The Clark Doll Experiment: Race, Identity, and the Case That Ended School Segregation

Four dolls. Two white, two brown. A child between three and seven years old seated across a table from a researcher. The researcher asks, quietly and without expression: "Which doll is the nice doll?" "Which doll looks bad?" "Which doll looks like you?" Kenneth and Mamie Clark ran this procedure hundreds of times in the early 1940s, in churches, schools, and community centers across the American South and North. What they found — that a majority of Black children assigned goodness to whiteness and badness to their own skin — became scientific evidence in the most important civil rights case of the twentieth century.

Who Were Kenneth and Mamie Clark?

Kenneth Bancroft Clark and Mamie Phipps Clark were both products of the Black intellectual community that flourished despite the systemic exclusion of the Jim Crow era. Mamie Phipps grew up in Hot Springs, Arkansas, and enrolled at Howard University in Washington, D.C. — one of the few institutions where Black students could pursue advanced academic work at the time. She met Kenneth Clark there.

Mamie Clark earned her master's degree from Howard in 1939 and her doctorate in psychology from Columbia University in 1943, one of the first African-American women to do so. Kenneth Clark received his PhD from Columbia in 1940.

Source: Smithsonian Magazine (2018), "How a Psychologist's Work on Race Identity Helped Overturn School Segregation" — biographical profile of Mamie Phipps Clark and the genesis of the doll studies.

The doll experiment grew directly from Mamie's master's thesis at Howard, which examined how Black children develop racial identification. The core question was deceptively simple: at what age do Black children recognize their own race, and what feelings do they attach to that recognition?

Kenneth and Mamie Clark expanded her initial work into a full research program. They published their central findings in two key papers: one in 1939 (in the Journal of Social Psychology) and a more comprehensive study in 1950 in a volume titled Readings in Social Psychology, edited by Theodore Newcomb and Eugene Hartley.

Source: Clark, K.B. & Clark, M.P. (1950), "Emotional factors in racial identification and preference in Negro children," in Newcomb & Hartley (Eds.), Readings in Social Psychology — the paper cited directly in the Brown v. Board footnote.

The Design: Four Dolls and Eight Questions

The materials were deliberately minimal. The Clarks used four commercially available dolls — two with light (white) skin and blond hair, two with brown (dark) skin and black hair. Every other feature was identical. The dolls were presented to children one at a time, and the researcher read from a scripted sequence of eight requests:

Procedure (the eight requests, in order):

1. Show me the doll that you like to play with — or the doll you like best.
2. Show me the doll that is the nice doll.
3. Show me the doll that looks bad.
4. Show me the doll that is a nice color.
5. Show me the doll that looks like a white child.
6. Show me the doll that looks like a colored child.
7. Show me the doll that looks like a Negro child.
8. Show me the doll that looks like you.

Requests 1 through 4 measured preference and value attribution. Requests 5 through 7 measured racial classification ability. Request 8 — "show me the doll that looks like you" — was the critical moment. It confronted the child directly with self-identification after they had just, in most cases, assigned that same doll negative attributes.

Source: NAACP Legal Defense Fund, "Brown v. Board: The Significance of the 'Doll Test'" — summary of the experimental protocol and the eight-question sequence.

Misattribution — the core psychological concept at stake here — refers to the process of assigning one's own feelings or characteristics to an external target. The Clarks were not merely documenting preference; they were measuring whether children had internalized the racial hierarchy that surrounded them, attributing negative qualities to their own racial group.

The Sample: North and South

The Clarks deliberately tested children in both segregated and non-segregated settings to determine whether the source of psychological damage was segregation specifically or anti-Black racism more broadly. Their combined sample across studies included approximately 253 Black children between ages three and seven, tested in segregated schools in Arkansas and in (nominally) integrated schools in Massachusetts.

Study Location	School Type	N (approx.)	Age Range
Segregated Arkansas schools	Fully segregated	~160	3–7 years
Springfield, Massachusetts	Mixed-race enrollment	~93	3–7 years
Combined data	Mixed	~253	3–7 years

Source: Clark, K.B. & Clark, M.P. (1939), "The development of consciousness of self and the emergence of racial identification in Negro preschool children," Journal of Social Psychology, 10(4), 591–599 — the initial study establishing baseline data.

