Bandura (1961) Transmission of Aggression

Aim:

• To see whether children will imitate aggressive behavior – even if in different environment and without a model present.
• Hypotheses:

1. Children will imitate the aggressive behavior of models.
2. Non-aggressive models will have an inhibiting effect.
3. Same-sex model will have more influence.
4. Boys will imitate aggression more than girls.

Sample:

 

• 72 children: 3 to 5 years old/ Stanford University Nursery school.

• 24 with aggressive role model

• 24 with non-aggressive role model.

• 24 in control group.

12 kids saw the same-sex role model, whereas the other 12 saw the opposite sex.

 

Procedure:

Laboratory Experiment with Observation.
Participants were matched for their aggression levels – researchers and nursery teacher rating 51 of the children on a scale from 0 to 5.

Independent Variables:

1. Role Model: Non-aggressive, Aggressive, No model.
2. Gender of Role Model: Male or Female.
3. Gender of Participant: Male or Female.

Dependent Variables:

1.Imitative aggression responses:

• Physical: Any acts imitated.
• Verbal: Phrases imitated.

2. Partially imitative responses:

• Mallet aggression: using it on other toys.
• Sitting/Bouncing on Bobo doll.

3. Non-imitative aggressive response:

• Slapping Bobo doll.
• Physical and verbal aggression not imitated.
• Aggressive gun play.

Stage 1: Modelling

Children were individually out into a corner of a room containing toys for 10 minutes while either:

• Watching a male or female behaving aggressively towards the Toy called “Bobo doll”. Hit with a hammer, threw it in the air or shouted “Pow, Boom.”
• Exposed to a non-aggressive model who played quietly with Tinker Toy.
• Used as a control group, therefore not exposed to any model.

Stage 2: Aggression Arousal

Children individually taken to a room with colourful and attractive toys.

After 2 minutes – Experimenter told child that the toys were reserved for other children since they were the very best toys they had and taken to a different room.

Stage 3: Test for Delayed Imitation

Room contained aggressive toys – mallet and peg board/dart guns/Bobo doll – and non-aggressive toys – tea set, crayons, bears, plastic farm animals.

20 minutes inside the room – observed and behaviour rated trough a one-way mirror.

240 response unit for each child – 5 second intervals observations.

Results:

• Children who observed the aggressive models made far more aggressive responses than the other 2 groups in all 3 conditions.
• Boys showed more physical aggression, whereas girls showed more verbal aggression.
• Boys were more likely to imitate same-sex models, however both sexes tended to do so.
• Girls in the aggressive model condition showed more physical aggression if the model was male, whereas they showed more verbal aggression if the model was female.

Conclusions:

• The study supports the Social Learning theory as the children did observe and imitate the behaviours.
• Not all behaviour is shaped by reward or punishment.
• It has important implication for the effects of media violence on children.

Strengths:

 

• Sample: Equal number of girls and boys and the children were matched across all 3 conditions, therefore the results are highly generalisable.

• High control: Since it was a laboratory experiment, the researchers had a lot of control over variables, which helps to decrease the chances that the behaviour was due to extraneous factors. This makes the results more reliable

• Replicable: Standardized procedures and instructions were used, allowing it to be repeated and checked for reliability.

 

Weaknesses:

 

• Ethics: Protection of participant – children were subjected to aggressive behaviour, which could have harmed their mental health.

• Generalisability: The study only looked at young children , therefore the results cannot be generalised to all children of different ages.

• Low ecological validity: The aggressive situation was not very realistic since it is rare that a social situation involves a child and an adult model. Also, they are strangers which is unlikely in life because modelling usually happens in families.

• Experimenter bias: One of the male model was an observer, which could have caused bias since he knew the aim of the study and what to look for.