Which of the Following Studies Would Need IRB Approval?
Ever stared at a research proposal and wondered, “Do I really need to go through the whole IRB maze for this?In practice, the line between “needs review” and “doesn’t need review” is blurrier than most textbooks admit. Which means ” You’re not alone. The Institutional Review Board (IRB) can feel like a bureaucratic gatekeeper, but it’s also the safety net that protects participants from harm. Below, I walk through the most common study types you’ll encounter, explain why the IRB cares (or doesn’t), and give you a cheat‑sheet you can actually use next time you’re filling out that paperwork Easy to understand, harder to ignore..
What Is an IRB, Anyway?
Think of the IRB as a committee of people—researchers, ethicists, community members—who double‑check that a study treats human subjects ethically. Their job isn’t to police your science; it’s to make sure the people you’re studying aren’t put at unnecessary risk. S.Now, in the U. , the federal regulations (45 CFR 46, aka the Common Rule) define “human subject research” as a systematic investigation that involves living individuals about whom an investigator obtains data through interaction or identifiable private information And it works..
The Two Big Questions
- Is it research? If you’re systematically gathering data to answer a question, you’re probably in research territory.
- Does it involve human subjects? If you’re talking to people, observing them, or using data that could be linked back to them, the answer is yes.
If you answer “yes” to both, you’ve got a candidate for IRB review. But there are nuances—some studies fall into “exempt” or “expedited” categories, which still need a formal determination.
Why It Matters
Skipping the IRB isn’t just a paperwork shortcut; it can have real consequences. Day to day, imagine a psychology graduate student who runs a stress‑induction experiment without consent. Participants could walk away feeling anxious, the university could face a lawsuit, and the researcher’s career would probably be over. On the flip side, over‑filing for IRB review on a truly low‑risk activity can waste months of time and drain resources. Knowing which studies truly need IRB approval saves you from both legal trouble and needless red tape.
How to Decide: Study Types and Their IRB Status
Below is the meat of the article. I break down the most common study designs you’ll see in grant proposals, class assignments, or industry projects, and I flag whether they typically need IRB approval, an exemption, or no review at all.
1. Surveys and Questionnaires
Typical scenario: You email a 15‑question poll about dietary habits to a list of volunteers.
- Do you need IRB approval? Usually yes. Even if the survey is anonymous, the fact that you’re collecting data from living people about their behaviors makes it human subjects research.
- When can it be exempt? If the survey is anonymous (no way to link responses back to an individual) and the questions are non‑sensitive (e.g., “What brand of cereal do you eat?”), many IRBs will grant an exemption. Still, you must submit the survey to the IRB for an official determination.
Key tip: Keep the survey short, avoid collecting identifiable info, and label it as “anonymous” in your protocol. That’s the fastest route to an exemption The details matter here..
2. Interviews (Structured or Semi‑Structured)
Typical scenario: A sociology student conducts 30‑minute face‑to‑face interviews about experiences with remote work Not complicated — just consistent..
- Do you need IRB approval? Yes, almost always. Interviews involve direct interaction and can reveal personal, sometimes sensitive, information.
- Exempt possibilities? If the interview content is strictly about public behavior (e.g., “What time do you usually start work?”) and you record no identifying details, an IRB may deem it exempt. But the default is to go through the board.
Key tip: Draft a consent script and let participants know they can skip any question. That small step can make the IRB review smoother.
3. Observational Studies
Typical scenario: A public health researcher watches foot traffic in a city park and tallies how many people wear masks.
- Do you need IRB approval? It depends. If you’re observing public behavior where there’s no reasonable expectation of privacy, you generally don’t need IRB review.
- When does it become a human subjects study? If you start recording identifiable details (faces, license plates) or you move into a private setting (e.g., a clinic waiting room), the IRB steps in.
Key tip: Stick to counting or noting non‑identifiable traits when you’re in a public space. If you need to capture images, blur faces and get a waiver Not complicated — just consistent..
4. Experiments Involving Deception
Typical scenario: A psychology experiment where participants think they’re being tested on memory, but the real focus is on stress response Less friction, more output..
- Do you need IRB approval? Absolutely. Deception raises ethical flags because participants aren’t fully informed.
- Special considerations: The IRB will require a debriefing plan and a justification that the deception is essential and that no less risky method exists.
Key tip: Write a clear debrief script before you even start the experiment. The IRB loves to see that you’ve thought about participants’ feelings after the fact Still holds up..
5. Biological Sample Collection (Blood, Saliva, DNA)
Typical scenario: A biotech startup collects saliva swabs from volunteers to study microbiome diversity.
- Do you need IRB approval? Yes. Even though the sample is a tiny bit of tissue, it’s considered identifiable private information.
- Exempt? Rarely. The only time you might get an exemption is if the samples are completely anonymized at the point of collection and you’re not planning any future re‑identification.
Key tip: Use a coded labeling system and keep the key separate from the data. That can sometimes move the study into an expedited review rather than full board review.
6. Secondary Data Analysis
Typical scenario: A data scientist re‑analyzes a publicly available dataset of hospital admissions.
- Do you need IRB approval? Usually no. If the dataset is publicly available and de‑identified, you’re not dealing with human subjects.
