You are viewing a free preview of this lesson.
Subscribe to unlock all 10 lessons in this course and every other course on LearningBro.
Before criminological psychology can explain why people offend, treat offenders or advise the courts, it must first settle two deceptively simple questions: what counts as a crime, and how much crime is there? Neither is as straightforward as it appears. A "crime" is not a natural fact about a behaviour in the way that mass or temperature are natural facts about an object; it is a social and legal construction that varies across cultures, changes over time, and depends on who is doing the defining. Measuring crime is equally treacherous, because the vast majority of offences never appear in any official record — the so-called "dark figure" of unrecorded crime. This lesson examines the legal and social ways crime is defined, shows how definitions shift across culture and history, and then critically compares the three principal ways psychologists and criminologists measure crime: official statistics, victim surveys and self-report studies. Understanding these foundations matters because every subsequent topic in criminological psychology — explanations of offending, eyewitness testimony, offender profiling, treatment — rests on data whose reliability and validity are shaped by exactly these definitional and measurement problems.
Key Definition: Crime is any act that violates the criminal law of a society and is punishable by the state. Because the criminal law differs between jurisdictions and changes over time, what counts as a crime is not fixed but socially and legally defined.
This lesson addresses the introductory content of Edexcel 9PS0 — Paper 2, Topic 6: Criminological Psychology, specifically the requirement to understand how crime is defined (including the problems of defining crime and how definitions vary across cultures and over time) and how crime is measured (official statistics, victim surveys and self-report studies, with the strengths and weaknesses of each).
Connects to…
The most widely used definition of crime is legal: a crime is an act (or omission) that breaks the criminal law of a jurisdiction and is punishable by the state. On this view, a behaviour is criminal if and only if the law says so. This definition has clear practical advantages — it is precise, it is written down, and it tells police, courts and citizens exactly where the boundaries lie. It also underpins the principle of the rule of law: people can only be punished for acts that were already defined as criminal at the time they were committed.
The legal definition rests on what sociologists call a consensus view — the assumption that there is broad social agreement that certain acts (murder, theft, assault) are wrong and should be prohibited. For the most serious offences, this consensus is genuine and near-universal: no known society treats the arbitrary killing of its own members as acceptable.
Key Definition: The consensus view holds that laws express a shared moral agreement across society about which behaviours are unacceptable. The conflict view counters that laws reflect the interests of powerful groups and are used to control the less powerful.
A purely legal definition, however, has an obvious limitation: it makes crime entirely dependent on the current state of the statute book. Many behaviours that most people regard as seriously harmful (some forms of exploitation, environmental damage, or morally objectionable but technically lawful conduct) are not crimes, while some behaviours that harm no one are. This motivates a social definition of crime, which focuses on acts that violate the norms and values of a society and provoke social condemnation, whether or not they are formally illegal. The related concept of deviance captures behaviour that departs from social norms; crime is best understood as the subset of deviance that the law has chosen to prohibit.
The two definitions frequently diverge, and mapping that divergence is a useful analytic exercise:
| Behaviour | Legally criminal? | Socially deviant? |
|---|---|---|
| Murder | Yes | Yes (strong consensus) |
| Minor speeding | Yes | Often not condemned |
| Adultery (in England & Wales) | No | Widely disapproved of |
| Queue-jumping | No | Disapproved of, not illegal |
| Assisted dying (in some jurisdictions) | Varies by jurisdiction | Contested |
The conflict view, associated with Marxist criminology, pushes the social definition further: it argues that the criminal law is not a neutral expression of consensus but a tool by which powerful groups protect their interests. On this reading, the acts most heavily policed and punished tend to be those committed by the less powerful (street theft, public-order offences), while the harms caused by the powerful (some corporate and financial misconduct) are less likely to be criminalised or prosecuted. This is not a claim that has to be accepted wholesale, but it usefully exposes the assumption of neutral consensus built into the legal definition.
graph TD
A[Human behaviour] --> B{Departs from social norms?}
B -->|No| C[Conformity]
B -->|Yes| D[Deviance]
D --> E{Prohibited by criminal law?}
E -->|No| F[Deviant but not criminal<br/>e.g. queue-jumping]
E -->|Yes| G[Crime]
style G fill:#e74c3c,color:#fff
style C fill:#27ae60,color:#fff
style F fill:#f39c12,color:#fff
Because crime is legally and socially defined, what counts as criminal varies markedly between cultures. This cultural relativity operates in several ways:
For criminological psychology, cultural variability creates a serious problem for cross-cultural comparison. Comparing the "crime rate" of two countries assumes the two are counting the same thing, but if the underlying definitions and recording practices differ, the comparison may be close to meaningless. This is a form of imposed etic — applying a category derived from one culture as though it were universal.
