Ethics Is Not a Checkbox. It Is the Foundation.
By Hakan Karaaytu, AU Assistant Professor & UX Researcher
Every usability session, every interview, every survey begins with an unspoken agreement. A participant has agreed to give you their time, their candor, and a window into how they think and feel. They are not giving you data. They are extending trust. And everything that follows — how you conduct the session, how you store what you learn, how you represent their words in a report — either honors that trust or quietly erodes it.
This is what makes ethics in UX research something more than a procedural requirement. It is the relational foundation on which all credible research rests. And treating it as anything less — a consent form to file, a policy box to tick — misses the point entirely.
Compliance Is Necessary. It Is Not Sufficient.
There is an important distinction between compliance and ethical reasoning, and conflating them is one of the more consequential mistakes a research practice can make.
Compliance is the checkbox. It means meeting legal requirements, following company policy, collecting the right signatures. It is necessary — but it is a floor, not a ceiling. You can be fully compliant and still conduct research that fails the people who trusted you with their participation.
Ethical reasoning is something different. It is a sustained habit of asking not just am I allowed to do this? but should I do this? Not just does this meet our policy? but does this respect the person in front of me?
My journalism background made this distinction visceral early on. The Society of Professional Journalists' code of ethics rests on principles — seek truth, minimize harm, act independently, be accountable — that are not reducible to rule-following. They require ongoing judgment, applied differently in every situation. UX research operates under the same ethical logic. The methods differ. The underlying obligation does not.
Four Pillars of Ethical Research Practice
Over years of conducting and teaching UX research, I have come to organize ethical practice around four pillars. Each addresses a distinct dimension of the researcher-participant relationship.
Consent — true understanding, not just a signature. Informed consent requires that participants genuinely understand what they are agreeing to, in plain language rather than legal terminology. It requires that consent be ongoing rather than a one-time transaction at the beginning of a session — checking in as topics shift, as recording choices arise, as observations are made that will be shared with a wider team. And it requires that withdrawal be genuinely easy, with no social pressure, implicit or explicit, to continue.
Consent fatigue is a real phenomenon worth taking seriously. When a participant is asked to sign multiple forms, acknowledge recording permissions, and agree to data policies before the session even begins, the technical validity of their consent does not guarantee its quality. A fatigued, socially pressured agreement is not the foundation of trust that research requires.
Privacy — treating participant data the way you would want your own treated. The principle is straightforward. The practice requires discipline. Anonymize wherever possible, including in internal presentations. Use participant IDs rather than real names in notes and reports. If you are sharing screen recordings, blur any personal information visible on screen. Be specific and honest about where data lives, who has access to it, and what happens to it after the study ends.
A particular failure mode worth naming: overpromising anonymity in small studies. Telling participants they will be anonymous and then describing their role with enough specificity that anyone on the team can identify them is an ethical failure regardless of intent. The protection promised has to be the protection delivered.
Bias — the hidden force shaping research before the first question is asked. Bias is not usually malicious. It is usually structural, and it begins long before the research session itself. It lives in recruitment: if you draw participants exclusively from convenience — friends, colleagues, people already in your professional orbit — you are not studying users. You are studying access. And access is shaped by privilege and proximity in ways that systematically exclude the perspectives most likely to challenge your assumptions.
Bias lives in question design: asking "why do you prefer this approach?" assumes a preference that may not exist. Bias lives in analysis: foregrounding observations that confirm your hypothesis while minimizing contradictory evidence is confirmation bias, whether it is deliberate or not. Ethical research practice means actively working against each of these tendencies — recruiting deliberately, writing questions that genuinely invite unexpected answers, and approaching synthesis with the explicit goal of representing what participants actually communicated, not what was convenient to hear.
Emotional impact — responsibility for the research experience itself. Participants are not passive data sources. Participating in research is an experience, and we are responsible for what that experience feels like. This is especially important when research touches sensitive territory: financial stress, health, discrimination, personal failure. These topics require preparation — explicit permission to skip questions, genuine check-ins throughout the session, acknowledgment that the conversation may surface difficult feelings.
In a student project I supervised on financial stress among college students, the initial protocol had not accounted for how emotionally exposed participants might feel discussing money anxiety. We revised it: added clear language about the right to skip any question, included campus financial counseling resources in the debrief, and trained the student researchers to check in if the conversation became heavy. The substantive research did not change. The ethical quality of it changed significantly.
Lessons from Journalism That UX Research Needs
The parallels between journalism ethics and UX research ethics are close enough to be instructive rather than merely analogous.
Seek truth and report it translates directly: observe reality rather than assumptions. Do not design research to confirm a direction already chosen. Do not present findings selectively to make stakeholders comfortable. Just as a journalist verifies before publishing, a researcher must interrogate before concluding — ruling out alternative explanations, checking that patterns are real rather than artifacts of a small or skewed sample, representing the full complexity of what was learned rather than the tidiest narrative.
Minimize harm means thinking ahead about how participants will be affected by how their words and experiences are used. Before sharing a compelling quote, ask whether it could embarrass the participant, whether it could be traced back to them, whether the insight justifies the exposure. Usually the answer is yes — but the question must be asked.
Act independently means resisting pressure to skew findings in the direction stakeholders want. Your loyalty is to what you actually learned, not to any particular business outcome. That does not require confrontation, but it does require honesty — especially when the research surface something inconvenient.
Be accountable and transparent means naming the limitations of your work clearly. Small sample sizes. Demographic gaps. Preliminary findings that need validation. Transparency about what a study cannot tell you builds the kind of credibility that makes what it can tell you actually influential. Teams trust rigorous, honest researchers. They are appropriately skeptical of researchers who present findings without caveats.
Empathy Without Ethics Is Not Empathy
There is a version of user empathy that has become almost fashionable in design and research circles — a disposition of warmth and attunement that is invoked frequently but not always examined critically. The uncomfortable truth is that empathy without ethical grounding is not automatically benevolent. You can understand people deeply and use that understanding to manipulate them. You can know what users need and design in ways that exploit those needs rather than serve them.
This is not hypothetical. Dark patterns in interface design — mechanisms that use knowledge of user psychology to obscure costs, manufacture urgency, or make cancellation unnecessarily difficult — are a product of exactly this dynamic: understanding applied without ethical constraint.
Ethical research practice is what keeps empathy honest. It is the discipline that ensures understanding is directed toward serving users rather than exploiting them. It means designing sessions that respect participant agency. It means representing findings faithfully even when the truth is inconvenient. It means asking, at every stage of the research process, not only what will I learn from this? but how will my process make someone feel?
Will they feel respected? Will they feel heard? Will they walk away from this experience glad they participated — or will they feel, in some quiet way, that they gave more than they received?
Those questions do not have procedural answers. They require ongoing judgment, sustained attention, and a genuine commitment to the people whose trust makes research possible in the first place. That is what ethical practice actually looks like. Not a form, not a policy, not a phase of the research process — but a continuous orientation toward the humans at the center of the work.
Research is ultimately about relationships. And relationships are built on trust — or they are not built at all.
Hakan Karaaytu is an Assistant Professor and UX researcher working at the intersection of journalism, academic inquiry, and human-centered design. He can be reached at hakanhakan@hakankaraaytu.com
The ideas explored in this post were inspired by The Research Room , a podcast about how research shapes design, education, and everyday life.
Comments