Social Psychological Theories of Agression


A number of theories have been put forward by psychologists to explain aggression in terms of social or psychological factors. These include:

The tendency of human beings to imitate the aggressive behaviour of others, (sometimes called modelling),  especially where such behaviour is seen to go either unpunished or rewarded in some way.

The idea that a sense of social, economic or political injustice can lead to aggressive behaviour (sometimes called relative deprivation)

The idea that individuals who are in some way anonymous, i.e wearing a uniform or in a crowd can behave more aggressively than individuals who are clearly identifiable as individuals. This is called deindividuation theory

In addition there are a number of other theories such as cue arousal theory, desensitisation, disinhibition, excitation transfer theory, relative deprivation theory and social constructionism which have sought to explain aggression in social psychological terms of reference


Social learning theory originated from the work of Gabriel Tarde (1912)

Learning is:

a relatively permanent change in behaviour due to experience.

Classical behaviourism = direct experience (classical/operant conditioning

Neo-behaviourism= indirect / vicarious experience (observational learning through modelling)

Social Learning Theory claims we learn aggression primarily through observation of significant people around us= Modelling people who are similar (age/sex) or higher status (parent/teachers).

According to Tarde the key characteristics of imitation are :

1. Behaviour of role models (Fixed – zero sum)

2. Copying behaviour of a higher status (Variable)

3. Degree of contact with role model (Variable)

4. Degree of understanding of the behaviour (Variable)

Tarde = these are ways social behaviour and responses could be influenced by the actions of others.

Tarde combines both logic of social/cognitive psychology in social cognitive perspective of human behaviour.

Behaviour motivated by inherent psychological factors/ socio-environmental factors:

Individual and social environment linked= reciprocal determinism

SLT consists of four basic processes:

1. Attention= observation- watching the behaviour

2. Retention= cognitive process schema- storing the behaviour

3. Reproduction= behaviour- copying behaviour

4. Motivation= state- having reason for displaying behaviour


SLT  helps in explaining inconsistencies of aggressive behaviour – where we learn to act/or not- or be aggressive depending on different situations/contexts

Explanation of SLT based on research= ‘Bobo doll studies’

According to Bandura (1973) most behaviour, including aggression, is learned.

“The specific forms that aggressive behaviour takes, the frequency with which it is displayed, and the specific targets selected for attack are largely determined by social learning factors”



These three elements of social learning were demonstrated in Bandura’s research with children and the Bobo doll. In this study it was found that exposure to an aggressive model led to imitation of specific acts, generally increased levels of aggression, and aggression was directed at the same target.


Young children watched as an adult behaved aggressively towards a Bobo doll. The adult punched the doll and hit it with a hammer. After 10 minutes the children were moved to another room where there were some toys, including a hammer and a Bobo doll. Other children who had either observed a model behaving non aggressively or  had been exposed to no model were also led into the room. Once in the room they were watched through a one-way mirror and rated for their aggression.


The children who had watched a model behaving aggressively were more violent and imitated exactly some of the behaviours they had observed, as compared with children who either had seen no model or watched an adult (model) behaving in a non-aggressive manner.

Bandura (1965) carried out a further study on aggressive behaviour this time introducing reinforcement schedules.


One group of children simply saw a film of an adult model kicking and punching the Bobo doll. A second group saw the same aggressive behaviour performed by the adult model, but this time the model was rewarded by another adult for his aggressive behaviour by being given sweets and a drink. A third group saw the same aggressive behaviour but the model was punished by another adult, who warned him not to be aggressive in future.


Those children who had seen the model rewarded and  those who had seen the model neither rewarded nor punished behaved much more aggressively towards the Bobo doll than did those who had seen the model punished. All the children were rewarded for imitating as much of the model’s behaviour as they could remember. Thus, the children in all three groups showed comparable levels of observational learning, but those who had seen the model punished were least likely to apply this learning to their own behaviour.


Both experiments clearly demonstrated a modelling effect

The importance of reinforcement schedules in increasing or moderating aggressive behaviour was firmly established

Bandura exaggerated the extent to which children imitate the behaviour of models. Children are very likely to imitate aggressive behaviour towards a doll but they are much less likely to imitate aggressive behaviour towards another child.

Bandura consistently failed to distinguish between real aggression and playfighting, and it is likely that much of the aggressive behaviour observed by Bandura was only playfighting (Durkin, 1995).

