Sign up to the PsychLogic newsletter at the bottom of this page to download printable AQA A-level Psychology notes + AQA A-level Psychology revision guide + how to revise for A-level Psychology + more...
The best way to revise Psychology A-level...




  • The limbic system
  • Serotonin and testosterone


  • Twin and adoption studies
  • The MAOA gene
  • Evaluation of genetic factors


  • Innate releasing mechanisms and fixed action patterns
  • Evolutionary explanations of human aggression


  • The frustration-aggression hypothesis
  • Social learning theory as applied to human aggression
  • Deindividuation


  • Prisons & dispositional and situational explanations


  • Television & computer games
  • Desensitisation, Disinhibition, Cognitive priming



Aggression is an innate, instinctual drive present in all complex organisms.

Early psychodynamic theories viewed aggression as arising from frustration of the Id’s attempts to satisfy basic desires. Later behavioural theories focused on how aggression could be acquired through learning and reinforcement from one’s social environment (e.g. family and peers) and the role of the newly-emerging media (film, TV, video games) in acquiring aggressive responses via imitation of powerful and admired role models (e.g. violent media heroes).

Advances in neurology, biochemistry and genetics have revealed how hormones, brain structures and inherited genes modify aggressive responses in individuals. Evolutionary psychology has highlighted how a variety of forms and expressions of aggressive behaviour can be accounted for by deep, instinctual inherited behaviour patterns aimed at maximising survival and reproductive success via food acquisition (e.g. hunting), defence, gaining and protection territory, group status, punishment for perceived social wrongs, and mate acquisition and retention.

Humans are one of the only species who kill other members of their own species, and from the 19th C. onwards the mechanisation of warfare has relocated aggression between groups from relatively small scale, hand-to-hand combat to mass, large scale, long range killing. Some human aggression seems purely sadistic – taking pleasure in the suffering of others.





The brain’s limbic system coordinates basic motivational/emotional states such as fear and aggression. It contains 2 structures that have been associated aggression.

  • The amygdala evaluates the emotional importance of stimuli and produces an appropriate response. Kluver (’37) found that removal of a dominant monkey’s amygdala caused it to lose its place at the top of the group hierarchy, and electrical stimulation of the amygdala will cause animals to automatically exhibit aggressive responses: e.g. cat arching back, hissing, etc.
  • The hippocampus stores long-term memories of previous experiences with stimuli. Therefore, animals may respond to other animals with aggression or fear depending on what they experienced in previous encounters. Damage to the hippocampus may cause the amygdala to respond inappropriately to stimuli – thus, aggression may be displayed at inappropriate stimuli (e.g. someone who is not a threat). Boccardi (’10) found that repeatedly violent offenders often displayed abnormal hippocampus functioning.

A study into the amygdala. Pardini (’14)

A longitudinal study was carried out on males from childhood to adulthood. Aged 26, MRI scans revealed a negative correlation between aggression and amygdala size: i.e. high levels of aggression = smaller amygdala size. This was true even when other potential confounding variables were taken into account. This implies that the amygdala does play an important role in accurately evaluating emotional stimuli and that lower amygdala volume may make inappropriate aggressive responses more likely.


Serotonin is thought to inhibit (prevent) behaviours. Therefore, if a person is provoked it will play a role in preventing the expression of an aggressive response. Low levels of serotonin are linked to impulsive behaviours including aggression and violent suicide.

  • Mann (‘90) gave 35 volunteers dexfenfluramine – a substance that reduces serotonin. The subjects self-reported increases in hostility and aggression, supporting the idea that aggression is at least partially biologically determined.
  • Raleigh (‘91) fed Vervet monkeys a diet either high or low in tryptophan to increase or decreases serotonin levels. High tryptophan diets decreased aggressive tendencies, whilst low tryptophan diets increased aggression.

A study into serotonin. Lindberg (’85)

Lindberg compared the serotonin levels of 16 males convicted of violent crimes (including murder), 22 males who had attempted suicide and 39 healthy males. Serotonin levels were measured from cerebrospinal fluid. Lowest levels of serotonin were found in those who had killed a partner or attempted suicide. This suggest that low levels of serotonin are associated with a failure to regulate emotions, resulting in negative emotions (anger, depression) and impulsive behaviours.


