What you'll learn

• How frustration, social learning, and de-individuation can explain human aggression.
• Key studies such as Dollard et al. (1939), Bandura et al. (1961), and Zimbardo (1969).
• How to apply each theory to scenarios for AO2 and evaluate it for AO3.
• Common exam traps, especially confusing the three explanations.

Big picture: what counts as a social explanation?

A social psychological explanation looks for the causes of behaviour in a person’s social environment: other people, group situations, social rewards, social norms, and immediate triggers.

In this topic, you are not trying to explain aggression mainly through genes, hormones, brain structures, or evolution. Instead, you are asking: what is happening in the social situation that makes aggression more likely?

Keep the three pathways separate as you revise:

Frustration-aggression hypothesis

The frustration-aggression hypothesis was proposed by Dollard et al. (1939). It argued that frustration always leads to some form of aggression, and that aggression is always caused by frustration.

For example, if someone is prevented from getting home, winning a game, gaining respect, or getting something they feel entitled to, this may create frustration.

Dollard et al. suggested that frustration produces an aggressive drive: an internal pressure or motivation to behave aggressively. If the person cannot aggress against the real source of frustration, aggression may be displaced onto a safer or more available target.

Later, Berkowitz (1989) revised the theory. He argued that frustration does not automatically cause aggression. Instead, frustration creates anger or negative affect, which makes aggression more likely if there are suitable aggression cues in the environment.

An aggression cue is something associated with aggression, such as a weapon, threatening language, or a hostile facial expression.

AO3 evaluation of frustration-aggression

A strength is that the theory has supporting evidence. Berkowitz and LePage (1967) found that angered participants gave more electric shocks to another person when weapons were present than when neutral objects were present. This supports the idea that aggression cues can increase aggressive behaviour.

There is also real-world support. Harris (1974) found that people reacted more aggressively when someone pushed into a queue close to the front, where their goal was nearer. This supports the link between goal-blocking and aggression.

However, the original version is too deterministic. Not every frustration leads to aggression: people may cry, withdraw, problem-solve, or seek social support. Also, not all aggression follows frustration. Instrumental aggression, meaning aggression used as a means to achieve another goal, may be planned rather than emotionally triggered.

Another weakness is the mixed evidence for catharsis. Catharsis is the idea that releasing aggression reduces aggressive drive. Research such as Bushman (2002) suggests that “venting” anger can actually increase later aggression, so the hydraulic “release” idea is questionable.

There are also ethical and methodological issues. Shock-based laboratory studies often involve deception, limited fully informed consent, and potential emotional distress. Researchers must protect participants from harm, allow the right to withdraw, maintain confidentiality, and debrief them afterwards. Field studies of queue-jumping have higher ecological validity but often lack prior consent.

Social learning theory and aggression

Social learning theory, associated with Bandura (1977), argues that people can learn behaviours by observing others. This is especially important for aggression because children and adults may learn aggressive acts without being directly rewarded themselves.

A model is a person whose behaviour is observed and may be copied. Imitation means copying the model’s behaviour. Vicarious reinforcement occurs when an observer sees someone else being rewarded, making the observer more likely to copy that behaviour.

Bandura also argued that learning involves mediational processes, which are mental processes between observing and performing behaviour. These include:

• Attention: noticing the aggressive behaviour.
• Retention: remembering what happened.
• Reproduction: being physically or socially able to copy it.
• Motivation: expecting a reward or benefit from copying it.

A child is more likely to imitate a model they identify with. Identification means seeing the model as similar, attractive, powerful, high-status, or worth copying.

The classic evidence comes from Bandura, Ross and Ross (1961). Children watched an adult behave aggressively towards a Bobo doll. Later, children who had seen the aggressive model showed more imitative aggression than children in non-aggressive or control conditions. Boys showed more physical aggression, and children were more likely to imitate same-sex models.

Bandura’s later work on vicarious reinforcement showed that children were less likely to perform aggressive behaviour if they had seen the model punished, but they could still reproduce it when offered an incentive. This suggests they had learned the behaviour even if they did not immediately perform it.

AO3 evaluation of social learning theory

A major strength is strong experimental support. Bandura et al. controlled the environment carefully, making it easier to show a cause-and-effect relationship between observing aggression and later imitation.

The theory also has useful real-world applications. It helps explain why aggressive behaviour may be transmitted through families, peer groups, gangs, or media. It suggests interventions such as rewarding non-aggressive behaviour, reducing exposure to aggressive models, and showing clear negative consequences for aggression.

However, the Bobo doll studies may lack ecological validity. A Bobo doll is designed to be hit, so children may have believed aggression was expected. This raises the issue of demand characteristics, where participants change their behaviour because they guess the aim of the study.

There are also ethical concerns. The children were exposed to aggressive adult behaviour and may have experienced distress or learned aggressive actions. Researchers needed parental consent, protection from harm, and debriefing. Modern ethical standards would require careful justification.

A further limitation is that social learning theory may underplay biological factors and individual differences. Not everyone exposed to aggressive models becomes aggressive. Temperament, self-control, family relationships, culture, and biological factors may affect whether observed aggression is copied.

De-individuation

De-individuation is an explanation for aggression in crowds or anonymous situations. It suggests that when people feel less personally identifiable, they may feel less accountable for their actions.

Key factors include:

• Anonymity: not being easily identified, such as wearing a mask or being online.
• Reduced self-awareness: paying less attention to personal values and standards.
• Diffusion of responsibility: feeling less personally responsible because responsibility is spread across a group.
• Weakened internal restraints: reduced guilt, fear, or concern about social judgement.

This can increase aggression because the person feels less likely to be blamed, punished, or personally evaluated.

Zimbardo (1969) tested this using female participants who delivered electric shocks. Some wore hoods and were not named, making them more anonymous. Others were identifiable and wore name tags. The anonymous participants gave longer shocks, supporting the link between anonymity and aggression.

Other evidence comes from Diener et al. (1976), who studied children trick-or-treating. Children who were anonymous and in groups were more likely to steal extra sweets or money. Mullen (1986) found that larger lynch mobs were associated with more savage killings, suggesting that group size and anonymity can intensify aggression.

AO3 evaluation of de-individuation

A strength is that de-individuation explains forms of aggression that happen in crowds, gangs, riots, online abuse, and masked groups. It has practical applications: increasing accountability through CCTV, visible names, moderation, or clear group rules may reduce aggression.

However, the theory does not always predict aggression. Johnson and Downing (1979) found that anonymity did not automatically increase aggression; behaviour depended on the group cues. Participants dressed in nurse-like clothing were less aggressive, suggesting that anonymity can increase conformity to pro-social norms too.

This links to social identity theory, especially work by Reicher et al. (1995). They argued that people in crowds do not simply “lose themselves”. Instead, they may shift from personal identity to group identity. Crowd behaviour depends on the norms of that group. This is an important challenge to the simple de-individuation explanation.

Methodologically, some evidence is correlational. For example, Mullen’s lynch mob research cannot prove that group size caused brutality; other factors may have been involved. Shock studies also raise ethical issues, including deception, potential distress, and the need for careful debriefing.

Bringing the explanations together

In essays, you can compare the explanations directly.

Frustration-aggression is strongest when aggression follows a blocked goal. Social learning theory is strongest when aggression has clearly been learned from models. De-individuation is strongest when aggression happens in anonymous group situations.

They can also work together. For example, a person may be frustrated after a defeat, have learned aggressive chants from peers, and then become de-individuated in a masked crowd. A good answer can acknowledge that human aggression is usually multi-causal.