If You Want the Public to Trust Science, Show Us You're Trustworthy

Created with ChatGPT based on my original writing and the themes of this essay.

There is rigor before a researcher is ever supposed to sit across from another human being and say, Tell me your story.

I know this because I have had to go through it.

I have completed human-subjects research training and maintained my CITI certifications. My dissertation methods have been examined by a faculty committee. I have had to explain how I will recruit participants, what I will ask them, how I will obtain consent, how their information will be stored, how I will protect their privacy, what risks participation might create, and what I will do if something does not go according to plan. Then I have had to submit that work to an Institutional Review Board and allow people outside my project to decide whether the access I am requesting to other people's lives is ethically justified.

The process can be frustrating. I have revised language that seemed obvious to me, documented decisions I had already thought carefully about, and waited for approvals when I was ready to move forward.

But going through it has changed the way I understand research authority.

We do not treat the right to conduct research with human beings as automatic. We make researchers demonstrate that they are prepared to hold that authority responsibly.

My intentions are not enough.

My belief in the importance of my research is not enough.

My training, by itself, is not enough.

Before another person is asked to trust me with their experiences, their information, their family, their health, their history, or some other part of their life, I am expected to account for what I intend to do with that access.

That is why the current federal conversation about revising human-subjects protections has landed differently for me than I expected.

The Department of Health and Human Services has included a proposed regulatory action in the 2026 Unified Agenda concerning exemptions and clarifying provisions related to Institutional Review Board oversight under the Common Rule. Among the stated goals are expanding exemptions for some low-risk research, allowing greater flexibility for de minimis protocol changes, and reducing administrative burden so that IRBs can focus more heavily on research posing greater risks.

I understand why that sounds appealing. I also think there is a difference between unnecessary bureaucracy and the productive friction of accountability.

Going through the IRB process has made that distinction much harder for me to ignore.

Some of the friction is the point.

It interrupts the researcher's momentum long enough to require questions that are easy to skip when you are excited about your study and convinced that what you are doing matters.

Do I actually need this information?

Could this question expose someone to a risk I have not considered?

Who will have access to what this person tells me?

What happens to these data after I am finished analyzing them?

What have I promised the participant?

What power do I have in this interaction that they do not?

And beneath all of those questions is one that I think every researcher should have to encounter at some point:

What gives me the right to ask another human being for this in the first place?

That question has stayed with me.

It is also why I keep thinking about human-subjects protections in relation to the much larger conversation happening around public trust in science.

Scientists, physicians, academics, researchers, engineers, and people across STEM keep asking some version of the same question: How do we get the public to trust us?

I think we are asking the question backward.

Trust is not something the public owes science because scientists are highly educated. A PhD does not create trust. An MD does not create trust. A university affiliation does not create trust. Neither does a peer-reviewed publication.

Those things can establish expertise. They can establish competence.

They do not automatically establish trustworthiness.

Trust requires social proof.

People need to be able to see what institutions do with power. They need to see what happens when our own interests come into tension with theirs. They need to see what we protect, what we are willing to constrain, and what we are willing to make harder for ourselves in order to reduce the possibility of harm to someone with less institutional power.

Human-subjects protections are part of that evidence.

The Common Rule did not appear because researchers collectively decided their work needed more paperwork. The modern federal system of human-subjects protections grew out of a history of serious ethical failures in biomedical and behavioral research. The National Research Act of 1974 was passed two years after the Tuskegee syphilis study became public. It created a national commission to examine the ethical principles that should govern research with human beings. The Belmont Report that followed articulated principles that remain foundational: respect for persons, beneficence, and justice.

That history does not mean regulations can never change.

It means we should remember why they exist when we change them.

When a safeguard becomes cumbersome, the first question cannot simply be whether we can remove it. We should also understand what problem it was built to address, whether that problem still exists, and what takes over its protective function if the safeguard disappears.

That becomes especially important when we start talking about research as "low risk."

Low risk to whom?

Risk is not experienced in the abstract. Something that appears inconsequential from inside a university may be deeply consequential inside someone's family or community. Information that looks like a variable in a dataset may represent disability, immigration status, medical history, trauma, financial insecurity, identity, stigma, or an experience a participant has trusted a researcher not to mishandle.

A question may look harmless until you understand the power relationship in which it is being asked.

A dataset may look anonymous until you understand the community represented inside it.

A research interaction can satisfy every procedural requirement and still leave someone feeling extracted from rather than engaged with.

Researchers can miss those things without being malicious.

That is precisely why independent accountability matters.

