Jane Elliott’s blue eyes/brown eyes exercise, first conducted in 1968 with third-graders in Riceville, Iowa, is widely considered unethical because it deliberately inflicted psychological distress on children through deception, public humiliation, and manufactured hostility, all without the safeguards that would normally protect human subjects in any research or educational setting.
How the Exercise Worked
The day after Martin Luther King Jr. was assassinated, Elliott asked her class if they wanted to try an exercise about discrimination. The students agreed. She then divided them by eye color: blue-eyed children were declared the superior group. They received extra food at lunch, more recess time, and seats at the front of the classroom. Brown-eyed children wore fabric collars to mark them as the inferior group, sat in the back, and were reprimanded more harshly for mistakes.
Elliott went further than simply creating unequal conditions. She deliberately lied to the children, telling them that the melanin responsible for blue eyes also produced higher intelligence and learning ability. She encouraged blue-eyed students to play only with each other and to ignore their brown-eyed classmates. The next day, she reversed the roles. The exercise was later repeated with new classes for years and eventually adapted for adult diversity training workshops.
Deception Used on Children
The most fundamental ethical problem is that Elliott knowingly told children a scientific falsehood to manipulate their behavior. She presented a fabricated link between eye color and intelligence as fact, leveraging her authority as their teacher. Third-graders lack the developmental capacity to critically evaluate a claim like that from an adult they trust. The brown-eyed students initially pushed back on the idea that blue-eyed children were better, but Elliott’s insistence, backed by her position of authority, overcame their resistance.
In any formal research context, deceiving participants requires strict justification: the deception must be necessary, the potential harm must be minimal, and participants must be fully debriefed afterward. Elliott’s subjects were eight- and nine-year-olds who could not meaningfully consent to being lied to, and whose emotional responses were not something they could simply reason through after the fact.
Psychological Harm to Participants
The exercise reliably produced significant emotional distress. In adult versions of the training, researchers found that participants experienced high levels of stress, with some breaking down in tears and at least one quitting the exercise mid-session. Facilitators themselves expressed concern about the intensity of negative emotions the exercise provoked. For children, the effects could be worse, precisely because they had fewer tools to process what was happening to them.
One participant who went through a version of the exercise in fifth grade around 1971 described the experience decades later in stark terms. “The experience was degrading and humiliating,” they wrote to PBS. At the end of the day, when the brown-eyed students were asked for their reactions, some expressed anger while others burst into tears. The reaction was so intense that the teacher halted the experiment early, meaning the brown-eyed children never got the reversed experience of being the “superior” group. That participant reported carrying feelings of humiliation and unresolved hostility toward both the teacher and the blue-eyed students long into adulthood, calling it “an experiment gone wrong.”
This points to a core problem: the exercise could inflict real harm, and there was no reliable mechanism to repair it. Reversing the roles the next day was supposed to provide balance, but not every replication followed that structure, and even when it did, the emotional damage from day one didn’t simply cancel out.
No Meaningful Consent
Elliott did ask her students if they wanted to participate, but asking eight-year-olds in a classroom whether they’d like to try something their teacher is proposing is not informed consent in any meaningful sense. The children had no way to understand what they were agreeing to. They didn’t know they would be lied to, publicly singled out, or subjected to peer hostility orchestrated by their own teacher. Parental consent was not obtained in advance. The power imbalance between a teacher and her students made genuine refusal nearly impossible.
Emotion Without Reflection
Academic critics have raised a structural concern about how the exercise handles emotion. As researchers have noted, Elliott’s approach sets up raw emotional experience in a way that is largely inaccessible to rationality, reflection, or criticism. The distress becomes the lesson. But distress without careful processing can reinforce negative feelings rather than produce empathy or understanding.
The participant who described their fifth-grade experience captured this precisely: “I don’t recall any discussion of what other people might feel like if we discriminate against them. What I remember is the humiliation of that day and the hatred/vengeance that I felt toward the teacher and the blue-eyes, which was never redirected or satisfied.” The intended lesson about discrimination never landed. What stuck was the pain.
Researchers evaluating the exercise have questioned whether a prejudice-reduction approach built on confrontation and guilt is genuinely useful, particularly when numerous alternative diversity-training methods exist that use positive messages rather than induced suffering. Even scholars sympathetic to experiential learning about sensitive topics have acknowledged that there is a line separating justifiable discomfort in education from harm, and that Elliott’s exercise sits on the wrong side of it.
The Teacher as Both Authority and Antagonist
A less discussed but significant ethical issue is the dual role Elliott played. She was simultaneously the person inflicting the distress and the person responsible for the children’s wellbeing. In any ethical research design, there are safeguards: independent oversight, the ability to withdraw without consequence, someone outside the experiment monitoring for harm. In Elliott’s classroom, she was all of those roles at once, and she had a vested interest in the exercise producing dramatic results.
Children in the “inferior” group couldn’t appeal to a neutral adult. Their teacher was the one telling them they were less intelligent. Their classmates, following the teacher’s instructions, were the ones enforcing the social exclusion. For a child in that position, there was no safe exit.
Why It Still Sparks Debate
Elliott’s exercise occupies an uncomfortable space because many former participants, including some from the original 1968 class, have spoken positively about the experience as adults. A 1985 PBS Frontline reunion showed several of them crediting the lesson with making them more empathetic. This is part of why the exercise has persisted for decades and why Elliott continued conducting it with adults well into her later career.
But positive outcomes for some participants don’t resolve the ethical problems. The fact that an intervention sometimes produces good results doesn’t justify the harm it inflicts on others, especially when those others are children who cannot consent. The ethical standard isn’t whether the exercise worked for the majority. It’s whether it was acceptable to subject any child to deliberate humiliation, deception, and peer hostility in a setting where they had no real choice and no independent protection.