How to Write Your First IRB Protocol: A Step-by-Step Guide

An IRB protocol is the narrative document that tells the review board exactly what you will do, to whom, and why. Most institutional forms are structured as a set of prompts that together reconstruct the protocol, but the underlying document is the same everywhere. If you can write a strong standalone protocol, filling in any institution-specific form becomes mechanical.

What the protocol is for

The protocol serves three audiences at once: the IRB reviewer assessing risk, your advisor confirming feasibility, and your future self six months into data collection trying to remember what you promised. Write it so all three can read it without outside context. Precision is more important than eloquence.

Section 1: Background and significance

Two to three paragraphs. State the problem, summarize the most relevant prior findings (three to five citations is plenty), and identify the gap your study fills. Reviewers do not want a dissertation literature review here — they want enough context to judge whether your study design is proportional to what is already known.

Section 2: Research questions and hypotheses

State each question as a single sentence. If the study is hypothesis-driven, state the hypothesis. If it is exploratory (common in qualitative research), say so explicitly and describe the research aims instead. Do not mix the two without noting which is which.

Section 3: Study design and methods

This is where most protocols fall apart. Describe, in order:

• Study type (observational, experimental, qualitative, mixed methods).
• Participant activities, step by step, in the order they will occur.
• Duration of each activity and total time commitment.
• Location (in person, online platform name, phone).
• Instruments (surveys, interview guides — attach as appendices).

Use the verb "will" throughout. Reviewers are deciding whether to authorize future activity, not reading a past-tense report.

For quantitative studies, the sample size justification belongs here. Power analyses should cite the software used and the effect size assumed. A good primer on sample size justification is available from Stats for Scholars — reviewers notice when the sample size language is vague or hand-waved.

Section 4: Participants and recruitment

Specify inclusion and exclusion criteria as numbered lists. State the target sample size. Describe recruitment channels (flyers, listserv, social media, snowball) and attach every recruitment text as an appendix. If you recruit through a gatekeeper (a school, clinic, community organization), include a signed letter of cooperation.

Equity matters here. Reviewers will ask why you excluded certain groups. "Convenience" is not an acceptable reason for excluding protected populations unless you can justify it methodologically.

Section 5: Informed consent process

Do not just attach the consent form — describe the process. Who will present consent? In what setting? How will you confirm comprehension? How will you document consent (wet signature, electronic signature, verbal with log)? If you are requesting a waiver of documentation or a waiver of consent, cite the applicable regulatory basis, generally 45 CFR 46.117(c) for documentation waiver and 45 CFR 46.116(f) for consent waiver.

Section 6: Risks and benefits

Be honest and specific. The three risk categories are physical, psychological, and social (including confidentiality breach and economic harm). For a survey about stressful experiences, acknowledge emotional distress as a possible risk and describe your mitigation (ability to skip questions, debrief with hotline numbers).

Benefits are often overstated. "Advancing scientific knowledge" is not a direct benefit to participants — it is a societal benefit. List it honestly, but do not dress it up.

Section 7: Data management

Describe what you will collect, where it will live, who has access, and how long it will be retained. Specify encryption at rest and in transit, access controls, and the destruction plan. Our data management plan template covers the common fields. If your study produces identifiable qualitative data, see our consent form guide for how to describe recording and transcription in the consent document.

Section 8: Compensation

State the amount, timing, and prorating rules. Compensation that is contingent on completing the entire study can be seen as coercive — reviewers will ask you to prorate.

Section 9: Conflicts of interest

Disclose any financial or personal interest that could bias the study. If none, say so.

Before you submit

Read the protocol out loud once. Every sentence that requires a "it depends" clarification is a sentence the reviewer will flag. Rewrite those sentences as definite statements. Then run through our pre-submission checklist to catch the missing pieces most first-time applicants forget — typically the recruitment text, the debriefing script, and the data destruction date.

The first protocol takes twenty hours to write. The second takes six. The friction is front-loaded, and every hour you spend on the first one compounds every protocol you write after.