Psychology II | Unit: Behaviourism and Aggression, pp. 29–33
Difficulty: Intermediate
Prerequisites: Read Part 1 (Social Learning Theory of Aggression, core theory) first. You need to know Bandura's four processes and the three types of vicarious learning before tackling these studies.
Bandura did not just propose social learning theory as an idea. He tested it experimentally. His Bobo doll studies (1961 and 1963) are among the most cited experiments in psychology and form the empirical backbone of SLT. Later researchers extended these findings beyond the lab, looking at whether real-world media exposure and family violence predict aggression over months and years. This set of notes covers the key studies you need to know, what each one found, and how they fit together.
Bandura's 1961 study showed that children who watched an adult behave aggressively toward a Bobo doll imitated that aggression. His 1963 follow-up showed the same effect from filmed and cartoon models. Later studies by Eron, Huesmann, Joy, Charlton, and Ehrenaft tested whether the same principle holds outside the lab, with mixed but broadly supportive results.
Bobo doll
An inflatable, weighted toy shaped like a clown that rocks back upright when pushed or hit. Used in Bandura's experiments as a target for measuring aggressive behaviour in children.
Laboratory experiment
A study conducted in a controlled setting where the researcher manipulates an independent variable and measures the effect on a dependent variable. Bandura's 1961 and 1963 studies are lab experiments.
Ecological validity
The degree to which findings from a study can be generalised to real-life settings. A common criticism of Bobo doll studies is low ecological validity: hitting an inflatable toy is not the same as real aggression toward a person.
Longitudinal study
Research that follows the same participants over an extended period. Eron et al. (1972) and Eron and Huesmann (1986) are longitudinal.
Correlation
A statistical relationship between two variables. A correlation does not prove that one variable causes the other. Several of the supporting studies report correlations between media violence and aggression.
Natural experiment
A study in which the researcher takes advantage of a naturally occurring event (rather than creating conditions in a lab). Joy et al. (1986) and Charlton et al. (2000) are natural experiments.
Demand characteristics
Cues in an experiment that may lead participants to guess the purpose of the study and adjust their behaviour accordingly. A concern with the Bobo doll studies is that children may have thought they were expected to hit the doll.
Aim
To investigate whether exposure to a real-life aggressive model increases aggression in children.
Procedure
72 children (36 boys, 36 girls), aged 3–6, from Stanford University nursery
Three conditions: (1) aggressive model, (2) non-aggressive model, (3) no model (control)
In the aggressive condition, the child watched an adult enter a room and behave aggressively toward the Bobo doll: punching it, hitting it with a mallet, kicking it, and shouting phrases like "Sock him in the nose"
In the non-aggressive condition, the adult played quietly and ignored the Bobo doll
Each child was then mildly frustrated (told they could not play with some attractive toys) and placed in a room with a Bobo doll, a mallet, and other toys
Observers behind a one-way mirror recorded the child's behaviour
Findings
Children in the aggressive model condition showed significantly more physical and verbal aggression than those in the non-aggressive or control conditions
Children often reproduced the exact behaviours and phrases of the aggressive model (imitative aggression)
Boys were more physically aggressive than girls overall
Children were more likely to imitate a same-sex model
Conclusions
Children can learn aggressive behaviour through observation alone, without direct reinforcement
The results support SLT: observation and imitation are sufficient for learning aggression
Key finding
Children exposed to an aggressive adult model imitated specific aggressive acts they had witnessed, demonstrating that aggression can be acquired through observational learning.
Aim
To investigate whether children would imitate aggression observed through filmed (on screen) models as well as live models, and whether a cartoon character would produce the same effect.
Procedure
Similar design to the 1961 study, but this time there were four conditions: (1) live aggressive model, (2) filmed (on-screen) aggressive model performing the same acts, (3) a cartoon character ("Herman the Cat") performing the same aggressive acts, (4) control group with no model
After exposure, children were again mildly frustrated and placed in a room with a Bobo doll and other toys
Behaviour was observed and recorded
Findings
All three experimental groups (live, filmed, cartoon) showed significantly more aggression than the control group
The live and filmed model conditions produced similar levels of aggression
The cartoon condition also produced significantly more aggression than the control, although slightly less than the live and filmed conditions
Conclusions
Aggressive behaviour can be learned through observation of filmed and cartoon models, not only live ones
This has implications for the influence of television and media on children's behaviour
The major new thing we learn from this study
Media representations (film, cartoons) are enough to transmit aggressive behaviour to children. The model does not need to be a real person in the same room. This extended SLT's relevance to the real world of television and screen media.
These studies extend Bandura's lab findings to real-world settings, longer timeframes, and different populations.
Eron et al. (1972)
A longitudinal study that followed children in New York State. Found a positive correlation between the amount of violent television watched at age 8 and the level of aggressiveness at age 18. This suggests that early exposure to media violence may be linked to later aggressive behaviour, consistent with SLT.
Eron and Huesmann (1986)
A follow-up and extension of Eron's earlier work. Continued tracking the same participants and confirmed that the correlation between childhood TV violence exposure and adult aggression persisted over time. Also found cross-cultural support for the link in several countries.
Joy et al. (1986)
A natural experiment studying a Canadian town (referred to as "Notel") that had no television until 1973. Researchers compared children's aggression levels before and after television was introduced, and also compared Notel with two similar towns that already had television. Found that aggression levels in Notel children increased significantly after television arrived, supporting the idea that media exposure can increase aggressive behaviour.
