PsychOS
The Blue Eyes, Brown Eyes Experiment: The 1968 Classroom Exercise That Turned Third Graders Into Bullies Overnight
A teacher told her all-white third-grade class that blue-eyed kids were smarter than brown-eyed kids. By lunch, best friends were tormenting each other, grades were dropping, and an entire room of...
Explore PsychOS
PsychOS combines psychology education with structured self-reflection tools. Explore assessments, articles, experiments, and related resources while treating results as educational signals rather than diagnosis.
The morning after Martin Luther King Jr. was shot dead in Memphis, a teacher in Riceville, Iowa walked into her third-grade classroom and lied to twenty-eight children. She told them blue-eyed kids were smarter, cleaner, and more civilized than brown-eyed kids. She gave the blue-eyed children extra recess, first place in the lunch line, and open praise. She made the brown-eyed children wear paper collars so everyone could spot them at a glance, and told the class that people with brown eyes just weren't as good.
By eleven a.m., children who had been inseparable the day before were calling each other names on the playground. By that afternoon, the brown-eyed kids, now branded inferior, were failing math problems they'd solved easily twenty-four hours earlier. Nothing about any child's intelligence had changed. A teacher had simply told them a lie, and their entire social world rearranged itself around it in under a day.
THE HYPOTHESIS
Jane Elliott wasn't running a university study. She was a schoolteacher trying to answer a question her students couldn't articulate: what does discrimination actually feel like when it's happening to you? Her working idea was direct — if she built a fake hierarchy inside her own classroom and put every child on both sides of it, they would understand prejudice in a way no lecture ever could.
This is worth stating plainly and up front: Elliott's exercise was not a controlled scientific experiment. It had no control group, no randomization, and none of the safeguards that let researchers rule out other explanations. What makes it valuable isn't methodological rigor. It's how vividly it puts a face on a mechanism that more controlled research had already begun mapping in the lab, and how many separate, well-documented psychological patterns it manages to touch at once.
THE METHODOLOGY
Elliott split her twenty-eight students into two groups by eye color. On day one, blue-eyed children were declared superior — more recess, first in line, praised more freely for right answers. Brown-eyed children wore collars, sat at the back, were denied direct access to the water fountain, and had their mistakes singled out. The next day, Elliott reversed it entirely, telling the class she'd been wrong and that brown eyes were actually the superior trait, flipping every privilege and restriction in the opposite direction.
She tracked academic performance, playground behavior, and the language kids used toward each other across both days, and repeated modified versions of the exercise with later classes for years afterward. Decades later, several of her original 1968 students were interviewed on camera as adults, giving at least an informal, longitudinal thread to compare against her original classroom notes.
STEP-BY-STEP EXECUTION
The shift was fast and ugly. On the day blue-eyed children were told they were better, they grew visibly more confident and outperformed their own prior scores on timed tasks, while also turning openly cruel toward brown-eyed classmates who had been friends the day before. Brown-eyed children, told they were lesser, grew withdrawn, made more mistakes on identical material, and in later interviews described feeling "stupid" or "bad" during the exercise itself.
When Elliott flipped the roles the next day, the cruelty flipped with it. Kids who had mocked their classmates twenty-four hours earlier now found themselves on the receiving end of it, and several later described that specific discomfort, recognizing their own behavior turned back on them, as the moment the lesson actually landed.
There's no statistical control group here, and Elliott's documentation, however careful, is one teacher's observation rather than blinded, independent measurement. What strengthens it is repetition: she ran the exercise again with new classes for years and later with adult groups, reporting the same pattern each time.
PSYCHOLOGICAL REVELATION
Elliott's exercise showed how fast a false label from a trusted authority figure can rewire how children treat each other and how they perform academically, using nothing but words and a paper collar. Eye color predicts nothing about ability. The label alone did the damage — a mechanism closely related to In-Group vs. Out-Group Bias Explained , where arbitrary categories produce real favoritism with no factual basis behind them at all.
