Criminal Psychology — AQA GCSE Psychology Revision Notes

What you'll learn

• Why crime is defined socially rather than being a fixed category of behaviour
• Three ways of measuring crime, and the different blind spot each one has
• Biological explanations: Lombroso, Eysenck's criminal personality, and Raine's brain research
• The social learning explanation, and what it explains better than biology does
• Punishment, rehabilitation and why recidivism is the measure that matters
• Behaviour modification, anger management and restorative justice
• How to argue a nature–nurture question where both sides have real evidence

Key terms and definitions

Crime — behaviour that breaks the law of a particular society at a particular time.

Dark figure of crime — the gap between crime committed and crime officially recorded.

Victim survey — a survey asking members of the public about offences committed against them.

Offender survey — a survey asking people to report offences they have committed.

Atavistic form — Lombroso's idea that criminals are an evolutionary throwback identifiable by physical features.

Criminal personality — Eysenck's proposal that high extraversion and high neuroticism predispose someone to offending.

Social learning theory — the view that behaviour is acquired by observing and imitating others.

Model — a person whose behaviour is observed and imitated.

Vicarious reinforcement — seeing someone else rewarded, which makes imitation more likely.

Deterrence — discouraging offending through fear of punishment.

Recidivism — reoffending after a punishment or intervention.

Rehabilitation — an approach aiming to change behaviour so the offender does not reoffend.

Token economy — a scheme in which desirable behaviour earns tokens exchangeable for privileges.

Anger management — a programme teaching offenders to recognise and control their anger.

Restorative justice — a process in which the offender confronts the harm caused and agrees how to make amends.

Core concepts

What counts as a crime

Behaviour is criminal because the law of that society defines it so, and laws reflect that society's values, religion and history. An act legal in one country may carry a serious penalty in another.

Definitions also change within a country over time. As social attitudes shift, behaviour once prohibited may be legalised and behaviour once tolerated may be criminalised — so the same act can move in or out of the category of crime without the behaviour itself changing.

This matters more than it first appears. It means crime is socially defined rather than a fixed category of behaviour, and it complicates any comparison of crime rates across countries or decades.

Measuring crime

Three methods, each partial in a different way.

Official statistics record offences reported to and recorded by the police. They substantially understate the true amount. Many crimes are never reported — victims fear the consequences, think nothing will be done, or do not recognise what happened as a crime — and some reported offences are not recorded. The gap is called the dark figure of crime. Changes in public confidence or in recording practice can move the statistics without any real change in crime.

Victim surveys ask the public about offences committed against them, so they capture crimes never reported and reach part of the dark figure. They still depend on respondents remembering and being willing to disclose, and they miss crimes with no identifiable individual victim.

Offender surveys reach offences no victim reported, but rely on offenders admitting to crimes. Under-reporting is likely where the offence is serious or shameful, and some respondents exaggerate instead.

Because each method has a different blind spot, a change in the figures may reflect a change in reporting or recording rather than in crime.

Biological explanations

Lombroso proposed the atavistic form — the idea that criminals are a throwback to an earlier stage of human evolution, identifiable by physical features.

The theory is rejected, but its place in the syllabus is instructive. Lombroso was the first to attempt a systematic, evidence-based study of offenders rather than treating crime as simple wickedness, and he established that criminality might be investigated scientifically at all. His method, however, had a fatal flaw: he studied only offenders and compared their features with no control group of non-offenders, so he could not know whether those features were any more common among criminals. His conclusions also linked criminality to physical and racial characteristics in ways later used to justify discrimination, and he ignored social circumstances entirely.

Eysenck's criminal personality proposes that high extraversion combined with high neuroticism predisposes someone to offending — a person who seeks stimulation and conditions poorly to punishment, so learns less well from being disciplined. He later added psychoticism.

It is testable and measurable by questionnaire, and it proposes a mechanism — poor conditioning — rather than merely describing traits. But the questionnaire relies on self-report, most people scoring highly on those traits never offend, a single criminal personality cannot cover offences as different as fraud and assault, and the evidence is correlational.

