You are viewing a free preview of this lesson.
Subscribe to unlock all 9 lessons in this course and every other course on LearningBro.
Bandura, Ross and Ross's 1961 study — universally known as the "Bobo doll" study — is one of the most cited experiments in all of psychology and the classic study for the developmental theme of external influences on children's behaviour. Its claim is deceptively simple and, at the time, genuinely radical: children learn how to behave, including how to be aggressive, by watching other people and imitating them. Before Bandura, the dominant learning theories held that behaviour was shaped by direct reinforcement — a child did something, was rewarded or punished, and adjusted accordingly. Bandura's demonstration that children would acquire and reproduce a novel aggressive act simply from observing an adult do it, with no reward to the child at all, was the foundation of social learning theory and a decisive challenge to a purely reinforcement-based account of development.
This lesson tells the study in the OCR "tell the story" format: the theoretical background, the aims and hypotheses, the method (its careful matched design, the sample of nursery children, and the step-by-step procedure), the results with their real figures, the conclusions, and a full evaluation of its method, data, ethics, validity, reliability, sampling and cultural reach. It closes by linking the study to its theme, its area, the behaviourist perspective and the debates it informs. Because Bandura anchors the "external influences" theme and the whole social-learning tradition, knowing it precisely is one of the highest-yield investments for Component 02.
| This lesson covers | OCR H567 Component 02 element | AO focus |
|---|---|---|
| Background: social learning theory versus reinforcement accounts | Section A — Developmental; theme: external influences on behaviour (classic) | AO1 knowledge |
| Aims and hypotheses (imitation; same-sex modelling; sex differences) | Section A — Core study (Bandura) | AO1 |
| Method: laboratory experiment; matched design; sample (n and who); procedure | Section A — Core study | AO1; AO2 |
| Results (imitative and non-imitative aggression; the model-sex and child-sex effects) | Section A — Core study | AO1 |
| Conclusions (observational learning of aggression) | Section A — Core study | AO1; AO3 |
| Evaluation: method, data, ethics, validity, reliability, sampling, ethnocentrism | Section A; Section B debates | AO3 |
| Links to theme, area (Developmental), behaviourist perspective, debates | Section B — Areas, perspectives, debates | AO1; AO3 |
The specification is referenced descriptively; consult the official OCR H567 specification document for its exact published wording. This lesson develops AO1 (aim, procedure, results, conclusions), AO2 (applying social learning to novel cases of children's behaviour) and AO3 (evaluating the study's ethics, validity and generalisability).
By 1961, learning theory was dominated by the idea that behaviour is acquired through direct reinforcement — the operant conditioning of Skinner, in which the consequences of one's own actions shape future behaviour. Bandura did not reject conditioning, but he argued it was radically incomplete as an account of how children acquire the vast repertoire of behaviour they display. Children plainly learn enormous amounts — language, social rituals, skills — far too quickly to have been directly reinforced for each element, and often before they have ever performed the behaviour themselves. His alternative was social learning theory: the proposal that much learning happens vicariously, by observing others (called models) and imitating them, and that this observational learning can occur without any reinforcement to the observer at the moment of learning.
Aggression was an ideal test case. If children could be shown to acquire a specific, novel aggressive act — one they were very unlikely to invent for themselves — merely by watching an adult perform it, that would be powerful evidence for observational learning. The inflatable Bobo doll, a weighted clown figure that rocks back upright when struck, provided a target on which a distinctive set of aggressive behaviours could be modelled and later scored. Crucially, some of the aggressive acts Bandura's models performed were unusual (for example, sitting on the doll and punching its nose, or striking it on the head with a mallet while saying set phrases) — precisely so that if a child later reproduced them, imitation could be inferred rather than coincidence.
Bandura also wanted to test who children imitate. Social learning theory predicts that observers imitate models they identify with, and one strong basis for identification is sex. He therefore built into the design a comparison of same-sex versus opposite-sex models, expecting children to imitate a same-sex model more readily — an expectation especially strong for boys imitating the (culturally sex-typed) physical aggression of a male model.
There is an important logical feature of the design worth drawing out, because it is what makes the study evidence for observational learning rather than merely a demonstration that frustrated children hit things. If the aggressive-condition children had simply been more aggressive in general, that would not prove they had learned anything from the model — they might just have been more aroused. What clinches the case is specific imitation: the children reproduced the model's particular and unusual acts and words. No child would spontaneously invent "sit astride the doll and punch its nose while saying 'sock him in the nose'"; if a child performs exactly that after watching a model do it, the only plausible source is the observed model. Bandura therefore engineered the situation so that a positive result could only be explained by observational learning, and he built in comparison conditions (non-aggressive model, no model) so that any general tendency to hit the doll would show up in those groups too and could be subtracted out. This is a beautiful piece of experimental logic, and articulating it is one of the surest ways to lift an evaluation of the study from description to genuine analysis.
