The doll study: what it shows, and what people sometimes make it say

A research programme, not a single scene

Kenneth and Mamie Clark conducted several studies between the late 1930s and 1950 on racial awareness, identification and preferences among Black children. The doll procedure was part of this programme, but it does not sum up the whole body of work.

In the historical synthesis most often cited, around 253 Black children aged three to seven took part across different administrations. The questions did not all measure the same thing. Some asked for a preference, some associated a doll with positive or negative qualities, and others concerned racial identification and self-identification.

These distinctions matter. Choosing a white doll in response to one question is not equivalent to saying “I hate myself.” An answer may reflect learned social associations, understanding of the instructions, the testing context, the options imposed, or several mechanisms at once.

What the answers made visible

The Clarks’ work showed that young children were already perceiving the racial categories and hierarchies around them. In the samples studied, many assigned more positive qualities to white dolls and negative qualities to Black dolls. The moment when a child had to indicate which doll looked like them could make the tension especially visible.

The result had political force because it challenged the idea that segregation was merely an administrative arrangement with no effect on childhood. It showed that the social meanings attached to color entered children’s worlds early.

But the study cannot isolate segregation as the single cause. Children also encountered images, messages and social relationships beyond school. The procedure was not designed to measure “global self-esteem” using the psychometric tools that this expression implies today.

Brown v. Board: real influence, misleading shortcut

Kenneth Clark contributed to desegregation legal cases and testified in some of the cases brought together under Brown v. Board of Education. In its 1954 ruling, the Supreme Court concluded that separate schools are inherently unequal.

Footnote 11 of the decision cites a report by Kenneth Clark among several social-science references. That is enough to show that this work formed part of the intellectual and legal landscape of the case. It does not mean the Court decided “because of the doll test.” The ruling rested first on constitutional analysis and a broader set of arguments.

Reducing Brown to a psychological experiment also erases the legal, activist and collective work that made the case possible. The fuller story is no less striking: research conducted by two Black psychologists helped make visible what an institution claimed it could separate without causing harm.

A historic study complicated by later findings

In 1970, Joseph Hraba and Geoffrey Grant used a similar procedure with Black and white children. In their sample, a majority of Black children preferred the Black doll. This result does not mechanically “disprove” the Clarks. It shows that observed preferences are not a natural or immutable property of Black children. They change with historical moment, social context, Black pride movements and the way a study is conducted.

Other researchers have questioned the validity of the procedure and the way its findings were transformed into a story about low self-esteem. These critiques reinforce an essential lesson: the experiment reveals associations within a constructed situation. It does not provide an individual diagnosis.

What this changes for adults today

It may be tempting to reproduce the test at home to “see what a child thinks.” That is not a good idea. Without a protocol, context or support, a child’s choices can easily be overinterpreted. A preferred doll today does not sum up either their identity or their psychological health.

You can, however, observe the environment. Which children are described as beautiful, intelligent, gentle, brave or desirable in stories? Who occupies the center of the picture? Do Black characters have multiple personalities, families, skin tones and adventures, or do they always carry the same lesson?

When a child expresses a hierarchical preference, avoid shaming them. Ask: “What makes you say that?” Then offer a clear counterpoint: no skin color makes a person kinder, more beautiful or more worthy. The aim is not to make the child pass a moral test, but to help them notice a learned association and build other ones.

What the research does not say

The procedure does not prove that a child who chooses a white doll rejects themselves. It does not directly measure the effect of a book, cartoon or family conversation. Nor does it allow us to claim that a historical percentage applies to children in France today.

Its value lies elsewhere: it reminds us that children notice social rankings early and that their preferences do not emerge in a vacuum.