Bandura et al. (1961)

Aim To investigate whether a child would learn aggression by observing a model and would reproduce this behaviour in the absence of the model, and whether the sex of the role model was important.

To investigate whether a child would learn aggression by observing a model and would reproduce this behaviour in the absence of the model, and whether the sex of the role model was important.


  1. Observed aggressive behaviour will be imitated.
  2. Observed non-aggressive behaviour will be imitated.
  3. Children are more likely to copy a same-sex model.
  4. Boys will be more likely to copy aggression than girls.

Children copy adults. The social setting makes the child imitate what he/she is watching.
Imitative learning: Learning a new behaviour through observing a role model, and imitating it later in the absence of the model.
Children are also differently rewarded for their copying behaviour depending on their sex.

Research Method
It was a laboratory experiment with controlled situations.

Research Design and Variables
Independent measures design was implemented as different children were used in each condition of the study. There was matched participants design where children were divided into 3 groups based on their initial levels of aggression.
Independent Variable:
• model-type: aggressive or non-aggressive model
• model-gender: whether the model is the same gender as the child or not.
• learner-gender: whether the child was a boy or girl.
Dependent Variable: The behaviour the child displayed, which was measured through a controlled observation.

72 children aged 3-6 years. There were 36 boys and 36 girls obtained from the Stanford University nursery. The sampling method used was opportunity sampling.

Before the experiment, the children were observed in their nursery school by the experimenter and a teacher who knew them. They were rated on 4 five-point scales measuring physical aggression, verbal aggression, aggression to inanimate objects, and anxiety. They were assigned to 3 groups with their aggression levels matched. 51 children were rated by 2 observers and similar ratings were produced therefore, there is inter-rater reliability.

12 boys and 12 girls were in control groups who saw no model. The remaining children were divided equally by sex between aggressive and non-aggressive model groups, and within those, between same and opposite same-sex models.


The experimenter and children were in the 'play area' where they made potato prints. In the opposite corner of the room there was a table, chair, Tinkertoy set, mallet, and a 5-foot Bobo doll. This is where the model sat, in the conditions where there was one. The experimenter remained in the room but appeared to be working quietly at their desk.
➔ In the non-aggressive condition, the model assembled the Tinkertoy set for 10 minutes.
➔ In the aggressive condition, the model assembled the toy for a minute and then attacked the Bobo doll. For 9 minutes the aggressive sequence was repeated 3 times. Aggressive comment made: "Kick him."
➔ In the control condition, the student did not see any model.

The experimental procedure continued when all participants were deliberately mildly annoyed. Children were told that they could play with the toys in the room, but 2 minutes into playing, they were told to stop. They were informed that the toys were reserved for other children and that the toys were the experimenter's very best.

The child was observed for 20 minutes through a 1-way mirror. The experimental room had a 3-foot Bobo doll, a mallet and other toys. Their behaviour was observed in 5-second intervals. There was an inter-rater reliability score of 0.89.

Children exposed to aggressive models imitated their exact behaviour. They were significantly more aggressive than the children in other groups. Imitation was greater for boys than girls. Boys were more likely to imitate physical aggression and girls were more likely to imitate verbal aggression.

The average imitative physical aggression for male subjects with male models was 25.8 which is much higher than that for female subjects which was 7.2. This exhibits that boys imitated physical aggression more than girls.
However, with a female model girls imitated less aggression than with the male model. Girls with female models imitated an average of 5.5. Boys with female models imitated an average of 12.4.
For non-aggressive play, girls played with dolls and boys played with guns. Comments were made on sex-typed behaviour.


  • Observation and imitation cause behaviour to be learnt without reinforcement.
  • Observed aggressive behaviours are imitated.
  • Observed non-aggressive behaviours are imitated.
  • Children are more likely to copy a same-sex model.
  • Boys are more likely to copy aggression than girls.


  • It was a laboratory experiment, so it was possible to control extraneous variables. For example, all children had seen their models for the same length of time. Model behaviour was standardised. This makes the research more valid, reliable, and replicable.
  • There is high inter-rater reliability, and it leads to accurate data collection.
  • Children were unaware that they were being watched, and this increased validity as they are less likely to show demand characteristics.


  • Only 6 children were used in each experimental condition and this is a small sample. Further, they all attended the same nursery and had academically able parents. This could bias the sample and lower the validity.
  • Children may have imitated behaviour, due to demand characteristics and social desirability, as they might have thought that they had to imitate the model.
  • A longitudinal study would have better explained the durability of the newly learned behaviour.

Ethical Issues
Some children might have been harmed by becoming more aggressive. Children had been mildly annoyed, which could be psychologically distressing.
Children didn't have the opportunity to consent or withdraw. As the study may have caused distress to the children, this is a key ethical issue. The headteacher was aware, but the parents' consent is missing.

Application to Everyday Life
The study shows that when children (especially boys) are exposed to violence or aggression in real life or through the media, the models influence their behaviour.

Nature vs. Nurture
Boys imitated more aggressive behaviour than girls because they have more hormone testosterone, which is a nature factor. Boys may be more likely to imitate aggression because they have learned about male stereotypical behaviour.

Individual vs. Situational Explanation for Behaviour
The situational influence of models had led children to imitate aggressive behaviour.
Individual factors explain why the acquisition of behaviours differs between boys and girls, and it may be because they are differently rewarded for sex-typed behaviours.