The sex hormone testosterone is thought to influence aggression from young adulthood onwards (particularly in men) due to its action on brain areas controlling aggression.

  • Wenger (‘79) noticed that after mice were castrated (which would have prevented the production of any more testosterone) aggression, operationalised through counting biting attacks, decreased. However, when these same mice were injected with testosterone, levels of aggression rose to pre-castration levels.
  • Klinesmith (’06) had male students provide a saliva sample to measure testosterone, then either interact with a gun or a child’s toy for 15 minutes and then provide another saliva sample. Males who interacted with the gun showed significantly greater increases in testosterone than the control group.

A study into testosterone and aggression. Dabbs (‘87)

Dabbs measured the salivary testosterone of violent and non-violent criminals and found that those with the highest testosterone levels had a history of primarily violent crimes and those with the lowest levels mainly non-violent crimes. This suggests that testosterone levels are directly linked to aggression.  As this study is based on prisoners, however, it lacks population validity and does not address more minor, general acts of aggression in the general population.




  • Coccaro (‘97) used a self-report questionnaire to measure hostility in 182 MZ and 118 DZ twin pairs. He concluded that genes contributed 47% to direct assault, 37% to irritability and 28% to verbal assault. As concordance rates were not 100%, however, environmental influences must also play a role. Twin studies assume that the environmental influence for MZ twins will be the same as for DZ twins, however, MZ twins usually have a closer relationship than DZs so it is difficult to separate out the effects of genetic similarity and the environment.


  • Adoption studies help separate out the contribution of environment and genes as the adopted away child is raised in a different environment to their genetic parents. If a positive correlation is found between aggressive behaviour in adopted children and aggressive behaviour in their biological parent, we can assume some genetic inheritability in aggression.
  • Hutchings (‘75) studied over 14,000 adoptions in Denmark. Where neither biological nor adoptive father had a criminal record, the son received a criminal record 10% of the time. This rose to 11% where the adoptive father had a criminal record, 21% where the biological father had one, and 36% where both adoptive and biological father had a criminal record. Although this strongly suggests a genetic link, the study measured whether sons received criminal records and only a small proportion of crime involves direct aggression.


  • Although no individual gene for aggression has been identified in humans, a gene responsible for producing a protein called Monoamine Oxidase A (MAOA) has been associated with aggression. MAOA regulates the metabolism of serotonin in the brain, and low levels of serotonin are associated with impulsive, aggressive behaviour.
  • Brunner (‘93) studied a Dutch family where many male members behaved aggressively and had been involved in violent crime. They were found to have abnormally low levels of MAOA and a defect in the gene responsible for producing MAOA.
  • Caspi (‘02) researched 500 male children, and discovered a variant of the gene associated with high levels of MAOA and a variant associated with low levels of MAOA. Those with low levels of MAOA were more likely to grow up to exhibit aggressive behaviour, but only if they had been mistreated as children. Children with high levels of MAOA who were mistreated, and those with low levels of MAOA, did not display abnormal aggression behaviour. This indicates that there is an interaction between genes (MAOA) and the environment (physical abuse in childhood) in determining aggression.


  • Studies have often focused on criminal behaviour rather than actual aggression, and on twins and adopted Research has not focused on the more low-level non-criminal aggression normal members of the population are likely to engage in. Also, even violent criminals may only have been violent once. Thus, extraneous variables and population validity are issues in most genetic studies of aggression.
  • There are numerous difficulties in determining exactly what influence genes have on aggression.
    • More than 1 gene probably contributes to aggression.
    • A variety of environmental influences (childhood trauma, negative life events) have the potential to affect aggression and separating out the effect of genetic inheritability from environmental influences is difficult.
  • Different studies have operationalised aggression in different ways and these may all have problems with validity. For example, self-ratings and peer/parent ratings are subjective and biased by personal interpretation and the respondent providing socially appropriate answers (e.g. under-estimating their own or their child’s aggressiveness).




