Generalisability
A sample in all three studies was large:
72, 96 and 66, large enough that anomalies (eg disturbed children) might be cancelled out (eg by particularly mild mannered ones). The samples were all taken from the same nursery, which was for the students and staff at one of the world’s top universities. These children might have unusual home lives and particularly educated parents, making them unrepresentative of normal children.
Another problem is generalising from children to
adults. This might not matter if all of our important behaviour is learned in childhood (even if we don’t act it out until adulthood – the 1965 study shows children can learn behaviours but not act them out until later). However, the studies may not tell us much about how adults learn new behaviour because adults might be less influenced by role models.
Reliability
Bandura’s procedure is very reliable because it can be
replicated – as Bandura did, replicating the study in ’63 and ‘65. This was easy to do because of the standardised procedure (same script, same checklist categories, etc).
Bandura also used
two observers behind the one-way mirror. This creates inter-rater reliability because a behaviour had to be noted by both observers otherwise it didn’t count.
Finally, Bandura
filmed the 1963 study and the films can be watched by anyone, which adds to the inter-rater reliability.
ApplicationBandura, Ross & Ross (1961) can be applied to parenting and teaching styles. It suggests children observe and imitate adults, so if you want your children to grow up calm and well-behaved, you need to keep your temper and keep them away from aggressive role models. Calm role models seem to have a big effect, which might apply to “buddy” systems used in schools or prisons to help troubled students or prisoners learn from a role model.
Bandura, Ross & Ross (1963a) has much more application to TV censorship. Bandura claims the study was inspired by a news story about a boy in San Francisco who was seriously hurt when his friends re-enacted a TV fight scene. The study suggests even cartoon violence (like Tom & Jerry) might be causing children to learn aggressive behaviour. This study is used to support censoring TV, films and video games aimed at children.
The
1963 study also counts against the “catharsis” argument which is often used to defend violent sports like boxing or WWE (which is very popular with young boys). Defenders often say watching wrestling helps audiences “vent” their aggression harmlessly, but Bandura suggests the opposite is true. If Bandura is right, these sports should not be shown to children.
Bandura (1965) also applies to media censorship. Heroes in TV shows, films and video games are often rewarded for using violence: James Bond, Spider-Man, Lara Croft and every Arnold Schwarzenegger movie. In video games, violence is explicitly rewarded by “levelling up”; in films the violent hero saves the day or gets the girl. More media censorship might reduce violence in society. Alternatively, films and games should be made to show the real consequences of violence rather than the rewards.
ValidityThe main criticism of all Bandura’s studies is that they lack validity. The children were put in a strange situation, exposed to some unusual adult behaviour and given toys to play with which encouraged them to act unnaturally. For example, a Bobo Doll is designed to be hit and knocked over (it bounces back upright); children would suppose the experimenters wanted them to play with the Bobo Doll in this way. This sort of behaviour is called
demand characteristics, because the participants do the stuff they think the researchers demand of them.
Bandura did address this by creating a film in which an adult woman attacked another adult dressed as a clown; when an actor in a clown costume entered the observation room, the children used plastic mallets to aggress against him.
(NB. Although this "real life clown" variation is mentioned in many textbooks and websites, I can't find a citation for it. I wonder if it's one of those Psychology 'urban myths'. Nevertheless, it's so well-attested in textbooks it is surely appropriate for students to use in the Exam)
The other criticism of Bandura's conclusions is that there are other explanations for aggression - biological ones. The study by
Raine et al. (1997) shows that aggression is linked to certain brain deficits, like a weak prefrontal cortex; people with these deficits might need no excuse to start behaving aggressively and misinterpret the role model's behaviour as an invitation to do so.
There are many ethical issues with Bandura’s studies. The major issue is
harm and the wellbeing of participants. The children may have been distressed by the aggressive behaviour they witnessed and the aggressive behaviour they learned from the study may have stayed with them, going on to become a behavioural problem. Participants are supposed to leave a study in the same state they entered it, which may not have happened here. This is an example of what the BPS Code of Ethics calls "normalising unhelpful behaviours".
Although the children could not give
valid consent to take part, their nursery teachers agreed and it is assumed the children’s parents agreed to; this is presumptve consent. Nonetheless, the children could not withdraw from the study and no effort seems to have been made to debrief them afterwards (by explaining that the aggressive adults were only pretending).
Bandura would argue that the
benefits to society outweighed the risks to any of the children that took part. His research has shown us the influence that role models have on aggressive behaviour, especially role models on TV and film. This has been an importat contribution to the debate over censorship in TV, films, videos and games.