The wider theoretical stakes were considerable. If Bandura was right, then a great deal of socially significant behaviour — not only aggression but altruism, gender roles, language and prejudice — could be transmitted from one person to another simply by being seen, without any deliberate teaching and without the observer ever being rewarded. That is a claim with enormous reach: it implies that the models a child is exposed to, in the home, the playground and (increasingly, by the 1960s) on television, are not incidental to development but among its engines. The Bobo doll study was designed to establish the basic phenomenon under laboratory control so that this larger claim could rest on firm foundations.
Bandura's overarching aim was to test whether children would learn aggression by observing an aggressive adult model, and would reproduce that aggression in the model's absence. Within this, the study tested four specific predictions:
The study was a laboratory experiment with an independent-measures design across conditions, and — importantly — a matched-participants element: before allocation, every child was rated for their existing aggressiveness so that the groups could be equated. The independent variables were the behaviour of the model (aggressive / non-aggressive / no model), the sex of the model (male / female), and the sex of the child (boy / girl). The dependent variable was the amount and type of aggression the child later displayed, measured by structured observation.
The participants were 72 children — 36 boys and 36 girls — enrolled at the Stanford University Nursery School, ranging in age from about 37 to 69 months (roughly 3 to 5¾ years), with a mean age of around 4 years 4 months (52 months). This is an opportunity sample of nursery children available to the researchers.
To ensure the groups were comparable in baseline aggressiveness, the children were pre-rated on four five-point scales — physical aggression, verbal aggression, aggression toward objects, and aggression inhibition — by the experimenter and a nursery teacher who knew them well. The inter-rater reliability of these ratings was high (a correlation of about 0.89), and the children were then allocated to conditions by matched-pairs on the basis of these aggression scores, so that each condition contained children of comparable pre-existing aggressiveness. This matching is one of the study's methodological strengths and a frequent focus of exam questions.
The children were divided as follows: 24 in the aggressive-model condition, 24 in the non-aggressive-model condition, and 24 in the control condition (no model). The two model conditions were further subdivided by model sex and child sex, producing eight sub-groups of six children each in the modelling conditions.
The key materials were a room with attractive non-aggressive toys (a tea set, crayons, colouring paper, a ball, dolls, plastic farm animals), a set of aggression-eliciting toys (a mallet and peg-board, dart guns, and a five-foot inflatable Bobo doll), and, in the model's room, a smaller three-foot Bobo doll along with a Tinkertoy set and mallet.
The study proceeded in three stages, and the whole point of the design is the gap between the first two: the child observes in stage one but is not invited to imitate; the aggression measured in stage three is therefore spontaneous reproduction, not compliance.
Stage 1 — Modelling. Each child was brought individually to a room and seated at a table with the non-aggressive toys (potato prints, stickers) in one corner. In the aggressive condition, an adult model was brought to the opposite corner, given the Tinkertoys, mallet and the small Bobo doll, and — after a minute — turned to the Bobo doll and behaved aggressively toward it in a distinctive, scripted sequence: laying it on its side, sitting on it and punching it repeatedly on the nose; picking it up and striking it on the head with the mallet; tossing it in the air and kicking it around the room. This sequence was repeated about three times over roughly nine minutes, accompanied by verbal aggression — set phrases such as "Sock him in the nose," "Hit him down," "Throw him in the air," "Kick him," "Pow," and two non-aggressive remarks ("He keeps coming back for more" and "He sure is a tough fella"). In the non-aggressive condition, the model simply played quietly and gently with the Tinkertoys for the whole period, entirely ignoring the Bobo doll. Children in the control condition did not see any model at this stage.
Stage 2 — Aggression arousal. All children (including controls) were then taken to a second room containing attractive toys and, after starting to play, were told that these were the experimenter's very best toys, reserved for other children, and that they must stop. This mild frustration was included deliberately: it was intended to arouse a little aggression in all children, so that the crucial test in stage three would not simply be whether the aggressive-condition children were more aroused, but whether they channelled arousal into the specific imitative forms they had observed.
Stage 3 — Test for delayed imitation. Each child was taken to a third room for 20 minutes, containing both non-aggressive toys (tea set, crayons, dolls, bears) and aggressive toys (the three-foot Bobo doll, mallet, dart guns). The child was observed through a one-way mirror, and behaviour was recorded using time sampling: an observer noted what the child was doing at five-second intervals, giving 240 response units per child. Behaviour was scored into pre-defined categories: imitative aggression (exact reproduction of the model's physical and verbal acts — sitting on Bobo, mallet blows to the head, the specific phrases), partial imitation, non-imitative aggression (aggressive acts not modelled, such as shooting the dart guns or hitting other objects), and aggressive gun play. A second observer scored a proportion of the children independently to establish inter-rater reliability, which was very high (around 0.90).
The findings supported the core prediction clearly: children who had watched an aggressive model reproduced a great deal of the model's specific behaviour, whereas children in the other two conditions almost never did.
| Measure | Aggressive-model condition | Non-aggressive / control |
|---|---|---|
| Imitative physical and verbal aggression | Substantial — children reproduced the model's distinctive acts and phrases | Almost none in the non-aggressive and control groups |
| Non-imitative aggression | Also somewhat raised | Lower |
| Partial imitation | Present | Rare |
Several specific patterns are important for the exam:
Subscribe to continue reading
Get full access to this lesson and all 9 lessons in this course.