PPsychology.com

Theories / How behavior is learned

The Bobo doll experiment

Children watched an adult hit an inflatable clown with a mallet. Then they were left alone in a room with the same clown. The ones who had watched hit it too, often in the same way, sometimes with the same words.

Published November 4, 2026. 4 sources at the foot of the page.

What it says

Albert Bandura built the study to test one claim. A person can pick up a behavior by watching somebody else do it, with no reward and no practice of their own. Nobody paid the children and nobody taught them. They copied.

The design had three groups of nursery school children. One group watched an adult attack the doll, one watched an adult ignore it and play quietly, and one watched nobody at all. Then each child was left with a room of toys, the doll among them.

The children who had watched the attack hit the doll far more often than either other group. They also reproduced particular acts: the mallet, the kicks, the lines said out loud. Boys hit more than girls, and both copied an adult of their own sex more closely.

What the 1961 design compared

Three groups of children, one room of toys, and the same inflatable doll waiting in each.

  1. The adult who attacked the doll

    A model hit the doll with a mallet, kicked it and said a set of lines out loud while the child watched.

  2. The adult who played quietly

    A model ignored the doll for ten minutes and worked with a set of construction toys instead.

  3. No model at all

    The child went straight through, which gives the rate of hitting with nothing to copy.

  4. The room before the test

    Every child was shown attractive toys and then told they were for somebody else, which left them mildly annoyed.

  5. The test room

    Each child was then left alone with the mallet, the doll and other toys, and watched for twenty minutes.

Where it came from

Bandura ran the study at the Stanford University nursery school and published it with Dorrie Ross and Sheila Ross in 1961. Seventy two children took part, half of them girls, aged roughly three to six.

Two follow ups matter as much as the first. In 1963 the same team showed that a film of a model, and even a cartoon, produced the copying almost as well as a live person. In 1965 Bandura added consequences, and that is where the useful distinction comes from.

What holds up, and what does not

The core finding is solid and has been shown many times since. Children copy what the adults around them do, including a stranger they have seen once. A model on a screen works nearly as well as one in the room.

The design has real limits. An inflatable doll built to be knocked down and to bounce back is not a person. What was counted is a play act, not harm done to anybody. The doll was also a novelty, and hitting it is what it is for, so some children arrived with the game already in mind. The measure was taken minutes after the model left, which says nothing about next week. The wider claim the study is used to support is much weaker: that watching violence makes children violent. A 2015 review of the video game research found effects too small to separate from each team's choices (Ferguson).

What it does not mean

It does not show that a child who watches something will go on to do it. The study counted copying in one room, minutes later, with the thing to copy sitting in front of them. It does not make the copying automatic either. The 1965 study found that what a child chose to do depended on what they saw happen to the model. What they were able to do was wider than what they chose.

Check yourself

Three questions on what is above. Each answer says why.

  1. Pick the statement that describes what the 1961 study measured.

    1. How aggressive each child was before the study
    2. How much each child copied an adult they had just watched
    3. Whether the children liked the adult they watched
    Show the answer

    Answer: B. The three groups differed only in the model they watched, so the comparison is about copying. Nothing was measured about the children beforehand.

  2. In the 1965 follow up, children who saw the model punished copied less on their own. What happened when those children were later asked to show what the model had done?

    1. Most of them could reproduce the acts
    2. Almost none of them remembered the acts
    3. They refused and left the room
    Show the answer

    Answer: A. Bandura reported that nearly all of them could reproduce the behavior when asked. Seeing a punishment changed what they chose to do, not what they had taken in.

  3. Which criticism of the design is the strongest?

    1. The children were too old to copy an adult
    2. A doll made to be knocked over is not a person, so the count is of a play act
    3. The study had no group that watched nobody
    Show the answer

    Answer: B. The doll springs back up, which is also the game, so hitting it is not harm to anybody. The study did have a group that watched no model at all.

Sources

  1. Bandura, A., Ross, D., and Ross, S. A. (1961). Transmission of aggression through imitation of aggressive models. Journal of Abnormal and Social Psychology, 63(3).
  2. Bandura, A., Ross, D., and Ross, S. A. (1963). Imitation of film mediated aggressive models. Journal of Abnormal and Social Psychology, 66(1).
  3. Bandura, A. (1965). Influence of models' reinforcement contingencies on the acquisition of imitative responses. Journal of Personality and Social Psychology, 1(6).
  4. Ferguson, C. J. (2015). Do angry birds make for angry children? A meta analysis of video game influences. Perspectives on Psychological Science, 10(5).

Practice it on Learn

Learn teaches this study in Studies that changed the field, the first unit of the Landmark studies track. The unit reads it beside the Milgram, Robbers Cave and Stanford prison experiments, each against its original paper.

Practice the Bobo doll experimentSee what Learn is

Every lesson is free. Plus is optional. For adults 18 and over.

More theories