Ethologists state that aggressive responses in animals are fixed action patterns (FAPs) which are:

  • Innate – they are instinctual and do not need learning
  • Stereotyped – the behaviour always occurs in the same way
  • Universal – is the same in all conspecifics (same species)
  • Have specific triggers known as ‘sign stimuli’ (see stickleback fish below).

For example, Tintenberg, studying FAP aggression in sticklebacks, found that male sticklebacks exhibit a FAP of aggressive behaviours upon another male entering their territory. The sign stimulus here is the male’s red underbelly. If this red sign stimulus is covered up the FAP of aggression will not occur.

Tintenberg used the term innate releasing mechanism to refer to the part of the brain which is stimulated by a sign stimulus and activates the FAP of aggression.

FAPs operate around a hydraulic model

  • Aggressive energy builds up in the animal.
  • A perceived threat causes the FAP to be automatically released.
  • This release causes energy levels to decline and the FAP cannot be repeated until energy levels rise again.

Lorenz argued that animal aggression served constructive purposes: for example, intraspecies (between the same species) aggression has the effect of distributing animals through an area that makes most effective use of resources such as food and mates. From an evolutionary perspective, aggression also maximises survival (hunting and defence) and reproductive success – competition for mates and passing on genes to the next generation.

The FAP for most animals contains a ritualised form of aggressive display where animals make physical and verbal threat displays (e.g. gorilla beating chest and howling) which allow the combatants to assess their opponent’s chances of winning and, if judged appropriate, back down without engaging in actual combat and suffering injury. For example, when wolves fight, if a wolf recognises that they are losing they will expose their throat to their opponent to indicate submission. This submission sign-stimulus prevents the dominant wolf from attacking further. Thus, for nearly all animals, ritualised displays of aggression and submission stimulus signs prevent combatants dying as a result of aggression and the conflict (e.g. over food or mates) is resolved.


  • Anthropologists claim to have found evidence of such ritualised aggression in traditional, pre-modern societies. For example, Chagnon (92) states that amongst the Amazonian Yanomamo, ritualised club fighting (where opponents take turn to hit each other over the head with a club) allows conflicts to be settled without fatalities.
  • Although Lorenz viewed animal aggression as constructive (serving positive functions), human aggression in modern, industrialised societies has become destructive. The mechanisation of weaponry has led to weapons capable of killing large numbers indiscriminately and conflict is no longer simply about resources/mates. Therefore, human aggression can be regarded as pathological (psychologically abnormal) and dislocated from its original biological/evolutionary functions.
  • Killing of conspecifics. Despite Lorenz’s claim that animal aggression is largely ritualised, many species do deliberately kill conspecifics (ants, lions, chimpanzees, chickens, various bears, etc.) However, this aggression may have alternative functions. Male lions who have taken over a new pride of lions will kill existing cubs to bring females back into oestrus (sexual receptivity) so that they can pass on their genes (rather than investing in another male lion’s genes). Cannibalism (chimps) may simply serve the function of additional dietary protein.
  • Critics have argued that Lorenz under-estimated the role that environmental factors can play in species’ aggressive behaviours – therefore, they are not as fixed and innate as he suggested. There also seem to be small variations in aggressive behaviours shown by different species members in different parts of the world so behaviours may be less universal than he claimed.



Evolutionary psychologists argue that differences in male and female reproductive strategies led to evolved differences in jealousy. Male sexual jealousy is a frequent cause of violence in interpersonal relationships. In many cultures, the murder of an adulterous wife or her lover is not only excused but encouraged. It is estimated that sexual jealousy accounts for 17% of all UK murders

Unlike women, men can never be entirely certain that they are the fathers of their children. As a result, men are always at the risk of cuckoldry (raising another man’s child whilst believing it to be his own). The consequence of cuckoldry is that the man might invest his resources in offspring that are not his own – evolutionarily speaking – disastrous. The adaptive functions of sexual jealousy and violence, therefore, would be to deter a female mate from sexual infidelity thus reducing the risk of cuckoldry.

Buss (‘88) suggested that males have a number of strategies evolved specifically for preventing the risk of cuckoldry. Direct guarding refers to protecting and observing mates closely: e.g. keeping women out of public spaces. Negative inducements involve threatening violence against infidelity – either personal (bullying, domestic violence) or state-authorised (e.g. legal punishments for adultery).

