Before the study

In a study published in 1961, three researchers at Stanford University in California let young children watch an adult attack a large inflatable clown called a Bobo doll. Later, in another room with the same kind of doll and the adult gone, many of the children attacked it in the same way, using the same words. The study became one of the best-known experiments in psychology. This page explains what you need to know before you read about it.

The people who did it

Albert Bandura was born in 1925 in Mundare, a small farming town in Alberta, Canada, to a Polish father and a Ukrainian mother. He trained in clinical psychology at the University of Iowa, where he learned the learning theories of the time, and joined Stanford in 1953. He was in his mid-thirties when this study was done. His co-authors were the sisters Dorothea Ross and Sheila A. Ross. Dorothea Ross later became a well-known researcher in children's health psychology.

The study is often called "Bandura's Bobo doll study", but the paper has three authors. Cite all three: Bandura, Ross and Ross (1961). The paper does not say which of them ran the sessions or who the two adult models were.

How psychologists explained learning in 1961

In the 1950s the strongest school in American psychology was behaviourism. Behaviourists study what people and animals do, not what they think. They explain learning through the consequences of behaviour.

The key idea is reinforcement: a consequence that makes a behaviour more likely to happen again. If a child tidies her toys and her father praises her, the praise reinforces tidying. Punishment is a consequence that makes a behaviour less likely. The psychologist B. F. Skinner built a whole account of learning on these two ideas.

This account works well for many things. But it has a problem. Children pick up a great deal that nobody ever rewarded them for: an accent, a way of walking, a swear word they heard once. How?

In 1941 two Yale psychologists, Neal Miller and John Dollard, offered an answer in a book called Social Learning and Imitation. They argued that children learn to imitate because imitating has been rewarded in the past. A child copies an older brother, gets the sweet, and so copying itself becomes a habit. In their account, imitation still needs reinforcement.

Bandura's objection was that in Miller and Dollard's experiments the children only chose between acts they could already do. The harder question was whether watching could teach a child a new pattern of behaviour. He suspected that simply watching another person could be enough, even when nobody is rewarded and the child does not practise at the time. Learning by watching is called observational learning. The person being watched is called the model, and copying what the model did is imitation.

A second idea: identification

A different explanation came from psychoanalysis, the school founded by Sigmund Freud. Psychoanalytic writers said that young children identify with important adults. Identification means taking on another person's qualities as part of yourself, usually because of a strong emotional bond or a strong emotional threat.

The Bobo paper names two versions of this idea. Anna Freud, Sigmund Freud's daughter, described identification with the aggressor: a frightened child copes by becoming like the powerful person who threatens them. The American learning theorist O. H. Mowrer described a similar process, which he called defensive identification. In 1961 these ideas were "widely accepted as an explanation of the imitative learning of aggression" (Bandura et al., 1961, p. 581).

Bandura wanted to show that a simpler explanation was enough. If children copied an adult they had no relationship with, an adult who neither loved nor threatened them, then a special emotional process would not be needed. The models in this study were, in the paper's words, "more or less neutral figures" (p. 582).

Where Robert Sears fits

Many textbooks link this study to Robert Sears, who headed Stanford's psychology department from 1953 to 1961 and tried to combine psychoanalytic ideas with learning theory. Sears is not cited in the 1961 paper. The developmental psychologist Joan Grusec (1992), who had worked with Bandura, describes Bandura's work as growing out of Sears's project, while dropping its psychoanalytic and drive-based parts.

What psychologists meant by aggression

Aggression in psychology means behaviour aimed at hurting or damaging someone or something. In 1961 there were three main ideas about where it comes from.

1. Aggression as a drive. Freud thought humans have an inner aggressive energy that builds up and must come out. One version of this, the catharsis idea, says that letting anger out in harmless ways, such as hitting a pillow, reduces it.
2. Aggression as a reaction to frustration. In 1939 Dollard, Miller and colleagues at Yale proposed the frustration-aggression hypothesis: being blocked from a goal produces a push towards aggression.
3. Aggression as learned behaviour. Behaviourists and social learning theorists argued that people learn when, how and against whom to be aggressive, just as they learn other behaviour.

These ideas lead to different predictions. If catharsis is right, a child who watches violence, or hits a doll, might feel calmer afterwards. If aggression is learned, watching violence should teach the child new ways to be aggressive. The Bobo doll study was designed so that the learning account could be tested directly.

The world around the study

Television spread fast in the United States in the 1950s, and by 1960 most American homes had a set. Parents, politicians and researchers had begun to ask whether violent programmes harmed children. The 1961 paper does not discuss television. Bandura's later Bobo studies did, and the research became central to public arguments about screen violence. You will read about that on the Why it matters page.

Two features of the time are worth noticing now, because the later pages return to them. First, the researchers and the children came from one university community in one wealthy country. Second, the paper treats boys' and girls' behaviour through the idea of sex-typing, the belief of the time that some behaviours naturally "belong" to one sex. Modern psychology describes these as gender roles that societies teach, and you will see that difference matter in Sections C and D.

Try this

A three-year-old has never been praised or punished for saying 'Oh, bother!' One evening at dinner she says it, in exactly her grandmother's tone, after spilling her juice.

How would Miller and Dollard explain this, and how would Bandura's idea differ?

Show the reasoning

Miller and Dollard would look for a history of reward for copying. They would say she has been reinforced in the past for imitating adults, so imitating has become a habit, and this phrase is one more example of that habit.

Bandura's idea is simpler and more direct. She heard her grandmother say it, stored it, and later used it. No reward was needed for this particular phrase, and she did not need to practise it when she first heard it. The Bobo doll study was designed to test exactly this kind of learning, with aggression as the behaviour.