Raine and colleagues scanned the brains of people who had committed murder and found reduced activity in areas of the prefrontal cortex associated with controlling impulses. This uses objective measurement rather than self-report and identifies differences in regions plausibly linked to impulse control. But the design is correlational, so the difference may be a consequence of a violent life, head injury or substance use rather than its cause. The sample consisted of offenders with particular legal circumstances, so may not represent violent offenders generally. And no scan can distinguish someone who will offend from someone who will not.

Taken together, biological explanations have real evidence and treatment implications, but they cannot explain why crime rates differ sharply between societies with similar populations, and they raise troubling questions about responsibility and about identifying people as criminals before any offence.

The social learning explanation

Criminal behaviour is learned by observing and imitating others. A child observes a model — a parent, sibling or peer — offending and seeing it rewarded with money or status, and imitates the behaviour. Vicarious reinforcement means the observer need not be rewarded themselves; seeing someone else rewarded is enough.

Bandura's Bobo doll research is the standard supporting evidence for imitation of aggression.

This explains why offending clusters in families and neighbourhoods and why it often begins in adolescence among peers, and it suggests practical intervention through changing the available models and rewards. Against it: much supporting evidence comes from laboratory studies of aggression towards a doll rather than real crime; it cannot explain offenders who had no criminal models; and it says little about why some observers imitate while others exposed to the same models do not.

Note the overlap problem with the genetic account. Crime running in families is consistent with shared genes and with shared models — the family provides both, which is why the two explanations are so hard to separate.

Dealing with offending

Punishment protects the public during the sentence, expresses society's condemnation and may deter some people. But high recidivism rates show it frequently fails to change behaviour. Prison can damage the employment and relationships that support going straight, and it exposes offenders to criminal models — so release can leave them more likely to offend rather than less. Offenders may also not weigh the risk beforehand at all, which undermines deterrence as a mechanism.

Rehabilitation addresses the causes and has better evidence for reducing reoffending, but requires the offender's cooperation, takes longer, and can appear to the public and to victims as an insufficient response to serious harm.

Recidivism is the measure that settles this. It shows how many offenders go on to reoffend, which tests whether a punishment achieved anything beyond the period of the sentence itself. A punishment that satisfies public opinion while leaving recidivism unchanged has not reduced crime.

Token economies apply operant conditioning inside institutions: desirable behaviour is immediately rewarded with tokens exchangeable for privileges, so that behaviour is reinforced. They are cheap, easy to administer and demonstrably improve order and cooperation within the institution, with no need for the prisoner to gain any insight. But the behaviour is tied to the reinforcement, so it typically stops when the tokens do and rarely transfers to life after release. It also produces compliance rather than genuine change, and gives staff considerable power over privileges.

Anger management teaches offenders to recognise the triggers and early signs of their anger and to practise alternative responses, treating the offending as a skills deficit that can be addressed.

Restorative justice brings offender and victim into contact so the offender understands the harm caused and agrees how to make amends. It confronts the offender with the actual human consequence of what they did, which conventional sentencing rarely achieves, gives victims a voice and an explanation they otherwise never receive, and has evidence of reducing reoffending and improving victim satisfaction. But it depends on a willing victim and a genuinely remorseful offender, a badly handled meeting can retraumatise the victim, it is unsuitable for some offences, and it can be perceived as too lenient for serious crime.

Nature and nurture in criminality

Social factors — poverty, poor education, family disruption, criminal models — correlate strongly with offending and explain why rates differ so much between societies, which a purely biological account cannot. But the biological evidence from twin studies and brain imaging is real.

The two interact. An inherited tendency towards impulsivity may only produce offending given the right circumstances. The useful question is how they combine rather than which one wins, and an answer that says so will outscore one that picks a side.

Worked examples

Example 1: Explaining a measurement problem

Explain why official crime statistics may understate the amount of crime. (3 marks)

Many offences are never reported to the police, because victims fear the consequences or believe nothing will be done (1). Some offences that are reported are not formally recorded (1). The difference between crime committed and crime recorded is known as the dark figure of crime (1).

Example 2: Identifying a methodological flaw

Explain one weakness of Lombroso's method. (2 marks)

He examined only offenders and had no control group of non-offenders for comparison (1), so he could not establish whether the physical features he identified were any more common among criminals than in the general population (1).