Evolutionary theory predicts men to more sexually jealous than women. Buss (‘92) found that men showed greater distress than women (higher galvanic skin responses - measures emotions like fear or anger), when asked to imagine scenes of sexual infidelity by their partners. In this scenario, Takahashi (’06) found greater activation of the amygdala and hypothalamus (areas associated with aggression) in men than women. Females, on the other hand, showed greater distress in response to their partners being emotionally unfaithful. Evolutionary psychology would explain this as being due to females fearing abandonment and a subsequent removal of resources by a male provider.


From an evolutionary perspective the universal human tendency to aggress against neighbouring groups (war) would only have evolved if it brought adaptive benefits to the individuals involved and their offspring. At a basic level, warfare is aimed at acquiring more resources for the group (land, food, sexual partners) which will increase survival chances and reproductive success.

  • Warfare is present in very few species apart from man. However, our closest relatives – chimpanzees – do engage in remarkably similar inter-group conflict: e.g. killing, mutilation and cannibalism.
  • Successful fighters may be rewarded by access to female mates, and soldiers frequently engage in rape. Displays of aggression and bravery are, traditionally, attractive to females. Chagnon (’88) found that amongst the Yanomamo, successful warriors had more sexual partners and more children.


  • Evolutionary psychology can be accused of being deterministic, suggesting that men are locked into behavioural patterns of violence against sexual partners simply to avoid cuckoldry. Feminists might suggest that this argument rationalises and, therefore, legitimises male violence.
  • The evolutionary approach fails to take account of individual differences in jealousy and violent behaviour – for example, why some men are particularly jealous and/or violent.
  • A methodological problem with the use of surveys to collect information about partner violence is social desirability bias – both sexes may under-estimate or lie about levels of violence they exhibit or receive.
  • Evolutionary explanations suggest that man is genetically predisposed to group aggression and war is, therefore, natural. Anthropologists often report on non-western, hunter-gatherer societies as very peaceable, however. This may depend, however, on population density and competition for resources: e.g. tribes in Papua New Guinea live in high population density and tend to be warlike.




  • Dollard (’39) defined frustration as ‘any event or stimulus which prevents an individual from attaining some goal’ (i.e. when we cannot get what we want), and argued that frustration led to aggressive urges which might be expressed.
  • Whether aggression will be expressed depends on factors such as fear of punishment or legitimacy of target (e.g. most people would not act aggressively to a child who frustrated them). Thus, situational factors can inhibit It should be noted, however, that although we may not act in a physically aggressive manner, we may be verbally or emotionally aggressive to the source of our frustration.
  • Dollard argued that there was a direct cause-effect relationship between frustration and aggression. Furthermore, the act of aggression is cathartic: i.e. it releases us from the negative emotional/physical/psychological state caused by frustration and makes us feel better.
  • Frustration is heightened when (i) our motivation to achieve something is very strong; (ii) we expect gratification but do not receive it; (iii) there is nothing we can do to remove the source of frustration: i.e. it is out of our control.

Research by Pastore (’52) found that what we perceive as unjustified/unfair frustration produce more anger and aggression than frustration we label as justified in some way. For example, we would be angrier about being frustrated by poor exam results if we had worked hard than if we had not revised at all.

If the individual/situation causing frustration is not available to direct our aggression against (e.g. the government) or is not perceived as a legitimate target (e.g. a parent), then aggression may be displaced onto a weaker, more easily accessible target. This could be minor (e.g. a frustrated teenager displacing anger onto their younger sibling) or major (e.g. a nation suffering frustration as a result of poor economic conditions displacing aggression onto minority groups: e.g. Nazi Germany and Jews).