Definitions also shift historically within a single society. Acts once criminal are decriminalised, and acts once lawful are criminalised, as social attitudes and political priorities change. Within England and Wales alone, the twentieth and twenty-first centuries saw the decriminalisation of behaviours that were previously serious offences, alongside the creation of entirely new categories of crime — most obviously the raft of cybercrime offences (computer misuse, online fraud, digital harassment) that could not have existed before the relevant technology. The category of crime is therefore a moving target: a longitudinal graph of "total crime" over a century is partly a graph of changing definitions, not simply of changing behaviour.
This historical variability has a direct methodological consequence, examined below: because the law changes, apparent rises or falls in recorded crime may reflect the stroke of a legislative pen (a new offence created, an old one repealed, a recording rule changed) rather than any real change in how people behave.
Given that crime is definitionally slippery, measuring it is correspondingly difficult. Three principal methods are used, each capturing a different slice of the total picture. A guiding idea runs through all three: the amount of crime that is committed is far larger than the amount that is recorded, and the gap between them is the dark figure.
Key Definition: The dark figure of crime is the quantity of crime that is committed but never recorded in official statistics — because it is not witnessed, not reported, not recorded by the police, or not defined as a crime at the point of contact.
Official statistics are the counts of crime compiled by government from records generated by the criminal justice system — principally crimes recorded by the police, but also data from courts and prosecutions. In England and Wales these "police recorded crime" figures are published regularly and are the headline numbers most often quoted in the media.
Their strengths are practical and considerable: they are cheap to produce (they reuse data the system already generates), they cover the whole population rather than a sample, they are collected continuously so trends can be tracked over years, and they are readily available for research. For the most serious offences — homicide in particular — they are also reasonably complete, because such crimes are almost always reported and recorded.
Their weaknesses, however, are severe and systematic. An offence only enters the statistics if it survives a chain of filters, and at each stage crime "leaks out" of the count:
graph LR
A[Crime committed] --> B{Detected /<br/>witnessed?}
B -->|No| Z[Dark figure]
B -->|Yes| C{Reported<br/>to police?}
C -->|No| Z
C -->|Yes| D{Recorded<br/>by police?}
D -->|No| Z
D -->|Yes| E[Appears in<br/>official statistics]
style Z fill:#34495e,color:#fff
style E fill:#27ae60,color:#fff
The upshot is that official statistics measure recorded crime, which is a biased and incomplete sample of actual crime. They are best read as a measure of the activity of the criminal justice system as much as of crime itself.
A victim survey (or victimisation survey) asks a large, representative sample of the population whether they have been a victim of specified crimes over a set period, regardless of whether they reported those crimes. The best-known example is the large national crime survey conducted in England and Wales, which interviews tens of thousands of households each year. Crucially, victim surveys capture crimes that were never reported to the police, so they reach into part of the dark figure that official statistics miss.
The strengths of victim surveys follow from this design. Because they do not depend on reporting or police recording, they provide a fuller and often more stable picture of crime and its long-term trends, and they are less vulnerable to the recording-practice distortions that plague official statistics. They also yield rich detail about victims, circumstances and the impact of crime, and — by asking whether each incident was reported — they allow the reporting rate itself, and hence the size of the dark figure for different offences, to be estimated.
Their weaknesses, however, are real:
A self-report study takes the opposite perspective: instead of asking about victimisation, it asks respondents (often anonymously) whether they have committed particular offences, whether or not they were caught. This method is invaluable for research because it can reveal offending that never came to official attention at all, and — by asking about the respondent's own characteristics — it allows offending to be related to variables such as age, background or personality.
The strengths are that self-report studies can access hidden offending (including minor and juvenile offences that rarely reach the courts), they illuminate the characteristics of offenders rather than just the count of offences, and they are indispensable to developmental and longitudinal criminology (the Cambridge Study in Delinquent Development, discussed in later lessons, relied heavily on self-report alongside official records).
