record of specified research decisions in a registry or other system before a defined point in the research process.
A protocol may be used as the basis for preregistration, and some registries require substantial protocol information. A protocol can also exist without being publicly registered.
Likewise, registering selected hypotheses and analyses does not necessarily provide the operational detail needed to conduct the entire study.
Follow the requirements of the relevant registry, journal, funder, regulator, or research design where registration or preregistration applies.
A protocol is not automatically a published protocol paper
Some journals publish protocol articles describing planned studies. Publishing such an article can make the planned methodology publicly available and may be useful in particular fields.
But a research protocol exists independently of whether it is published as a journal article.
You can have a formal protocol that is never published as a protocol paper. Conversely, preparing a manuscript for a protocol article may involve journal-specific reporting and editorial requirements beyond the internal or ethics-review protocol used to run the study.
How detailed should the protocol be?
Detailed enough that the relevant reader can understand what will happen and evaluate or implement the study appropriately.
That reader may be the research team, ethics committee, funder, regulator, collaborating site, sponsor, data custodian, or another oversight body.
A complex intervention involving several sites will require more operational detail than a straightforward analysis of an existing public dataset. NIH's use of different protocol templates for behavioral interventions, observational research, repositories, retrospective reviews, and secondary research illustrates how protocol needs vary by study type.
The question is therefore not “How many pages should a protocol be?” but “Does it contain enough detail for this study and this purpose?”
If you are uncertain how much detail is appropriate at your current stage, consider how detailed the research plan needs to be before you begin.
Do not create protocol detail merely to make the document longer
A long protocol can still be vague where it matters.
Several pages of background literature do not compensate for unclear eligibility criteria. A detailed description of statistical software does not compensate for an undefined primary outcome. A beautifully formatted timeline does not solve an unresolved recruitment pathway.
Protocol detail is valuable when it reduces ambiguity, supports implementation, enables review, protects participants, or clarifies how the evidence will answer the research question.
Anything else risks becoming paperwork about paperwork, a genre with which academia is already adequately supplied.
