Defining Research with Human Subjects in Social, Behavioral, and Economic Sciences
If you've ever handed out a survey in a classroom, conducted interviews about people's shopping habits, or observed how strangers interact in public spaces — you've probably already crossed paths with human subjects research regulations. But here's the thing: many researchers in social, behavioral, and economic (SBE) fields aren't entirely sure where the line is between "doing research" and "needing IRB approval." And that ambiguity can get you into real trouble Practical, not theoretical..
Most guides skip this. Don't Not complicated — just consistent..
The truth is, defining human subjects research isn't as straightforward as it sounds, especially in SBE contexts where your "subjects" might be college students, online survey respondents, or people going about their daily lives in places you happen to be studying. So let's clear this up.
What Is Research with Human Subjects?
At its core, research with human subjects involves obtaining data from or about living individuals in a systematic way, with the intent to generalize findings or contribute to scholarly knowledge. But that broad definition hides a lot of nuance — and that's where people get stuck Most people skip this — try not to..
Here's what the regulations actually care about: you're conducting human subjects research when you're doing all three of these things:
- You're engaging in a systematic investigation designed to develop or contribute to generalizable knowledge
- You're obtaining data from living individuals
- You're either interacting with those individuals or obtaining identifiable private information about them
That third point is worth sitting with for a moment. Their test scores. Now, their responses to a survey. Their behavior in a focus group. It's not just about talking to people — it's about accessing information that they would reasonably expect to remain private. Their browsing history No workaround needed..
In SBE research specifically, this definition captures a wide range of activities. Practically speaking, a psychologist studying memory retention through cognitive tests. An economist analyzing bank records to understand spending patterns. A sociologist conducting ethnographic fieldwork in a community. Because of that, an education researcher surveying teachers about classroom practices. All of these typically involve human subjects Small thing, real impact..
Honestly, this part trips people up more than it should Simple, but easy to overlook..
What About Studies That Don't Count?
Here's what trips up a lot of researchers: not everything involving people is technically "human subjects research" under the regulatory definition And that's really what it comes down to..
If you're just collecting publicly available data — census figures, published statistics, information from news archives — you're not working with human subjects, because that information isn't private. Similarly, if you're analyzing existing datasets that have been fully de-identified (meaning there's no way to link the data back to specific individuals), you might not need full IRB review, though you should still consult with your institution.
And if you're simply gathering information for internal program improvement — say, a nonprofit collecting feedback from clients to make their services better — that's typically considered quality improvement or program evaluation, not research. The key distinction is whether you're trying to contribute to generalizable knowledge that will be shared beyond your immediate context.
The SBE Context: Why It Gets Complicated
Research in social, behavioral, and economic sciences has some unique features that make the human subjects definition particularly tricky.
For one thing, your "intervention" is often much lighter than in biomedical research. You're not giving people experimental drugs — you might be showing them a persuasive message, asking them to solve a puzzle, or having them play a budgeting game. But these interventions still affect people, and participants might experience psychological discomfort, embarrassment, or other harms you're not anticipating.
For another, the data you collect is inherently personal. People's attitudes, beliefs, behaviors, and economic situations are sensitive. Even "harmless" surveys about shopping preferences can reveal information participants don't want shared.
And then there's the matter of deception. Some SBE studies intentionally withhold information from participants — say, a psychology experiment where participants think they're evaluating one thing but you're actually measuring something else. Deception is allowed in some circumstances, but it triggers additional review and requires specific protocols for debriefing.
And yeah — that's actually more nuanced than it sounds Simple, but easy to overlook..
Why This Definition Matters
Why should you care about getting this right? Because the consequences of getting it wrong range from minor (delays in your project) to serious (federal funding pulled, research terminated, reputational damage) And it works..
The federal regulations governing human subjects research — known as the Common Rule — exist because research has caused real harm to people. The Tuskegee syphilis study. The Stanford prison experiment. And the CIA's MKULTRA program. These aren't just historical footnotes; they're the reason we have review boards, consent requirements, and protections for vulnerable populations But it adds up..
Not obvious, but once you see it — you'll see it everywhere.
In SBE research, the harms tend to be psychological rather than physical — stress, embarrassment, loss of privacy, or even retaliation if sensitive information gets out. But those harms are still real, and participants deserve protection.
Beyond the ethical imperative, there's a practical one. And universities have legal obligations to oversee research conducted under their auspices. If you receive federal funding, you're required to comply with human subjects regulations. And if you ever want to publish your work in a peer-reviewed journal, editors will want to know you handled human subjects appropriately.
Here's what most people miss: the definition applies regardless of funding status. Even if you're running a study with no budget, on your own time, you're still subject to these requirements if you're conducting human subjects research. Your institution's policies apply to all research conducted by its faculty and students, funded or not.
How Human Subjects Research Is Reviewed
Once you've determined you're conducting human subjects research, the next question is: what level of review do you need?
Exempt Review
Some research is considered so low-risk that it doesn't require ongoing oversight. Exempt doesn't mean "not human subjects research" — it means the review board has determined the study poses minimal risk and meets certain criteria Most people skip this — try not to..