The weaknesses are, however, acute:
Because each method reaches a different part of the total picture, they are best understood as complementary rather than competing. Triangulating across all three gives a more trustworthy estimate of crime than any one alone.
| Feature | Official statistics | Victim surveys | Self-report studies |
|---|---|---|---|
| Perspective | The state's record | The victim's experience | The offender's admission |
| Reaches the dark figure? | Least (recorded crime only) | Partly (unreported victimisation) | Partly (unrecorded offending) |
| Coverage | Whole population | Large representative sample | Often convenience sample |
| Main threat to validity | Under-reporting & under-recording | Memory, telescoping, sensitive disclosure | Concealment / exaggeration |
| Best at capturing | Serious, always-reported crime (homicide) | Household/personal crime incl. unreported | Hidden minor & juvenile offending |
| Weakest at capturing | Unreported & unrecorded crime | Victimless / unaware-victim crime | Serious/persistent offenders |
No single method measures "crime itself" — each measures a construct shaped by its own procedure. Official statistics measure recorded crime; victim surveys measure reported-in-a-survey victimisation; self-report studies measure admitted offending. None of these is identical to the underlying quantity of criminal behaviour, and the differences between them are systematic rather than random. The implication is that any claim about "the crime rate" is only as good as the method behind it, and a psychologically literate reading of crime data must always ask which method produced the figure and what it therefore leaves out.
Official statistics have high coverage but low validity for most offences. Their whole-population, continuous, low-cost nature makes them irreplaceable for tracking long-run trends and for serious crimes such as homicide that are almost always recorded. But for the great bulk of offending, the reporting-and-recording filter introduces large, non-random bias: crimes that victims are ashamed to report, or that the police under-record, are systematically undercounted. The implication is that official statistics are a valid measure of the criminal-justice system's activity and of a few always-reported crimes, but a poor and biased measure of total offending — which is precisely why they cannot be used alone.
Victim surveys recover part of the dark figure but cannot recover all of it. By bypassing the reporting stage, they capture crimes official statistics miss and give a more stable picture of trends, and their design allows the reporting rate itself to be estimated — a genuine methodological advance. However, they inherit the frailties of memory (forgetting, telescoping) and self-disclosure (reluctance over sensitive victimisation), and they structurally exclude victimless crimes and crimes against the deceased or the uncounted. The implication is that victim surveys are the strongest single measure of household and personal crime and its trends, but they are not a universal solution and must be read alongside the other methods for offences they cannot see.
Self-report studies are uniquely able to relate offending to offender characteristics, which is why they are indispensable to psychology. For a discipline interested in who offends and why, a method that links admitted offending to age, personality and background is invaluable, and it reaches minor and juvenile offending that never troubles the courts. But its validity turns entirely on respondent honesty, which is threatened by concealment in some samples and exaggeration in others, and its typical reliance on convenient (often young, non-serious) samples limits what it can say about persistent or serious offenders. The implication is that self-report is powerful for studying the correlates of common offending — hence its central role in developmental criminology — but weak as a measure of the volume of serious crime.
Definitional variability undermines cross-cultural and historical comparison across all three methods. Because crime is legally and socially defined, and because those definitions differ between cultures and shift over time, comparing crime figures across countries or across decades risks comparing different things. A rise in "recorded crime" may reflect a new offence, a changed counting rule, or greater willingness to report, rather than more offending. The implication is that longitudinal and cross-cultural crime comparisons demand great caution: the safest inferences come from within a stable definitional and methodological frame, or from triangulating multiple methods that are unlikely to share the same bias.
The most defensible position is methodological triangulation. Because the biases of the three methods largely differ — official statistics understate through non-reporting, victim surveys miss victimless crime, self-report is limited to co-operative and often minor samples — convergence between them substantially increases confidence, and divergence flags exactly where the dark figure is largest. The implication is that criminological psychology should treat crime measurement not as a single authoritative number but as a triangulated estimate, reading each source for what it does well and cross-checking it against the others.
Assess the strengths and weaknesses of official statistics and self-report studies as ways of measuring crime. (12 marks)
Specimen question modelled on the Edexcel 9PS0 paper format. This extended-response question carries approximately 4 marks for AO1 (accurate knowledge of what each method is and how it works) and 8 marks for AO3 (evaluation of their strengths and weaknesses, ideally comparatively). There is no AO2 here unless a scenario stem is added. AO1 should define official statistics (police-recorded crime) and self-report studies (anonymous admission of offending), each with its procedure. AO3 should weigh coverage against validity for official statistics (whole-population and good for homicide, but under-reporting/under-recording and the dark figure) and the access-to-hidden-offending strength of self-report against its honesty and sampling weaknesses, drawing an explicit comparison and, ideally, concluding on triangulation.