  • Berkowitz criticised Dollard’s theory, arguing that (i) aggression often arises in the absence of any frustration and (ii) frustration does not always lead to aggression. Berkowitz argued that frustration is just one example of the negative emotions which could potentially be caused by frustration. Negative emotions in general cause a state of anger, and aggression is just one response to anger (for example, people may cry or become depressed rather than violent). Berkowitz argued that unanticipated frustration is much more likely to cause aggression rather than frustration which we can anticipate.
  • The original frustration-aggression hypothesis drew on a Freudian model of human drives: i.e. the Id’s blocked desires result in aggressive instincts and aggression is reduced through engaging in cathartic activities (acting out and expressing aggression). There is a lack of research support for such a view, however. For example, research into catharsis has produced contradictory findings with some research supporting the view that engaging in violence decreases aggressive urges whilst other research has found the reverse – i.e. that acting aggressively increases aggressive urges.
  • A study by Priks (’10) attempted to objectively assess the cause-effect relationship between frustration and aggression in sport.
    • Using Swedish football teams, Priks operationalised frustration through teams’ changing position in the league, and aggression through number of missiles fans threw during matches. He found that when teams performed worse than fans expected (g. unanticipated frustration) aggression rose with a 1-position drop in the league table causing a 5% increase in aggression.
  • Although the Frustration-Aggression hypothesis is appealing in its simplicity and application to minor and major everyday incidences of aggression which clearly result from frustration of some kind, it over-simplifies the relationship between these variables and fails to take into account biological factors, individual differences and the role of cognitions and previous learning experiences which will influence and modify whether we act aggressively and if so, how we express aggression.


SLT argues that aggression may be learnt via observation and imitation of others. Although SLT admits aggression is a biological instinct, it claims that the way it is expressed is learnt.

  • Modelling: the imitation of observed aggression: e.g. seeing a father commit domestic violence against a mother.
  • Vicarious Learning: imitation as a result of seeing another individual being positively reinforced for aggression: e.g. respect from peers for fighting, etc.

Bandura identified 4 mediational processes (cognitive factors) in social learning that will influence whether or not aggression displayed by a model will be learnt and then displayed.

  • Attention: the observer must observe the model behaving aggressively.
  • Retention: the observer must remember the aggressive acts they have seen.
  • Reproduction: the observer must be capable of imitating the observed aggression.
  • Motivation: the observer must be willing to imitate the observed aggressive act.

Bandura argued that aggressive behaviours may be acquired through social learning from a child’s social environment (family, friends) and from the media. Imitation is particularly likely if the model is of the same gender/age and is perceived as powerful, high status, friendly and/or liked.

The observer will form a mental representation of events they have seen, including possible rewards and punishments for their aggressive behaviour in terms of expectancies of future outcomes – when the reward is greater than the punishment they are likely to perform the action in the future.


Support for SLT comes from a series of laboratory experiments.

Bandura (61) exposed male and female children aged 3-5 to films of adult models interacting either aggressively or non-aggressively with life size, human shaped Bobo dolls. The aggressive models displayed distinctive behaviours such as hitting the doll’s head with a mallet, kicking, and verbal violence. Following exposure to the model, children were frustrated by being shown attractive toys which they were then not allowed to play with. They were then taken to another room where, amongst other toys, was a Bobo doll. Children who had observed the aggressive models tended to reproduce a good deal of the physical and verbal aggression they witnessed with 1/3rd of aggressive condition children (no gender difference) repeated the model’s verbal aggression and boys reproducing more physically aggressive acts. Children in the non-aggressive condition exhibited virtually no aggression. This study provides evidence for modelling.

Bandura’s study has been criticised on the following grounds. Firstly, children may have displayed demand characteristics believing the experimenter wanted them to aggress against the doll. Secondly, the study lacked ecological validity – children would have been able to tell the difference between attacking a doll and a real human. Thirdly, the sample only consisted of a fairly small group of 3-5 year olds.

In a modification of the original study, Bandura (63) found that children who observed models being rewarded for aggressing against the doll showed a higher level of aggression towards the doll whilst those who saw the model being punished displayed lower levels. This provides evidence for vicarious learning.

In an elaboration of the above study, Bandura (65) offered the children in both reward and punishment conditions further rewards if they could display aggression against the doll. All children did so, thus proving that even though the children in the punishment condition who did not originally aggress did not perform the behaviour they had in fact learnt it.

