Institutional Aggression

Social Learning Theory as a Foundation

Bandura's Social Learning Theory provides the theoretical basis for media effects on aggression: individuals learn aggressive behaviour vicariously by observing models in media. Learning is enhanced when the model is rewarded, attractive, or similar to the observer, and when the observer has the opportunity and motivation to perform the behaviour. Key processes include attention, retention, reproduction, and motivation.

Research Evidence

Television Violence — Huesmann et al. (2003)

Huesmann et al. conducted a longitudinal study following 329 participants from childhood (aged 6–10) into adulthood. Children's exposure to violent television in the 1970s significantly predicted aggressive behaviour in adulthood, even after controlling for childhood aggression, socioeconomic status, intellectual ability, and parenting. Correlations were r = 0.21 for males and r = 0.19 for females — modest but significant over a 15-year follow-up period.

Video Game Violence — Anderson et al. (2010)

Anderson et al. conducted a comprehensive meta-analysis including data from over 136,000 participants across 136 papers. They found consistent associations between violent video game exposure and increased aggression, increased aggressive thoughts, decreased empathy, and decreased prosocial behaviour. Longitudinal studies within the meta-analysis strengthened the causal inference.

Psychological Mechanisms

Desensitisation

Repeated exposure to media violence reduces physiological arousal to violence over time. Carnagey et al. (2007) found that participants who played a violent video game for 20 minutes showed significantly lower heart rate and galvanic skin response when subsequently shown footage of real violence, compared to those who played a non-violent game. Reduced emotional reactivity may lower the threshold for committing or tolerating aggression.

Cognitive Priming

Berkowitz (1984) proposed that exposure to violent media activates associated aggressive thoughts, feelings, and behavioural scripts — a network of associations in semantic memory. Once primed, aggressive schemas are more accessible and more likely to be applied when interpreting ambiguous social situations, increasing the probability of an aggressive response.

Evaluation

The longitudinal design of Huesmann et al. (2003) is a considerable strength: it tracks individuals over time and controls for confounding variables, making causal inference more defensible than cross-sectional studies. The meta-analytic scale of Anderson et al. (2010) increases statistical power and generalisability.

However, Ferguson (2015) re-analysed video game research and identified significant publication bias — studies reporting positive effects were more likely to be published. When corrected, effect sizes were substantially smaller and often non-significant. Furthermore, laboratory measures of aggression (e.g. noise blasts, electric shocks to confederates) lack ecological validity: these analogue behaviours may not translate to real-world violence. It is also difficult to isolate media exposure from other socialising influences — family, peers, and school.