- When does it change? If the dataset includes limited data sets under HIPAA (e.g., dates, ZIP codes) or you can link it back to individuals, the IRB must weigh in.
Key tip: Check the data use agreement. If it says “no re‑identification,” you’re safe. If it’s ambiguous, send a quick email to the IRB for clarification Surprisingly effective..
7. Quality Improvement (QI) Projects
Typical scenario: A hospital implements a new discharge checklist and measures readmission rates.
- Do you need IRB approval? Often not. QI projects are aimed at improving internal processes, not generating generalizable knowledge.
- But watch out: If you intend to publish the results or present them at a conference, the project may be re‑classified as research, triggering an IRB review.
Key tip: Label the project as “QI” in all documentation, and if you later decide to publish, submit a retrospective IRB request.
8. Usability Testing for Software
Typical scenario: A tech company asks users to complete tasks on a new app while recording screen interactions.
- Do you need IRB approval? Usually yes. Even though it’s a “product test,” you’re collecting data about how people think and behave, which can be sensitive.
- Exempt possibilities? If the test is anonymous, you don’t collect any personal identifiers, and the tasks are low‑risk, many IRBs will grant an exemption.
Key tip: Use a random ID for each participant and store the key separately. That small step can move you from full review to exemption.
9. Educational Research in Classroom Settings
Typical scenario: A teacher surveys her class about study habits and compares grades before and after a new teaching method.
- Do you need IRB approval? Often. Even though it’s “just a class,” you’re still studying human subjects.
- Exempt? If the data is anonymous and the study is strictly for internal improvement, an exemption is possible. Even so, if you plan to publish, the IRB will likely want a full review.
Key tip: Get parental consent for minors, and anonymize grades before analysis And that's really what it comes down to. And it works..
10. Social Media Mining
Typical scenario: A marketing analyst scrapes public Twitter posts about a new product launch.
- Do you need IRB approval? Generally no. Public tweets are considered public domain.
- When does it become IRB‑relevant? If you start linking tweets to user profiles, collect private messages, or combine the data with other identifiable sources, you cross the line.
Key tip: Stick to publicly available data and avoid any attempt to de‑identify users beyond what’s already public.
Common Mistakes / What Most People Get Wrong
-
Assuming “anonymous” = “exempt.” Many think that if a survey doesn’t ask for a name, the IRB can be ignored. The truth is you still need an official exemption determination But it adds up..
-
Mixing public and private data without clearance. A researcher might observe a public park, then later pull a participant’s name from a city registry. That combo triggers a full review.
-
Forgetting about “minimal risk.” Even low‑risk studies can be non‑exempt if they involve vulnerable populations (children, prisoners).
-
Skipping the debrief for deception studies. The IRB will reject any protocol that leaves participants in the dark after the experiment.
-
Thinking “quality improvement” never needs IRB. Publish a QI result and you’ll be asked for a retrospective review.
Practical Tips – What Actually Works
- Start with the IRB’s decision tree. Most institutions publish a flowchart; follow it step‑by‑step before you write your protocol.
- Create a one‑page “risk assessment.” List potential harms, how you’ll mitigate them, and who will monitor the study. IRBs love concise risk tables.
- Use a template consent form. Even if your study is exempt, having a consent script ready shows you’re serious about ethics.
- Separate identifiers from data early. Store the key in a locked folder (digital or physical) and give the data analyst a de‑identified set.
- Document everything. Email threads with the IRB, version‑controlled protocol drafts, and consent logs—these become the evidence you need if a question arises later.
- Know the “expedited” categories. Blood draws, non‑invasive physiological recordings, and some questionnaire studies can go through an expedited review, which is faster than a full board meeting.
FAQ
Q1: Do I need IRB approval for a short, informal focus group with 5 friends?
A: If the focus group is purely for personal learning and you won’t publish or share the data outside the group, you’re likely outside the scope of human subjects research. But as soon as you plan to use the data for a paper, you need IRB review.
Q2: My study uses only publicly available census data. Do I need an IRB?
A: No. Publicly released, de‑identified data sets are exempt from IRB oversight. Just keep the data source citation handy.
Q3: Can I get an exemption after the study is done?
A: Technically you can submit a “retrospective exemption request,” but it’s risky. Many IRBs will require you to stop the study until a determination is made Simple as that..
Q4: I’m collecting video footage in a public square. Do I need consent?
A: If the footage is non‑identifiable (no faces, license plates blurred) and you’re only counting behaviors, you likely don’t need consent. If you capture identifiable individuals, you need consent or a waiver Not complicated — just consistent..
Q5: Does a “pilot study” automatically need IRB approval?
A: Yes. A pilot is still research. The only time you might skip formal review is if the pilot is part of a larger exempt study and you’ve already received that determination.
Bottom Line
The short version is: **If you’re systematically gathering data from or about living people, you probably need IRB approval—or at least an official exemption.Plus, ** The safest route is to submit a brief protocol early, let the board tell you where you fall, and then adjust your methods accordingly. It may feel like an extra step, but the protection it offers to participants—and the peace of mind it gives you—makes it worth the time Which is the point..
So next time you draft a study proposal, pause, run through the checklist above, and send that email to the IRB office. It’s the smartest thing you can do for your research, your participants, and your career. Happy studying!
