Bandura, Ross, and Ross (1961)

Context

Concerned with the tendency of children to imitate adult social behavior, specifically aggression.

Observational learning- learning behavior by imitating others

Several earlier studies had demonstrated that children are influenced by witnessing adult behavior, tended to show kids repeating adult behavior in the same situation with the presence of the adult.

One purpose of this study was to test whether children will reproduce adult behavior in new situations without the adult

This study is also concerned with the learning of gender-specific behavior, previous studies have shown that kids are sensitive to it.

Ex. Agression is associated with masculinity

Also investigated whether boys were more likely to imitate aggression than girls, and whether they were more likely to imitate male than female models

Aim

To investigate observational learning of aggression. The study aimed to see whether kids would reproduce aggressive behavior when the model was not present, and to look for gender differences in learning aggression.

Participants

72- 36 male, 36 female

All selected from nursery school of Stanford

Ages from 37 months (3 years) to 69 months (5 years and 9 months)

Mean age was 4 years and 4 months

Design

Lab Experiment

Matched Pairs Design

Effect of 3 IVs-

The behavior of the model- aggressive or nonaggressive

Sex of model

Sex of children

8 conditions in all- kids in each condition were matched for their aggression levels so it wasn't a confounding variable (researcher and teacher rated 51 kids on a scale of 0-5, very good agreement achieved)

-An aggressive model was shown to 12 boys and 12 girls, six boys and six girls saw aggression modeled by a same-sex model, half saw if from a different sex model.

-A nonaggressive model was shown to 12 boys and 12 girls, half saw no-aggression modeled by same sex model, half saw if from a different sex model.

-A control group of 12 boys and 12 girls did not see a model display behavior

Procedure Stage 1

Modeling the Behavior

Kids were brought individually into a play room to play a game for ten minutes.

In the first two conditions there was additional adult in the room.

Agressive Condition- adult was aggressive toward a tall inflatable doll, saying aggressive things such as "kick him"

Nonaggressive Condition- adult assembled toys and did not interact with the doll

Control Condition- No other adult

Stage 2

Aggression Arousal

To annoy the kids and increase chances of aggressive behavior, they were taken to a different room with toys, they played with them for about 2 minutes but then were told they weren't allowed to play with them any more because they were "the very best" toys and were for other kids

Stage 3

Testing for Delayed Imitation

Children were observed playing for the next 20 minutes.

Experimenter was in the room occupied with paperwork, 2 more observers watch through a two-way mirror

The room had a lot of toys and smaller inflatable doll

Observers didn't know what condition the kids were in

Types of Aggression Recorded

Imitative- physical and verbal aggression identical to the ones modeled

Partially Imitative- Similar behavior to that carried out by the model

Non-imitative- New aggressive acts not demonstrated by the model

Results

Kids who witnessed an aggressive model were significantly more aggressive themselves

There was little difference between aggression in the control group and the non-aggressive conditions

Boys were significantly more likely it imitate aggressive male models, difference for girls was much smaller

Boys were significantly more physically aggressive, girls slightly more verbally aggressive

Conclusion

Witnessing aggression in a model can be enough to produce aggression by an observer

Children selectively imitate gender-specific behavior

Boys more likely to imitate physical aggression, girls more likely to imitate verbal aggression

Cautiously concluded that children selectively imitate same sex models

Evaluation

Lab Experiment

A lot of controls in place (the time kids watch model for, layout of room, which toys available), so reliable

Controls were high so the researchers could be confident that action of model cause the aggressive behavior

Setup was artificial because the kids were in a setting not familiar to them, low ecological validity

Some task expected from kids were unusual (watching an adult play with toys and not getting involved), low mundane realism

Quant. data

Ethics- protection, aggressive behavior might have continued