The geographic comparison yielded a counterintuitive result that the Clarks found troubling: children in segregated Southern schools showed somewhat less pronounced rejection of the brown doll than children in the mixed-race Northern schools. The Clarks' interpretation was not that segregation was benign but rather that children in segregated environments had less direct, daily exposure to the white ideal against which they were being measured — while Northern Black children, surrounded by white peers and white social standards, experienced the gap more acutely.

The Results: What the Children Said

Across the combined sample:

• 67% of children chose the white doll as the one they liked best.
• 59% identified the white doll as the "nice" doll.
• 59% said the brown doll "looks bad."
• 60% said the white doll was a "nice color."

When it came to self-identification (request 8), the majority of children correctly identified the brown doll as "the doll that looks like me" — demonstrating that they could accurately classify their own race. But many of those same children had just labeled that doll as bad, unpleasant, or undesirable.

Source: Clark, K.B. & Clark, M.P. (1950), Readings in Social Psychology — the consolidated data table including preference, attribute, and self-identification responses.

The collision between self-knowledge and self-rejection was what made the data so stark. A child pointing to the brown doll and saying "that's the bad one" — then pointing to the same brown doll and saying "that looks like me" — was not confused. They knew who they were. They had simply absorbed their culture's judgment of who they were.

Some children, when asked request 8 after having already called the brown doll "bad," became visibly distressed. Some pushed the brown doll away. Some cried. One child, the Clarks reported, pointed at the brown doll and said "that's a nigger, I'm a nigger." Another responded to request 8 by laughing — a reaction the Clarks identified as masking acute discomfort.

Source: History.com (2020), "How Dolls Helped Win Brown v. Board of Education" — narrative account of recorded child responses during testing.

Why This Happened: The Mechanism the Clarks Proposed

The Clarks did not argue that Black children were born with lower self-esteem. Their argument was structural and sociological: children internalize the values that their environment consistently signals to them.

Racial self-concept — a child's understanding of their own racial identity and what that identity means in social terms — develops between ages three and five, roughly in parallel with other identity formation. By age five, most children can correctly sort themselves and others by race. What varies is the valence — positive or negative — that the child attaches to those categories.

In 1940s America, the signals were unambiguous. Segregation itself communicated hierarchy: the better-resourced schools, swimming pools, restaurants, and neighborhoods were not available to Black children. The white doll, in this reading, was not an abstraction — it was a symbol children had been trained by their social environment to interpret as superior.

Think of it like a child who grows up hearing a specific food called "poor food" and a different food called "fancy food." Even without being explicitly told the hierarchy, they will absorb it. The Clarks' contribution was to make that absorption measurable and visible.

Source: NAACP Legal Defense Fund, "Brown v. Board: The Significance of the 'Doll Test'" — contextual analysis of the mechanism underlying the Clark findings.

Kenneth Clark wrote, in his 1955 book Prejudice and Your Child, that the damage was not simply a matter of hurt feelings. The sustained, systemic signal that one's racial group is inferior produces a distortion in the development of identity — a self-concept built partly on a foundation of internalized denigration. He called this a form of psychological injury as real and as measurable as a physical wound.

The Path to the Supreme Court

The Clarks were not initially working with legal strategy in mind. Their goal was to document psychological harm for its own sake — to generate scientific evidence of what segregation did to children's minds. The legal application came later, through a chain of connections.

Thurgood Marshall, who would later become the first Black Supreme Court Justice, was then the chief counsel for the NAACP Legal Defense Fund. Marshall was building a litigation strategy to challenge school segregation directly on constitutional grounds, and he needed social science evidence to demonstrate that "separate but equal" — the doctrine established in Plessy v. Ferguson (1896) — was not psychologically equal.

Kenneth Clark testified as an expert witness in Briggs v. Elliott (1952), the South Carolina case that became one of the five cases consolidated into Brown v. Board of Education. His testimony described the doll experiment and its results. He reported that 16 of the 23 Black children he tested in Clarendon County, South Carolina, had exhibited the same preference for the white doll.