The Bobo doll is of interest to young children, because it has a weighted base and so bounces back up when it is knocked down. Its novelty value is important in determining its effectiveness. Cumberbatch (1990) reported that children who were unfamiliar with the doll were five times more likely to imitate aggressive behaviour against it than were children who had played with it before.

Finally, there is the problem of demand characteristics. These are the cues used by participants to work out what a study is about.  In an experiment, participants try to guess what it is they should be doing  This leads them to search for cues which might help them and they use these cues, or demand characteristics, to direct their behaviour. Because experiments aim to have the same conditions for all participants all participants will be using the same cues and therefore they all end up behaving in ways that are predictable from the set up of the experiment.

As Durkin (1995, p.406) pointed out:

“Where else in life does a 5-year-old find a powerful adult actually showing you how to knock hell out of a dummy and then qiving you the opportunity to try it out yourself?

As such, the Bobo doll experiment provided cues which “invited” the participants to behave in certain predictable ways.


The essence of social learning theory is that new behaviours are learned indirectly as well as through direct reinforcement (traditional learning theory: classical and operant conditioning). Indirect reinforcement (vicarious reinforcement) results in observational learning. Vicarious reinforcement occurs when another person is observed to be rewarded for certain actions and this makes it more likely that an observer will imitate the actions. The imitator is not likely to repeat the behaviour immediately but may at an appropriate time in the future, reproduce the behaviour. Thus it is said that observational learning has taken place and the behaviour may be imitated or modelled at a later date. This means that a model must be stored internally, and implies the involvement of cognitive processes. This is a departure from traditional learning theory which rejects the involvement of any cognitive factors in learning. When aggression is imitated Individuals are more likely to imitate another’s behaviour if:

• The model is similar to themselves, such as being the same gender or similar in age.

•  The model is perceived as having desirable characteristics or is admired, as in the case of a rock star or an impressive teacher.

• The individual has low self-esteem.

• The individual is highly dependent on others.

• Reinforcement is direct, visible and immediate

Children respond most to direct reward, next to seeing a model in action, and least to a filmed model, especially a cartoon character (Bandura et al, 1963). Vicarious punishment may also occur, leading to a reduced response.

In addition social modelling may reduce the likelihood of a response because a different response has been strengthened. This was demonstrated in a study by Walters and Thomas (1963) who recruited participants for a study on the effects of punishment on learning. The participants worked in pairs, one was supposedly learning a task (this person was actually a confederate of the experimenters). The “true” participant was told to give the learner a shock following each error that was made. After each error, the participant was given the opportunity to select the level of shock to use for the next trial. Prior to the experiment all participants had been shown a film. Those participants who watched a violent scene were found to select higher shock intensities than those who watched a non-violent movie scene. This is an example of disinhibition. The participants observed socially unacceptable behaviour in the film and this weakened the pro-social behaviours they had previously learned. In other words their tendency to behave pro-socially was disinhibited or unlearned as a result of modelling.


KEY STUDY-Bandura, Ross and Ross (1961)

Experiment firstly involved: 72 child participants= different groups of children: experimental group –witnessed  adult model kick/punch Bobo doll/ control group- witnessed non-aggressive behaviour = After- different groups were put into the toy room with Bobo doll- their behaviour was observed.

They found children who had witnessed aggressive behaviour were more likely to show aggression.

Methodological issues:

Experiment had a well-defined way of coding the behavioural responses of the children to a measurable outcome.

Laboratory research- artificial setting/lacks ecological validity

Children’s behaviour- may have produced demand characteristics (wanting to ‘please the experimenter’)

Ethical issues:

Confidentiality- names of those involved maintained private- but films widely available on internet of experiment suggest compromise of confidentiality

Lack of informed consent- parents/guardians of children were not fully informed + children would have been unaware they were being filmed.

Study has face validity to explain children copying through imitating role models/ the media influence on aggression- such as influence of behaviour from t.v. personalities/ celebrities: e.g. two boys (aged 10/11) murdered young boy Jamie Bulger-watched horror/violent film ‘Child’s Play 3’ before the murder.

Bandura’s research focuses society’s attention on the power of the media- in areas of aggression and other related areas e.g. health (anorexia/bulimia)

Criticism of imposed etic/generalisibility- by assuming that processes of learning are the same for people in all countries/cultures universally.