Williams (86) studied Canadian children in a natural experiment and found that after the introduction of TV, incidents of aggression increased by 160% in children, in line with what SLT would predict.

A study by Phillips (86) provides some evidence for this effect on adults: after major boxing match the homicide rate in the US almost always rises for approximately a week.

SLT can be applied to concerns regarding children’s imitation of violent role models in the media. Huesmann suggested children may use TV models as a source of scripts that act as a guide for their own behaviour. Real life depictions of violence cause greater imitation than unrealistic violence such as cartoons.


 Le Bon (1895) proposed that individuals in crowd situations could be psychologically transformed. Anonymity creates a state of deindividuation characterised by

  • Decreased self-evaluation
  • Decreased concerns about evaluation by others
  • This leads to an increase in behaviour normally inhibited by personal/social norms: e.g. uninhibited aggression

 Zimbardo’s Stanford Prison Experiment (71) investigated social roles of guards and prisoners and the link between deindividuation and increased aggression.

  • 24 psychologically normal volunteer male students took the role of either guard or prisoner in a ‘mock’ prison. Prisoners had personal possessions removed, were dressed as prisoners and assigned ID numbers. Guards referred to the prisoners only by their number and were given uniforms, clubs and reflective sunglasses (to prevent eye contact).
  • Within a few days the guards became psychologically and physically abusive to the prisoners. The prisoners ‘rebelled’ against the guards’ authority by taking off their ID badges. The guards locked the prisoners in their cells. Rapidly, the guards seem to begin to enjoy sadistically (taking pleasure in others suffering) exercising power over the prisoners. Over the course of a few days, prisoners became passive, depressed and stressed The study was planned to run for 2 weeks but was called off after 6 days due to the guards’ brutal behaviour.


  • Ethics. Some participants suffered physical and psychological harm, and it is argued that Zimbardo had a moral responsibility to stop the study as soon as the guards showed any signs of brutality.
  • Ecological Validity. Clearly, the prison was not real and the participants (guards and prisoners) were engaged in a role play rather than a real-life situation, knew they could leave the experiment when they wished, and were only confined for a short period of time. To what extent we can generalise findings to real institutions and real aggression by guards against prisoners is, therefore, debatable.
  • Zimbardo’s interpretation of his participants’ behaviour was that when put in a social role with absolute power even psychologically normal individuals are at increased likelihood of behaving aggressively to those with no social power. The deindividuating effect of the prison and the uniforms seemed to encourage brutality and violence.

In a further laboratory experiment (Zimbardo ’69), female subjects were deindividuated by wearing lab coats and hoods and were instructed to deliver (fake) electric shocks to a victim. Compared to normal subjects, deindividuated ones delivered longer and more painful shocks.

There is a large body of experimental research and anecdotal (non-experimental/observations of everyday life) evidence which supports deindividuation theory – from football mobs to riots to bystanders urging potential suicides to kill themselves.

Prentice-Dunn (82) distinguishes between

  • Public self-awareness – a concern about the impression one gives to others. This can be reduced by anonymity and the fact that other members of the crowd act as role models whose social norms or standards of behaviour fellow crowd members may conform to. This reduces public self-awareness and make uninhibited, anti-social behaviour more likely.
  • Private self-awareness – a monitoring of our own feelings, thoughts and behaviours. To maintain self-esteem and our self-concept we usually inhibit anti-social or socially abnormal thoughts and behaviours. In an anonymous crowd we may ‘forget ourselves’, lose our usual private self-awareness and conform to other crowd members’ behaviour.

The major disagreement between psychologists studying deindividuation is whether crowds cause us to lose control over our own anti-social tendencies or whether we simply conform to the anti-social tendencies of others.

However, deindividuation does not always lead to aggression, and many crowd situations are peaceful (e.g. festivals) and marked by increases in pro-social behaviour and strong feelings of group belonging. For example, following natural disasters we may expect increases in looting – very often, however, we find group efforts focused on aiding victims.



Institutional aggression refers to violence between groups in institutions: prisons, the armed forces or mental institutions.

In the UK, assaults in prisons rose by 61% from ’00-‘09. A ‘07 report in the US estimated that 70,000 inmates were victims of sexual violence while serving their sentence.