Source: Cornell Undergraduate Law & Society Review, "Outside the Dollhouse and Into the Court: The Importance of the 'Doll Test' in Brown and Beyond" — detailed account of Clark's testimony in Briggs v. Elliott and the consolidation into Brown.

Case	State	Clark's Role
Briggs v. Elliott (1952)	South Carolina	Expert witness, direct testimony
Davis v. County School Board (1952)	Virginia	Written findings submitted
Brown v. Board of Education (1952)	Kansas	Research cited in brief
Bolling v. Sharpe (1952)	Washington, D.C.	Social science appendix
Gebhart v. Belton (1952)	Delaware	Supporting research cited

The five cases were argued together before the Supreme Court. The Court's unanimous 1954 decision, written by Chief Justice Earl Warren, contained a now-famous footnote — footnote 11 — which cited social science research including Clark's 1950 paper. The operative text of the decision stated that segregation "generates a feeling of inferiority as to [Black children's] status in the community that may affect their hearts and minds in a way unlikely ever to be undone."

Source: Brown v. Board of Education, 347 U.S. 483 (1954), footnote 11 — the Supreme Court's citation of the Clark research alongside six other social science sources.

Critiques and Complications

The doll experiment was influential, but it was not without critics — then and now. Three lines of criticism matter.

First, the interpretive question. Does choosing the white doll as "nice" necessarily indicate damaged self-esteem? Some researchers, including sociologist Robin Williams in reviews of the period, argued that preferring an object associated with dominant culture is not the same as rejecting the self. A Black child preferring the white doll might be expressing cultural literacy — knowing what the dominant group values — rather than self-hatred.

Second, the replication problem. Psychologist Doris Entwisle and others observed that the Clarks' sample sizes were modest and the experimental controls limited by the standards of later social science. The emotional reactions reported (children crying, pushing the dolls away) were observational notes rather than coded behavioral outcomes, and observer bias could have influenced their interpretation.

Source: Michigan Psychological Association (2022), "Almost 70 Years after Brown v. Board of Education, How Do Black Children View Themselves?" — review of replications and critiques of the original Clark methodology.

Third, the persistence question. In 2006, filmmaker Kiri Davis produced a short documentary, A Girl Like Me, in which she replicated the doll test with fifteen contemporary Black children in New York City. Eleven of the fifteen chose the white doll as the "nice" or preferred doll — a result nearly identical to the 1940s data. Whether this represents ongoing damage from structural racism or a different psychological process entirely remains debated.

Source: Davis, K. (2006), A Girl Like Me (documentary short) — informal replication of the doll test with contemporary Black youth.

A fourth critique came from a different direction: the legal use of the evidence. Some constitutional scholars argued that the Supreme Court should not have based its ruling on social science research at all, that Plessy should have been overturned on straightforward equal protection grounds. If the social science were later challenged or undermined, this argument went, the foundation of the ruling would be weakened. Chief Justice Warren was apparently aware of this concern and chose to place the Clark citation in a footnote rather than the body of the decision.

The Clarks' Later Work and Northside Center

The doll experiment is the best-known part of the Clarks' work, but their careers extended far beyond it. In 1946 — eight years before Brown — Kenneth and Mamie Clark founded the Northside Center for Child Development in Harlem. The Center provided psychological and educational services to Black children in New York City at a time when such services were effectively unavailable to them through public institutions.

Source: Howard University's The Dig, "Mamie Phipps Clark: The Pioneering Psychologist Behind the Famed 'Dolls Test'" — account of the Northside Center's founding and work.

Mamie Clark ran Northside Center as its executive director until 1979. The center addressed learning disabilities, emotional disorders, and the effects of poverty and racism on child development — the practical application of everything the doll studies had documented theoretically. Kenneth Clark, meanwhile, became the first Black president of the American Psychological Association in 1970.

The center still operates today. It has served tens of thousands of children and families in Harlem over eight decades.

Self-fulfilling prophecy — the sociological concept describing how a belief about an outcome can produce that very outcome through the behavior it induces — is the mechanism the Clarks spent their careers trying to interrupt. If Black children were taught, implicitly, that they were less capable, less worthy, or less beautiful, that belief would shape their behavior, their aspirations, and their outcomes. Breaking the mechanism required both social change (desegregation) and direct psychological intervention (the Center's services).