Common examples in SBE research include:
- Surveys or interviews where responses are anonymous and no sensitive information is collected
- Studies using publicly available data or existing datasets that are fully de-identified
- Observation of public behavior where individuals aren't identified in published work
- Educational research in normal instructional settings
Even exempt research usually requires some form of initial review to confirm it qualifies.
Expedited Review
Studies that pose greater than minimal risk but still fall within certain categories can often be handled through expedited review — a single reviewer rather than the full board. This covers things like studies collectingaudio or video recordings with consent, research involving sensitive topics where confidentiality is carefully protected, and studies with minor interventions Small thing, real impact..
Full Board Review
Research involving more than minimal risk, vulnerable populations (children, prisoners, pregnant women, people with impaired decision-making capacity), or significant interventions requires full board review. This is more time-intensive but ensures careful scrutiny of your protocols It's one of those things that adds up..
Common Mistakes and What People Get Wrong
Let me be honest: the line between what's reviewable and what isn't isn't always clear, and plenty of well-meaning researchers have made mistakes. Here are the most common ones:
Assuming small or informal studies don't count. Running a quick survey in your graduate seminar? That might still be human subjects research if you're collecting data systematically with the intent to learn something generalizable. The fact that it's small doesn't automatically make it exempt.
Confusing "anonymity" with "not human subjects." Even if you never collect names, your study might still qualify as human subjects research if you're interacting with living individuals to obtain information. Anonymity is a protection, not a blanket exemption.
Forgetting about existing data. If you're analyzing student records, medical data, or any dataset that contains identifiers, you may need IRB approval even though you're not "collecting" data in the traditional sense. The review board needs to ensure the original consent covered this use.
Underestimating risk in SBE research. Just because you're not handing out experimental medications doesn't mean participants can't be harmed. Psychological distress, embarrassment, and privacy breaches are real concerns. Review boards take these seriously.
Skipping the consultation. Most institutions offer free consultations where you can describe your study and get guidance on whether review is needed and what category applies. People who skip this step often end up having to revise their protocols anyway, which takes longer than asking first.
Practical Tips for Navigating This
If you're new to human subjects research or feel overwhelmed by the regulations, here's what actually works:
Start early. Here's the thing — iRB review takes time — sometimes weeks for expedited, months for full board. Build this into your timeline from the beginning, especially if you're working toward a grant deadline or academic calendar.
Write your protocol for the review board, not just for yourself. What if someone becomes distressed? In real terms, what if data is breached? Here's the thing — anticipate the questions a skeptical reviewer might ask. Be clear about what you're doing, why, and how you'll protect participants. What if a participant wants their data removed?
Use established templates and examples. Your institution likely has sample consent forms, protocols, and checklists. These exist because reviewers are familiar with them and know what to look for. Reinventing the wheel rarely helps.
When in doubt, ask. A quick email to your IRB office describing your study can save enormous headaches later. Most researchers who've been through this process will tell you: the board is there to help you protect your participants, not to block your research And that's really what it comes down to. Still holds up..
Document everything. In practice, keep records of your review determination, your protocol, consent forms, and any modifications. If questions come up later — during publication, a funding audit, or an inquiry — you'll want clear documentation.
Frequently Asked Questions
Do I need IRB approval for a survey I'm doing for my thesis?
Probably yes, unless your study meets very specific exempt criteria. But student projects generally do need review, even if they're small. Check with your institution.
What if I'm only observing people in public places?
It depends. Which means if you're observing behavior in public spaces where people have no reasonable expectation of privacy and you won't be identifying individuals in your research, this may not require full review. But you should still confirm with your IRB Worth knowing..
Can I use data I collected before without IRB approval?
This is tricky. If you collected data previously without proper approval, you may face limitations on how you can use or publish it. Here's the thing — future research using that data might require retrospective review or may not be possible. This is one area where it's much easier to get approval upfront than to deal with the consequences later.
What happens if I don't get approval when I should have?
Consequences vary. Your institution might require you to stop the research. Published papers might be retracted. That's why if you're receiving federal funding, you could face more serious sanctions. The best approach is always to get ahead of this.
Does working with existing datasets require IRB approval?
Often yes, especially if the data contains identifiers or could be linked back to individuals. You may need to verify that original participants consented to this type of secondary analysis, or your review board may need to make that determination.
The Bottom Line
Getting clear on whether your work constitutes human subjects research isn't just about checking boxes for the IRB — it's about taking seriously the responsibility that comes with studying real people and their lives. The regulations exist because we know from hard experience what can go wrong when researchers don't think carefully about participant protection.
Honestly, this part trips people up more than it should.
If you're working with people — whether you're surveying them, interviewing them, observing them, or analyzing their data — the safe move is always to have that conversation with your institutional review board early. They'll help you figure out what's needed, and you'll move forward with confidence that your research is on solid ethical and regulatory ground.
That's worth the wait Most people skip this — try not to..