Example 3: Applying social learning theory

Dan's older brother shoplifts and is admired by his friends for it. Dan begins shoplifting too. Explain Dan's behaviour using social learning theory. (4 marks)

Dan's brother acts as a model whose behaviour Dan observes (1). Dan sees the shoplifting rewarded with goods and with admiration from friends (1). This is vicarious reinforcement: Dan does not need to be rewarded himself, because seeing his brother rewarded makes imitation more likely (1). Dan therefore imitates the behaviour (1).

Example 4: Evaluating an intervention

Evaluate the use of token economies in prisons. (6 marks)

Token economies are cheap and easy for staff to administer, and they demonstrably improve order and cooperation within the institution without requiring the prisoner to gain any insight (2). However, the behaviour is tied to the reinforcement, so it typically ceases once the tokens stop and rarely transfers to life after release (2). This means the scheme may produce compliance within prison rather than genuine change, which is why recidivism rates are the proper test of whether it has achieved anything (2).

Common mistakes and how to avoid them

Treating crime as a fixed category. Crime is what the law of a society defines at that time. Saying so is often the first mark.

Saying official statistics are "wrong". They are incomplete in a specific direction — they understate. Explain the mechanism.

Claiming Lombroso's theory is simply stupid. Give the credit (the first systematic study) and then the flaw (no control group). Balanced answers score higher.

Saying Raine proved brain damage causes murder. The study is correlational. The difference may be a consequence of a violent life.

Using Bobo doll research as direct evidence about crime. It shows imitation of aggression towards a doll, not real offending. Note the gap.

Confusing deterrence with rehabilitation. Deterrence works through fear of punishment; rehabilitation works by changing the person.

Judging punishment by severity rather than by outcome. Recidivism is the measure.

Forgetting that restorative justice needs a willing victim. It is not universally applicable, and saying so is an evaluation mark.

Picking a side on nature versus nurture. The interaction position is what the evidence supports.

Exam technique for Criminal Psychology

Structure evaluation as credit, then flaw, then consequence. Every explanation in this topic has a genuine strength and a specific weakness. "No control group" for Lombroso, "correlational" for Raine, "laboratory aggression not real crime" for social learning, "most high scorers never offend" for Eysenck. Learn one per explanation and you can evaluate anything here.

Name the study and what it found, briefly. Lombroso, Eysenck, Raine, Bandura. One sentence each is enough — the marks are for what they show.

For measurement questions, use more than one method. The strongest answers point out that each method has a different blind spot, so no single source gives the true figure.

Link the explanation to the intervention. Biological explanation → medication or containment. Social learning → changing models. Behaviourist → token economy. Cognitive → anger management. Examiners reward seeing the connection.

Be careful with sensitive material. Write about people who have offended rather than using dehumanising language, and avoid implying that any group is inherently criminal — a point Lombroso's history makes directly relevant.

On ethics, remember the prediction problem. Biological explanations raise the prospect of identifying people as criminals before any offence, which is a strong evaluation point about the implications of the research rather than its method.

Quick revision summary

• Crime is defined by the law of a society at a time, so it varies between cultures and across decades
• Official statistics understate crime; the gap is the dark figure
• Victim surveys reach unreported crime; offender surveys reach undetected crime; each has its own blind spot
• Lombroso's atavistic form: first systematic study, but no control group and discriminatory conclusions
• Eysenck: high extraversion plus high neuroticism, with poor conditioning to punishment as the mechanism
• Raine: reduced prefrontal activity in murderers — correlational, so cause is not established
• Social learning: models, imitation and vicarious reinforcement explain family and neighbourhood clustering
• Bobo doll evidence concerns aggression towards a doll, not real offending
• Punishment protects and condemns but often fails to change behaviour; prison exposes offenders to criminal models
• Rehabilitation has better evidence on reoffending but needs cooperation and looks lenient
• Recidivism is the measure that tests whether any intervention worked
• Token economies improve order inside prison but rarely transfer after release
• Anger management treats offending as a skills deficit
• Restorative justice repairs harm and reduces reoffending, but needs a willing victim
• Nature and nurture interact: inherited impulsivity may only produce offending in certain circumstances