Bandura et al. · aggression
Follow the model-exposure experiment from observed adult behaviour to children’s later play. Distinguish imitation in a test room from claims about lasting or everyday aggression.
Exam scope and assessment notes
Original StudyForA teaching mapped to the Cambridge 2028–2030 syllabus. These lessons are independent study support, not Cambridge material, official mark schemes or awarded grades. No lesson reproduces a research article or Cambridge question. Human subject-expert review and full timed-paper grading remain pending. This lesson synthesises the named study from its original article and Cambridge's study outline; it is an original summary, not a reproduction of the paper.
Before you begin
- This is one of the 12 compulsory AS core studies in the 2028–2030 Version 1 syllabus. Check that your exam entry uses this syllabus year.
Work on paper. Try each question before revealing help, and explain why your method works.
Quick readiness check
Why did the study code distinctive actions that resembled the model?
What you will learn
- Explain the study aim, sample, procedure and findings
- Evaluate the method, ethics and evidence
- Apply the relevant approach and issues/debates without overclaiming
Approach and research focus
Learning approach. Bandura and colleagues tested whether children reproduce actions after observing an adult model. The study examines how observed behaviour can influence later behaviour even when the model is absent.
Background: Learning explanations emphasise experience and observable behaviour. The researchers asked whether children would reproduce a model’s aggressive acts in a new situation, including distinctive actions unlikely to occur spontaneously.
Aim and procedure
The sample was 72 children (36 boys and 36 girls), aged 37–69 months (mean 52), from a university nursery. Researchers matched children using prior aggression ratings and allocated them to aggressive-model, nonaggressive-model or control conditions. In an experiment, children observed an adult model or no model, then were observed later in a separate play situation without the model. Behaviour was coded using predefined categories; model sex and child sex were also compared.
Results and conclusion
Children exposed to the aggressive model later reproduced more physical and verbal actions resembling the model’s behaviour. Children who saw a subdued nonaggressive model and children in the no-model control group rarely performed those distinctive imitative acts. The paper also reports that model-sex patterns differed across physical and verbal aggression; these details should be interpreted in the study’s historical context rather than treated as universal gender rules.
The findings support the claim that observing a model can increase reproduction of particular aggressive actions in a later play setting. They do not establish that children became generally or permanently aggressive outside the study.
Evaluate the evidence
The control group, matched baseline ratings and coded observation allow structured comparisons across conditions.
The nursery sample was young and from one setting; a staged playroom with an inflatable toy is not everyday social behaviour.
Observers may interpret behaviour differently, although the original study assessed agreement on baseline ratings; measurement quality should be considered for each coding scheme.
The result demonstrates short-term imitation in a particular context, not long-term effects or aggression toward real people.
Ethics and issues/debates
Young children require careful protection, appropriate consent from guardians and age-appropriate assent where possible.
Showing children an aggressive model raises potential harm concerns; a modern replication would use safer, nonviolent actions, close supervision and a clear stopping procedure.
Do not claim that a specific debrief or consent step occurred unless the original report documents it.
The learning approach explains a short-term behaviour through observed environmental models; the study does not test every proposed mental process in social learning theory.
The experiment supports an effect of condition in this setting, but the artificial play task limits ecological validity.
Applying the finding to media or everyday aggression requires evidence beyond this one nursery study.
Sources for this lesson
Pause and explain
Which conclusion stays closest to the evidence?
Worked example
A report says, ‘Bandura proved that watching aggression makes children violent.’ Improve the claim.
Show the worked solution
- Identify what children observed and how behaviour was measured.
- The outcome was coded imitation in a later play situation, not violence in ordinary life.
- The experiment supports a context-specific short-term effect, with generalisation limits.
Answer The study found that children exposed to an aggressive adult model reproduced more modelled actions in a later play setting. It does not prove permanent or general violence.
Try a topic test
Choose Easy, Normal or Difficult. Save your answers and open worked review after finishing.
Test this topic →From guided practice to a new situation
Write your answer and reasoning first. Use a hint only if you are stuck. The model solution is for self-checking; your written work is not automatically marked.
The aggressive-model group later reproduced distinctive modelled actions more often than the other groups. State one supported conclusion and one boundary.
Give me a hint
Name the measured behaviour and the test setting.
Compare my reasoning
- Children who saw an aggressive adult model later showed more similar actions in the play test.
- This supports short-term imitation after observation in the study context.
- It does not establish that children became persistently aggressive in everyday life.
The experiment supports short-term imitation of modelled actions in a later play setting, not a claim about lasting aggression.
Look for these in your work
- Result linked to model exposure
- Test behaviour specified
- Long-term claim rejected
Explain how baseline matching and a no-model control help interpretation, then name one design limitation.
Give me a hint
Consider pre-existing behaviour and what the control does not rule out.
Compare my reasoning
- Baseline ratings helped distribute children with different starting aggression levels across conditions.
- The no-model group provides a comparison for changes after exposure to a model.
- The playroom task and young nursery sample limit how far the findings generalise beyond this setting.
Matching and the control group strengthen the comparison, while the sample and artificial play setting limit generalisation.
Look for these in your work
- Matching purpose explained
- Control purpose explained
- Limitation tied to external validity
Propose a safer study of whether children copy a helpful adult model during a puzzle task.
Give me a hint
Use a harmless action, guardian permission, child assent and a clear behaviour measure.
Compare my reasoning
- Show a brief helpful strategy, such as sorting puzzle pieces by colour, to one group and a neutral comparison to another.
- Observe whether children independently use the strategy on a comparable puzzle, using a pre-defined coding rule.
- Obtain guardian consent and child assent, supervise the task, allow withdrawal and avoid rewarding one group in a way that makes the comparison unfair.
A safe model-and-comparison design could measure strategy imitation without modelling aggression, with child protection and a standardised coding scheme.
Look for these in your work
- Outcome observable
- Comparison is fair
- Child ethics addressed
Common mistakes
- Describing the 72 children as all being in the aggressive-model condition
- Confusing a no-model control with a nonaggressive-model group
- Treating imitation of a toy-directed action as real-world violence
- Presenting historical model-sex findings as universal truths
What can you explain now?
What did the control group experience before the test-room observation?
Compare with the explanation
No adult model exposure
The nonaggressive-model group observed a subdued model; it is a separate comparison condition.
After trying it yourself, choose your next review. This is your self-assessment.
Your review choice appears on Today. Sign in to sync it across devices.
Make it stick
Name the aim, the best-evidenced result and one methodological limitation of Bandura et al..
If you needed the solution, close it and solve the problem again from a blank page. A correct answer is stronger when you can explain the reason for each step.