What the Experiment Got Permanently Right

Whatever the methodological limitations, the Clark doll experiment captured something that subsequent decades of research in developmental psychology have consistently confirmed: children absorb social hierarchies early, before they can articulate them, and those hierarchies shape self-concept in ways that have measurable consequences.

Research on implicit bias — beginning with the Implicit Association Test developed by Greenwald, McGhee, and Schwartz in 1998 — has repeatedly shown that both Black and white children and adults show faster automatic associations between white and positive concepts than between Black and positive concepts. The underlying psychological mechanism is different from the Clarks' methodology, but the pattern is the same.

Source: Greenwald, A.G., McGhee, D.E., & Schwartz, J.L.K. (1998), "Measuring individual differences in implicit cognition: The Implicit Association Test," Journal of Personality and Social Psychology, 74(6), 1464–1480 — foundational IAT research extending and contextualizing the Clark findings.

The Clarks were measuring, with four dolls and eight questions, the same thing that modern researchers measure with millisecond-level reaction times. They were measuring what a culture has taught its children to believe about themselves.

The experiment's legacy is specific: it demonstrated, in a format accessible to judges, journalists, and the general public, that psychological harm from segregation was not theoretical. It was visible in the choices of three-year-olds. And once visible, it was very difficult to dismiss.

---

Q: What exactly did the Clark doll experiment find?
A: Between 59% and 67% of Black children aged three to seven preferred the white doll, called it "nice," and identified the brown doll as "bad" — while also correctly identifying the brown doll as the one that "looks like me." The data showed simultaneous accurate racial self-identification and negative valuation of their own racial group.

Q: How did the doll experiment reach the Supreme Court?
A: Thurgood Marshall, building the NAACP's legal challenge to segregation, recruited Kenneth Clark as an expert witness. Clark testified in Briggs v. Elliott (1952), one of five cases consolidated into Brown v. Board of Education. The Supreme Court's 1954 decision cited Clark's 1950 paper in footnote 11.

Q: Were the Clark findings replicated?
A: Subsequent studies have found mixed results. Filmmaker Kiri Davis's 2006 informal replication found similar patterns. Controlled academic replications have produced less consistent findings, partly due to methodological variations. Modern implicit bias research, using the IAT, consistently finds analogous patterns of automatic race-positive associations favoring white stimuli.

Q: Did the Clark experiment prove that segregation caused psychological damage?
A: The study demonstrated correlation: Black children in a segregated society showed systematic devaluation of their own racial group. Whether the direct causal agent was segregation specifically or broader anti-Black racism in American culture was debated at the time and after. The Clarks argued that segregation was the vehicle for that racism's transmission to children.

Q: What did Mamie Clark do after the doll experiment?
A: Mamie Clark founded the Northside Center for Child Development in Harlem in 1946 with Kenneth Clark, providing psychological services to Black children largely unavailable elsewhere. She ran the center as executive director until 1979. She is now recognized as one of the most influential applied psychologists of the twentieth century.

Q: Why do some critics dispute the experiment's use in Brown v. Board?
A: Some constitutional scholars argued the ruling should have rested on equal protection grounds alone, independent of social science evidence. If the research were later weakened by methodological critique, they feared the legal foundation would be vulnerable. Chief Justice Warren apparently shared some of this concern, placing the Clark citation in a footnote rather than the core reasoning of the opinion.

---

The Takeaway

Kenneth and Mamie Clark built an experiment simple enough for a three-year-old to participate in and powerful enough to reach the Supreme Court. Their finding — that children raised inside a racist social structure absorb that structure's judgments about their own worth — was not a comfortable truth. It was a measurable one. The doll test worked not because four plastic toys revealed some deep psychological secret, but because they gave visible form to what Black children were already experiencing: a world that consistently told them their skin color made them less. The experiment's enduring contribution is not methodological precision. It is that it made invisible damage impossible to ignore.

---

Read next:

• The Bobo Doll Experiment: How Children Learn Aggression Through Observation
• The Milgram Experiment: Obedience, Authority, and the Banality of Harm
• The Asch Conformity Experiment: Why We Deny What We See