Charlton et al. (2000)
Studied the island of St Helena in the South Atlantic, which received television for the first time in 1995. Unlike Joy et al., found no significant increase in antisocial or aggressive behaviour among children after television was introduced. This study is important because it challenges a simple causal link between media and aggression, suggesting that other factors (community, parenting, social norms) may moderate the effect. St Helena had strong community oversight and low anonymity, which may have constrained any media influence.
Ehrenaft et al. (2003)
A longitudinal study that found children who witnessed domestic violence between their parents were significantly more likely to become involved in abusive relationships as adults. This supports SLT applied to family aggression: children observe and later reproduce the aggressive behaviour modelled in their home environment.
Students often state that the Bobo doll studies prove media causes aggression. They demonstrate that children can learn and reproduce observed aggression in a lab. That is an important distinction: learning in a controlled setting is not the same as causing real-world violence.
A frequent error is forgetting that the Bobo doll is a toy designed to be hit. Critics argue the children may have thought they were supposed to hit it (demand characteristics). Always mention this as a limitation.
Students sometimes treat all the supporting studies as if they produced the same result. Charlton et al. (2000) is a notable exception: television did not increase aggression on St Helena. You must include this as a counterpoint in essays.
Confusing correlation with causation is extremely common. Eron's studies found a correlation between TV violence and aggression. That does not mean TV violence caused the aggression. Third variables (e.g. family environment, personality) could account for the relationship.
⚠️ You will almost certainly be asked to describe Bandura's 1961 study. Know the aim, the three conditions, the key findings (imitative aggression, sex differences), and at least two criticisms.
⚠️ The 1963 study extends the 1961 findings to media. Be clear about the three model types (live, filmed, cartoon) and the conclusion that media models are also effective.
⚠️ Essay questions on media and aggression require you to weigh the evidence. Use Eron, Joy, and Charlton as supporting and contradicting evidence. The strongest answers note that Charlton et al. (2000) complicates the picture.
⚠️ Always evaluate the studies. Key criticisms of Bandura include low ecological validity, demand characteristics, ethical concerns (exposing children to aggression), and the fact that hitting a toy is not the same as aggressing toward a person. Key strengths include high internal validity (controlled lab conditions) and clear demonstration of observational learning.
True or false: In Bandura's 1961 study, children in the control group showed the same level of aggression as those in the aggressive model group. (False: aggressive model group showed significantly more aggression.)
Fill in the blank: Bandura's 1963 study added two new conditions beyond a live model: a ______ model and a ______ model. (Filmed; cartoon.)
True or false: Eron et al. (1972) found that watching violent TV at age 8 correlated with higher aggression at age 18. (True.)
True or false: Charlton et al. (2000) found that introducing television to St Helena significantly increased children's aggression. (False: no significant increase was found.)
Fill in the blank: A key criticism of the Bobo doll study is that the doll is designed to be hit, which raises concerns about ______ ______. (Demand characteristics.)
Q: Describe the aim, procedure, and findings of Bandura et al. (1961).
A: The aim was to investigate whether exposure to a real-life aggressive model increases aggression in children. 72 children aged 3–6 were assigned to three conditions: aggressive model, non-aggressive model, or control (no model). In the aggressive condition, an adult attacked a Bobo doll physically and verbally. Children were then mildly frustrated and placed in a room with a Bobo doll. Those exposed to the aggressive model showed significantly more imitative aggression (physical and verbal) than the other groups. Boys were more physically aggressive overall, and children were more likely to imitate same-sex models.
Q: What was the key contribution of Bandura et al. (1963) beyond the 1961 study?
A: The 1963 study demonstrated that children could learn and reproduce aggressive behaviour from filmed and cartoon models, not only from live models in the room. This extended SLT to media contexts and raised questions about the influence of television on children.
Q: Outline one study that supports and one study that challenges the claim that media violence increases aggression.
A: Supporting: Joy et al. (1986) found that aggression among children in a Canadian town ("Notel") increased significantly after television was introduced, compared with towns that already had television. Challenging: Charlton et al. (2000) found no significant increase in aggression among children on St Helena after television arrived. The community's strong social bonds and low anonymity may have moderated the effect.
Q: Give two criticisms of Bandura's Bobo doll studies.
A: (1) Low ecological validity: the Bobo doll is a toy designed to be hit, so the children's behaviour may not reflect real aggression toward people. (2) Demand characteristics: the children may have inferred that they were expected to hit the doll because an adult had done so in front of them, making the imitation a response to perceived expectations rather than genuine learned aggression.
Q: How does Ehrenaft et al. (2003) support SLT as an explanation for aggression?
A: Ehrenaft found that children who witnessed domestic violence between their parents were more likely to become involved in abusive relationships as adults. This supports SLT because the children observed aggressive behaviour modelled by their parents and later reproduced it in their own relationships, consistent with observational learning and imitation.
These studies connect to the broader debate about media influence on behaviour, which appears in media psychology modules and in discussions of censorship and regulation.
The methodological variety across the studies (lab experiment, longitudinal, natural experiment) is itself a useful discussion point for research methods topics. You can use these studies as examples when discussing internal vs. ecological validity, or correlation vs. causation.
Ehrenaft et al. (2003) links SLT to the cycle-of-violence hypothesis in developmental psychology, which examines how childhood experiences of abuse or witnessing abuse predict adult relationship patterns.
Bandura 1961, Bandura 1963, Bobo doll experiment, Bobo doll study, Ross and Ross, Eron 1972, Eron Huesmann 1986, Joy et al 1986, Notel study, Charlton et al 2000, St Helena, Ehrenaft 2003, media violence aggression, observational learning, imitative aggression, laboratory experiment, natural experiment, longitudinal study, ecological validity, demand characteristics, correlation causation, cycle of violence, Psychology II, FIU