Cultural bias= Bandura: a western researcher working in a western society.

Deterministic approach- to suggest child would passively absorb the observed behaviour and imitate it without logical thought for the implication of it.



Theory 2.


Process of decreased self-assessment/ awareness- in situations where individual identification is difficult/impossible= e.g. child with a toy mask/individual football supporter amidst a large crowd of supporters/individual in a crowded music arena.

Situations were individual identification is restricted ensures changes in the normal standards of behaviour/ inhibitions are lowered- Singer, Brush and Lublin (1965): Group situation->Inhibitions reduced-> Change in normal standards of behaviour/anti-social behaviour:

Milgram (1964/1965) – study of the situational determinants of destructive obedience= found participants were more likely to give higher levels of electrical shocks when they could not see/be seen by their victim.

Similar to this study:

Zimbardo (1969)- study showed the effect of reduced inhibitions:

He used female undergraduate participants in a ‘study of learning’. The female participants play teacher/ a stooge is used to play student. ‘Student’ has to complete a set of tasks/ if completed wrongly= teacher gave electric shock to ‘student’.

Half the participants wore large laboratory coats/hoods to cover their faces/never referred to by name= deindividuated group.

Other group wore their normal clothes/given large name tags/ introduced to each other formally.

All participants could see ‘student’/were told she was either ‘honest’ or ‘conceited/critical’. Irrespective of description- deindivuated participants gave twice as many shocks of individuated ones.


Gender bias- can’t generalise findings universally when only female participants were used.

Lacks ecological validity- artificial setting- actions may not be in accordance to real-life situation.

Ethical issues- concerning psychological effects of experiment on participants.

Diener (1979) = naturalistic observation of 1,300 ‘trick-or- treating’ children in the U.S. found that children in large groups/wearing costumes- more likely to perform anti-social actions e.g. stealing money/sweets. – the group ‘reduces the possibility of personal identification’- behaviour may deviate from normal moral standards

Factors that Affect Deindividuation:

Does physical anonymity always lead to deindividuation? Nurses uniform study

Diminished Self-Awareness Unself-conscious, deindividuated people are less restrained, less self-regulated, more likely to act without thinking about their own values and more responsive to the situation than self-conscious people

Loss of self-awareness and evaluation apprehension; occurs in group situations that foster anonymity and draw attention away from the individual

Group Size: Bigger crowds lead to more anonymity

Physical Anonymity Study: Women who wore white coats and hoods delivered more electric shock to “victims” than women who were visible and wore big name tags

Physical Anonymity Halloween Candy Study: Children were observed on Halloween night. They were told to take one candy and the experimenter left the room.

Deindividuation Loss of self-awareness and evaluation apprehension; occurs in group situations that foster anonymity and draw attention away from the individual


Deindividuation Loss of self-awareness and evaluation apprehension; occurs in group situations that foster anonymity and draw attention away from the individual


Fundamental problem of this theory= doesn’t provide explanation for the fact- not all crowds/groups perform aggressive actions:

Evidence linked to deindividuation and anti-social behaviour- but evidence also suggests deindividuation may lead to ‘prosocial’ behaviour=

Gergen et al. (1973)- study in which lowered levels of individuation didn’t result in aggressive actions= in a dark room – most participants were involved in intimate contact/at least half cuddled/ about 80% of the group felt sexually aroused.

Saying aggression is caused by lowering one’s inhibitions is narrow/deterministic:

Postmes and Spears (1998)= meta-analysis of over 60 studies investigating deindividuation= no consistent research to support argument- reduced inhibitions/anti-social behaviour more likely to be seen in large groups/crowded situations where anonymity is maintained.  = They suggest behaviour change of individuals in large groups- more influence from ‘group norms’.

Relating this theory to groups e.g. football hooliganism is too stereotypical: Marsh et al. (1978)= found what might appear to be an anti-social group actually consist of several different groups-with different places in the status ‘hierarchy’/ aggression was highly ritualised rather than physically violent.


Runciman (1966)= argues aggressive behaviour could be due to one’s relative deprivation- perceived difference between what you have and what you think you should have.

Dollard et al. (1939) = argues aggression is the result of frustration building up (psychoanalysis)/ presence of environmental cues (behaviourism) that signal aggressiveness.

Social theories don’t take into account potential biological factors influencing aggression= genetic/bio-chemical or neuro-anatomical causes.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s