Mid-band response (6/12):
Official statistics are the numbers of crimes recorded by the police and published by the government. A strength is that they cover the whole population and are cheap because the data already exists. A weakness is the dark figure — lots of crime is never reported or recorded, so the statistics are too low. People do not report crimes if they are embarrassed or think the police cannot help.
Self-report studies ask people if they have committed crimes, usually anonymously. A strength is that they find crimes the police do not know about, including minor crimes by young people. A weakness is that people might lie — they might hide serious crimes or exaggerate small ones — so the data may not be accurate. Overall both methods have strengths and weaknesses, so it is best to use more than one.
Examiner-style commentary: This earns the AO1 marks for correctly identifying both methods and their basic procedures, and it makes valid evaluative points (the dark figure; honesty problems). It sits mid-band because the evaluation is asserted rather than developed — each point stops at what the weakness is without explaining why it biases the data or what follows. To reach the next band the answer needs point–evidence–explanation–implication development (e.g. spelling out that under-reporting is systematic, not random, so official statistics are a biased sample) and an explicit comparison of the two methods rather than treating them separately.
Stronger response (9/12):
Official statistics are counts of police-recorded crime compiled by government. Their strength is coverage: they include the whole population, are collected continuously so trends can be tracked, and are near-complete for serious crimes like homicide that are almost always reported. Their central weakness is the dark figure: crime only enters the statistics if it is witnessed, reported and then recorded, and at each stage crime leaks out. Under-reporting (embarrassment, distrust, fear of reprisal) and under-recording (police discretion, changing rules) mean the figures are a biased sample of actual crime and partly measure the activity of the justice system rather than crime itself.
Self-report studies ask respondents, usually anonymously, whether they have committed offences whether or not they were caught. Their strength is that they reach hidden offending the police never see, especially minor and juvenile crime, and they relate offending to offender characteristics, which is why psychology relies on them. Their weakness is validity: respondents may conceal serious offences or exaggerate minor ones, and convenience samples (often students) rarely reach serious, persistent offenders. Comparing the two, their biases differ — official statistics understate through non-reporting, self-report is limited by honesty and sampling — so the safest approach is triangulation, using each to check the other.
Examiner-style commentary: This is a strong, well-organised answer that earns most of the AO1 and much of the AO3. Each method is defined precisely, the strengths and weaknesses are explained (why the filter biases official statistics; why concealment and sampling limit self-report), and there is a genuine comparison plus a triangulation conclusion. To reach top-band it would add a discriminating point that shows deeper understanding — for example, noting that official statistics are best read as a measure of criminal-justice activity, or that self-report's real strength is the study of the correlates of offending rather than its volume, and weaving the "different biases" argument through the whole answer rather than only at the end.
Top-band response (12/12):
The two methods measure different constructs, and a strong assessment turns on that distinction. Official statistics measure recorded crime — police-recorded offences published by government. Their strengths are coverage and continuity: whole-population data, collected continuously, cheap because it reuses existing records, and near-complete for homicide and other always-reported crimes, making them irreplaceable for long-run trends in serious offending. Their decisive weakness is the dark figure. An offence enters the statistics only if it is detected, reported and recorded, and at each filter crime leaks out through under-reporting (shame, distrust, fear, triviality) and under-recording (discretion, reclassification, changing counting rules that can shift the figures with no change in behaviour). The bias is therefore systematic, so official statistics function partly as a measure of criminal-justice activity, not offending as such.
Self-report studies invert the perspective, asking respondents to admit offences anonymously regardless of detection. Their strength is uniquely valuable to psychology: they access hidden and juvenile offending and, by linking admissions to age, personality and background, illuminate the correlates of offending — hence their centrality to developmental work such as the Cambridge Study. Their weakness is validity, threatened in opposite directions by concealment of serious offences and exaggeration of minor ones, and by convenience samples that under-reach persistent offenders. The comparison is the key evaluative move: because the two methods have different biases — official statistics understate volume through non-reporting; self-report captures correlates but not serious-crime volume — their convergence raises confidence and their divergence maps the dark figure. The defensible conclusion is neither method alone but triangulation, ideally with victim-survey data, each read for what it measures well.
Examiner-style commentary: This answer shows the discriminators examiners reward. It frames a thesis (the methods measure different constructs) and sustains it; it distinguishes recorded crime from actual crime and self-report correlates from volume; it explains that biases are systematic and in opposing directions; and it reaches an integrative triangulation conclusion rather than a list of pros and cons. The register is analytic throughout and every evaluative point carries an explicit implication.
This content is aligned with the Edexcel A-Level Psychology (9PS0) specification.