The instant boost in confidence and performance among the "superior" blue-eyed children mirrors a mechanism studied under far more controlled conditions in The Growth Mindset Study , where Carol Dweck found that praising a child's intelligence, rather than their effort, could measurably change how they responded to challenge afterward. Elliott's exercise suggests the same sensitivity to labeling runs in the opposite direction too, since a label of inferiority produced a documented performance drop within hours. That collapse in the brown-eyed children's performance also echoes Learned Helplessness Explained , where repeated signals that effort won't be recognized fairly tend to reduce effort shortly after, independent of any real change in ability.
The speed of the switch resembles obedience research more broadly. In The Milgram Obedience Study , ordinary adults complied with instructions from an authority figure that conflicted with their own sense of right and wrong, largely because the instruction came from someone they'd been primed to trust. Elliott's students didn't need much convincing either. A single trusted adult reframed the social order of the room, and the room reorganized itself accordingly, a dynamic explored more broadly in Conformity Explained and demonstrated experimentally in The Asch Conformity Study , both showing that individuals will bend their own stated judgment to match group pressure even when that judgment is visibly wrong.
A related pattern of assigned identity overriding individual disposition appears in The Stanford Prison Experiment , where students randomly assigned to be "guards" or "prisoners" began behaving in ways consistent with those roles within days, despite no real difference between the groups beyond the label they'd been given. The way an entire classroom's social dynamic bent around a single authority figure's framing also has a family resemblance to Groupthink Explained , where the pressure to preserve group cohesion can override individual judgment even among capable, well-intentioned people.
There's also a domestic version of this exact dynamic already documented outside the classroom. The Golden Child and Scapegoat describes how, within a single family, one child can be elevated and another blamed based on a parent's arbitrary framing rather than anything the children actually did, producing strikingly similar patterns of confidence and withdrawal to what Elliott documented in her classroom. And the underlying cognitive shortcut, letting one trait quietly color judgment of everything else about a person, is the same mechanism explored in The Halo Effect Explained , where a single positive or negative impression, however unrelated to competence, ends up shaping how a person's abilities are perceived across the board.
See Your Own Response to Assigned Labels More Clearly
Reading about the blue eyes, brown eyes experiment is different from noticing how much your own confidence shifts, in real time, when you're told you belong to a favored or disfavored group, and whether you've ever caught yourself living up, or down, to a label you never actually chose.
→ Take the Self Awareness Test
Free. No account required. Maps how much your self-perception bends under external labels and social feedback.
Curious how group identity and social belonging shape your broader behavior patterns?
→ Explore the full PsychOS Assessment Lab
Frequently Asked Questions
What was the blue eyes, brown eyes experiment?
A 1968 classroom exercise by teacher Jane Elliott, conducted the day after Martin Luther King Jr.'s assassination, in which she told her third-grade class one eye-color group was superior, then reversed it the next day to demonstrate how prejudice forms and feels from the inside.
Was it a real scientific study?
No, not in the controlled sense. It lacked a control group and randomization, unlike formal minimal group experiments from the same period. It's best understood as a documented, repeated classroom demonstration rather than a peer-reviewed controlled study.
How did the kids' behavior actually change?
The "superior" group grew more confident and performed better academically while acting cruelly toward the other group. The "inferior" group grew withdrawn and made more errors on identical tasks. The pattern reversed almost completely when Elliott flipped the assigned groups the next day.
How does this compare to the Stanford Prison Experiment?
Both show how quickly an assigned label or role, with no real basis behind it, can change behavior. The Stanford Prison Experiment used randomly assigned "guard" and "prisoner" roles among college students; Elliott's exercise used eye color among eight-year-olds, with a trusted teacher actively reinforcing the hierarchy.
Did the effects last beyond the classroom?
Several of Elliott's original students, interviewed decades later as adults, described the exercise as a lasting influence on how they understood prejudice, though this comes from personal recollection rather than a formal longitudinal study.