The Dispositional Model states that high levels of prison violence are the result of what prisoners bring to prison in terms of attitudes, personality traits and norms.

  • Irwin (62) argues that prisoners bring unusually violent traits and normative systems which legitimise violence into prison and high rates of violence are not, therefore, surprising. Wang (’99) found that anger, anti-social personality and impulsivity were stronger predictors of institutional aggression than ethnicity or type of crime committed.
  • Gang membership is associated with high levels of prison violence, and pre-prison gang membership is an important predictor of prison misconduct. Gang members in the US are x10 times more likely to commit murder than a non-gang member of the public. Again, it is unsurprising that importing and concentrating gang members who hold norms which legitimise violence into a confined environment breeds high levels of violence.

The Situational Model states that prisoner aggression is the result of stress, oppression and deprivation experienced within the institution. For example:

  • Overcrowding. A 2014 UK government report blamed record numbers of murder, suicides and assaults on increased overcrowding in prison.
  • Heat & noise. High temperatures and noisy conditions act as sources of stress and frustration which may trigger violence.
  • Job burnout. Overworked, demotivated prison officers lead to poorer relationships with prisoners that has been linked to decreased concern about prisoner violence.
  • Sykes (58) found that perceived threats to personal security increased inmate anxiety that in turn made aggressive outbursts more likely. Individuals reacted differently to this threat – some withdrawing whilst other rebelled against prisoners or staff.


The Situational Model received support from McCorkle (’95) who found that overcrowding, lack of privacy and lack of meaningful activity significantly increased violence.

A real world application of findings from the Situational Model was implemented at HMP Woodhill (a UK prison) in the early 90’s. It was argued that violence could be reduced by reducing the environmental stressors of heat, noise and overcrowding. 2 units were set up for prisoners which were less claustrophobic, cooler and quieter, yet despite the fact that these changes virtually eradicated assaults on prison staff and other inmates, political pressure (not wanted to be seen to be soft on prisoners) meant these changes were not permanently implemented or exported to other prisons.

The Dispositional Model received support from Harer’s study (’06) which found that US Black inmates had higher levels of violent behaviour in prison whereas Whites had higher levels of alcohol and drug-related misconduct. These differences parallel differences in these crimes in the American civilian population, therefore behaviours from ‘outside’ prison are ‘imported in’.

Research in the US by Kane (’81) found that those who had long periods of unemployment, low levels of education and a more serious criminal record were more likely to use aggression in prison, suggesting that dispositional factors rather than situational factors lead to prison violence.

Mills (‘98) found that alcohol dependency was associated with greater levels of aggression. Again, this provides support for the view that individual characteristics of the prisoner are responsible for violence rather than environmental factors/deprivation experienced within prison.




Since the 60’s there has been increasing concern that aggression portrayed in film, TV, music and computer games may be imitated.



  • Parke (77) exposed young offenders to either aggressive or neutral films. Participants were observed for a week to gain a baseline measurement of aggression, then watched either violent or non-violent films for a week, then were observed for a further week to measure changes in aggression. Boys who were rated in the first week as aggressive showed increases in aggressive behaviour after the week of violent viewing.
  • Field experiments suffer from a lack of control over extraneous variables. Therefore, it is possible that other external factors were causing increases in the violent TV group’s aggression.

Correlational Studies

  • Evidence from correlational studies is fairly consistent – viewing and/or a preference for violent TV is related to aggressive attitudes and behaviours. Atkin asked 9-13 year olds what they would do if another child pushed them off their bicycle. The children were given response options including verbal or physical aggression and the chance to avoid or minimise conflict. 45% of heavy violent TV viewers chose violent options compared to only 21% of low violent TV viewers.
  • However, correlational studies only indicate a relationship between these variables, not that violent TV leads causes real-life violence. It could well be that already violent children are attracted to watching more violent programmes.