Good intentions are not a system of protection.

That feels even more important now because our ability to produce knowledge from people's lives is expanding rapidly.

Human-subjects research no longer looks only like a participant sitting across from a researcher in a laboratory or interview room. We analyze institutional records, electronic health information, digital behavior, online communities, administrative data, images, and enormous collections of text. We link datasets created for different purposes. Computational methods allow us to find patterns across thousands or millions of records. Artificial intelligence can make inferences from language and behavior at a scale that would have been impossible for an individual researcher to achieve.

Our ability to extract knowledge is becoming more powerful.

That does not create a greater entitlement to other people's lives.

If anything, it creates a greater obligation to think carefully about what responsible access means.

My own work has made this increasingly difficult for me to treat as theoretical. The people I ask to participate are parents of children with disabilities, and what I ask them to share is their child's special education record and what it got wrong. That is not a low-risk question for the person answering it.

When I submit an IRB protocol, I have to write down what I intend to do with another person's trust. I have to explain who will see what they tell me, how I will store it, how long I will keep it, what they will know before agreeing, what I will do to protect their identity, and why I need the information I am asking them to share.

There is something important about having to explain those decisions to someone who is not invested in getting my study into the field as quickly as I am.

The process forces a researcher to confront the fact that being capable of producing knowledge from another person's experience does not make that knowledge ours by default.

That is a principle worth protecting.

And it is part of why I think the scientific community should be paying much closer attention whenever protections are discussed primarily in terms of the burdens they create for researchers and institutions.

If a requirement consumes enormous amounts of time without meaningfully protecting participants, identify it. Study it. Change it. Build something better.

But efficiency is not an ethical argument by itself.

We also have to ask what becomes easier when oversight disappears. We have to ask whose risk may become less visible when researchers have more discretion. We have to ask who gets to determine that a category of research is sufficiently harmless to require less independent scrutiny.

And we have to be willing to ask those questions even when the answers might make our own work slower.

This is where the conversation about public trust becomes real for me.

The public is constantly being told that science is trustworthy because science has standards. There are methods. There is peer review. There are ethical requirements. There are systems intended to constrain misconduct, catch errors, disclose conflicts, and protect the human beings whose bodies, information, experiences, and communities make research possible.

Those systems are part of the evidence.

You cannot point to safeguards as evidence that science deserves trust and then treat those same safeguards as peripheral when they become inconvenient.

That does not mean every safeguard is effective. It means the burden is on us to know what we are removing and what replaces it.

Scientists often approach the public-trust problem as though the primary task is communication. Explain the evidence better. Correct misinformation. Improve scientific literacy. Teach people how the process works.

All of that matters.

But sometimes people understand institutions perfectly well.

They are watching what we do.

They are watching what happens when ethics and efficiency come into tension. They are watching whether protections survive when the people benefiting from the research are also the people inconvenienced by the protections. They are watching whether accountability remains important when it constrains our own power.

That is social proof.

And the more I conduct my own research, the more I understand why it matters.

The CITI certifications. The committee review. The protocols. The consent documents. The revisions. The questions from people who are not emotionally invested in getting my project completed on my timeline. The waiting.

None of those things prove that I am an ethical researcher. Compliance is not the same thing as care, and an IRB approval cannot anticipate every decision I will make once I am sitting with another human being.

But the process communicates something I now believe is essential:

I do not get to decide by myself that my intentions are sufficient.

When someone participates in my research, they are extending trust before I have produced the findings that may eventually justify the study. They are trusting that I will continue to see the person after their words become a transcript, a code, a variable, a quotation, or a finding. They are trusting me to remember that the information I am analyzing came from a life I was temporarily allowed to enter.

They do not owe me that trust because I am a researcher.

Which means some of the burden is supposed to belong to me.

That may be the part of this conversation that has affected me most.

We spend so much time asking how to persuade the public to give science more trust that we sometimes skip the prior question of what scientific institutions are doing to deserve it.

If scientists want the public to trust us, we should give people evidence.

Show them that when a safeguard is cumbersome, we care enough to understand what it protects before removing it.

Show them that when research participants have less institutional power than the people studying them, we recognize the significance of that imbalance.

Show them that our ethical standards mean something even when they slow us down.

Show them that the rigor required before a researcher can sit across from another human being and say, Tell me your story, is more than a set of boxes we complete in order to obtain permission to begin.

The public does not owe science its trust.

We have to give people reason to extend it.

And sometimes the clearest evidence of our trustworthiness is what we are willing to make harder for ourselves in order to protect someone else.