  • Laboratory experiments have found increases in physiological arousal, aggression and hostile emotions after playing violent video games.
  • Anderson (‘07) surveyed violent video games use in 430 children aged 7-9. High exposure children were rated as more verbally and physically aggressive, and were rated by their teachers and themselves as less pro-social (acting in ways which benefit others).
  • Moller (’09) conducted a longitudinal survey into the relationship between violent video games and aggressive behaviours and thoughts among 13-16-year-old students. Questionnaires assessing violent video game usage and self-rated violent tendencies were taken at the start of the study then 30 months later. Increased game use was associated with increased violent tendencies. Although it is possible that extraneous variables caused increases in violence, this study does seem to indicate direct relationship between increases in violent video game playing and increases in aggressive behaviour and thoughts.


  • Although the above examples do indicate a relationship between violent media and violent behaviour many studies have found no relationship.
  • Aggressive behaviour is influenced by numerous variables such as gender, biological factors such as testosterone or genetics, low socio-economic status, low levels of education, peer group, family background, etc. It is virtually impossible, therefore, to isolate the media as a single cause of increased aggression.
  • Most people have seen huge amounts of fictionalised violence yet remain unaffected. If media violence does cause imitative behaviour it is likely that this only occurs with individuals who score low on measures of empathy and moral development, high on scores of fantasy, and are likely to possess violent temperaments already.
  • Traditionally, violence on TV was often illustrated as being ‘punished’. From a SLT viewpoint, therefore, this punishment should decrease imitative aggression. It has been argued that the way in which modern TV/film depicts violence is now more glamourised/put into a more positive/rewarding context. This is particularly the case with violent video games.



Assuming that we are normally inhibit aggressive impulses due to anxiety (for example, we fear punishment or have ‘learnt’ it is wrong), repeatedly viewing of media aggression may remove this anxiety, making us believe aggression is a more usual or appropriate response and thus reducing our anxiety about behaving aggressively. This is referred to as desensitisation.

Desensitisation can be measured through

  • Measures of physiological arousal (heart rate, electrical conductivity of skin, etc.) which are a good indicator of how anxious a person is.
  • Changes in cognitive (thoughts) and affective (emotions) reactions to aggressive stimuli. For example, a desensitised individual may show less sympathy for victims. These changes may make aggressive responses more likely.

Carnagey (’07) tested desensitisation by asking participants to play either a violent or non-violent computer game for 20 minutes, then watch scene of real life violence for 10 minutes. Those in the violent condition showed significantly lower heart rate and electrical skin conductivity when viewing the violent film compared to those in then non-violent condition. This certainly provides evidence that watching violence leads to a reduction in physiological anxiety, although it does not prove that those individuals are more likely to act aggressively.


People are usually inhibited from behaving aggressively due to learnt rules regarding what behaviours are appropriate/acceptable and that aggression can lead to punishment. By repeatedly viewing aggression which is either rewarded, glamourised or not punished, people may be less inhibited about behaving aggressively themselves and come to believe that using aggression is a legitimate response and is socially normal.

A heightened state of arousal can produce lowered inhibitions (we are more concerned with acting than calmly thinking). Thus, violence in the media can produce a state of arousal which in turn reduces inhibitions which can lead to aggressive responses.

Goranson (’69) investigated the effects of the consequences of violence on disinhibition. In 1 condition, participants saw a boxing match where there were no consequences to the violence; in condition 2 they saw the same boxing match but this time the loser died as a result of his injuries. Participants in condition 2 were much less likely to behave aggressively in a test of aggression after the experiment than participants who had been disinhibited from behaving aggressively by being shown the potential negative consequences of aggression.


Berkowitz argued that repeated viewing of violence will lead to a network of memories or ‘scripts’ concerning violent behaviours: e.g. how one should behave in potentially violent situations. These memories may be recalled in similar situations in real life if any aspect of the original media situation is present. For example, a child who has viewed a film where an argument results in the use of a nearby weapon may respond similarly in a real life argument. Thus, aggressive acts in films can activate similar aggressive thoughts and behaviours in viewers through their association in memory pathways.

Bushman (’98) tested participants’ reaction speed to aggressive words after they had watched either a violent or non-violent film. Those in the violent film condition showed much faster reaction speeds to aggressive words but did not show faster reaction time to non-aggressive words. Thus, viewing aggression certainly seems to increase availability of aggression thoughts in our memory.