The Gold Standard of Criminology

A broad question such as why crime happens contains many different questions. Why did recorded reports increase in one place? Why are some people repeatedly harmed? Does a particular prevention programme change an outcome? Each question requires a different design and a different kind of evidence.

State the population, place, period and outcome. A study of reported bicycle theft near a station is narrower than a claim about all offending in a city. Narrowing the question makes it possible to examine whether the evidence actually fits.

The crime, justice, policing and criminology research guide offers a route into finding relevant material. Start with a short question and a list of terms that need definitions. This prevents a search from becoming a collection of dramatic examples that all appear to confirm the first idea.

A useful learner can revise the question after discovering what records exist. Revision is part of inquiry when it makes the relationship between question and evidence more honest. Keep a note of the change so that an exploratory finding is not later presented as the result of a test planned from the beginning.

Contents · Next chapter

Crime refers to conduct defined as criminal within an applicable legal framework. Harm is a wider idea. Some harmful conduct may be addressed through other legal or institutional routes, while the boundaries of criminal offences vary across jurisdictions and time.

Keep these concepts separate when comparing places. A difference in recorded categories may reflect different legal definitions or recording practices. It cannot automatically be interpreted as a difference in the underlying behaviour. The Law root develops the relationship between rules, authority, evidence and procedure.

A researcher can ask how a category was created, what conduct it includes and whose experiences may fall outside it. Those questions do not require dismissing legal definitions. They make the limits of the category visible before it is used to explain a social pattern.

For learning, compare two fictional descriptions of the same behaviour under differently defined study categories. Ask which conclusions remain comparable and which need qualification. The result should be a precise statement about the study, rather than an unsupported claim that one society is inherently more lawful or more harmful.

Previous chapter · Contents · Next chapter

Explaining the conditions associated with offending is different from determining legal responsibility in an individual case. A population-level relationship does not establish that a particular person committed an act. Nor does investigating a mechanism automatically excuse conduct or remove concern for those harmed.

A useful explanation proposes how something might happen. It then identifies observations that would support, weaken or distinguish that explanation from alternatives. If every possible result is treated as confirmation, the explanation becomes difficult to investigate.

The Justice and Society roots provide connected routes. Criminology adds questions about patterns, mechanisms and the effects of responses. Its contribution becomes stronger when the level of the claim remains clear.

Ask learners to label three statements: an observation, a proposed explanation and a decision requiring legal authority. They may concern the same event while doing different work. This simple exercise prevents statistics, public opinion and procedural findings from being blended into a single apparently certain account.

Previous chapter · Contents · Next chapter

Police reports, court records and correctional records describe different stages and activities. Their units may be incidents, allegations, offences, cases or people. A person can appear in more than one record, and several events can be combined or separated under the relevant recording rules.

Before using a total, inspect the definition, coverage and point in the process at which the record is created. A report is not the same thing as a conviction. A case closed for an administrative reason should not silently be described as proof that an allegation was false.

The Data root develops the importance of definitions and provenance. In crime research, the path from an experience to an institutional record is especially important. Reporting access, recording practices and institutional priorities can influence the resulting dataset.

The UNODC Data Portal offers routes into international crime and justice indicators. Read the metadata alongside any number you use. The portal is a research starting point, not permission to compare every available country total as though the same observation process produced them.

Previous chapter · Contents · Next chapter

Surveys can ask people about experiences that may not appear in police records. Their value depends on the population reached, questions asked, reference period, sampling and handling of non-response. A survey also has boundaries: it does not capture every offence, person or setting equally.

Administrative records and surveys therefore provide different views. Agreement can be informative, while disagreement can generate a useful question about reporting or coverage. Neither source should automatically be treated as a complete count of all harm.

UNODC’s report on regional victimisation-survey work explains how such surveys complement national crime and criminal justice statistics by making unreported experiences more visible. For a real study, examine its questionnaire and technical documentation rather than relying on the general label victimisation survey.

In a classroom exercise, invent two evidence files about the same fictional town: reported incidents and responses from a defined household sample. Ask what each file establishes and who might be missing. Keep the example numerical and impersonal. Students do not need to disclose distressing personal experiences to learn how survey coverage affects interpretation.

Previous chapter · Contents · Next chapter

A count describes how many observations were recorded. A rate relates observations to a defined population or exposure. The appropriate denominator depends on the question. Comparing totals between places of very different sizes can answer a different question from comparing rates.

Suppose two fictional districts record 40 and 60 incidents during a year. Their resident populations are 2,000 and 6,000. The corresponding figures are 20 and 10 incidents per thousand residents. The district with the larger count has the smaller population-based rate in this simplified example.

That arithmetic does not settle every comparison. A busy commercial district may have many visitors, and the population at risk may differ from the resident count. State the chosen denominator and explain its relevance. Do not quietly switch between people, households and journeys.

The Statistics root develops this reasoning. A good learner names the quantity before calculating, checks compatible periods and avoids describing an area-level rate as a prediction of what will happen to every individual living there. The numerical result should remain attached to the definition that produced it.

Previous chapter · Contents · Next chapter

An increase in recorded incidents can reflect several processes: more underlying events, greater willingness to report, a new reporting channel, changed definitions or altered recording practices. The observed increase is important, but its cause needs further investigation.

Construct a timeline containing relevant institutional and measurement changes alongside the series. Check whether the category, geographic boundary and unit remain compatible. An abrupt break following a recording change deserves attention before it is interpreted as a sudden social transformation.

Consider a fictional online reporting form introduced halfway through a year. Reports rise afterward. One possible explanation is easier access to reporting. Another is an increase in events. More evidence is needed to distinguish their contributions.

Ask the learner to write an interim conclusion that preserves the observation without claiming a cause. Then list suitable next evidence, such as compatible survey findings or documentation of recording changes. This is a practical form of intellectual control: the learner can say something useful while leaving the unresolved part genuinely unresolved.

Previous chapter · Contents · Next chapter

Crime maps organise observations spatially. Their appearance depends on the measure, geographic units, reporting coverage and classification choices. A cluster of recorded events may identify a useful question, but it does not describe every person or activity in the area.

Consider exposure and opportunity. A location with many visitors can generate a larger count than a quiet residential area. A map of totals and a map of suitably defined rates may show different patterns. Neither should be presented without explaining what was measured.

Protect people when publishing detailed locations. Points can reveal homes, routines or sensitive experiences even when names have been removed. Use an appropriate level of aggregation and legitimate review for the actual purpose.

A learning exercise can use invented places and compare two different boundaries around the same observations. Ask how the visible pattern changes and what claims remain defensible. The point is to understand the representation, then choose a responsible interpretation. A coloured area is evidence about a defined dataset, rather than a moral identity assigned to a neighbourhood.

Previous chapter · Contents · Next chapter

Criminological theories emphasise different relationships: individual decisions, opportunities, social connections, institutions, learning, inequality or the way conduct is classified and controlled. A root introduction should show what a theory helps someone examine before asking them to memorise its label.

Take a simple fictional question about theft from unattended bicycles. One explanation may focus on access and opportunity; another may examine routines or the surrounding social setting. Each needs a proposed mechanism and evidence appropriate to the claim.

Different explanations can contribute at different levels. A situational change may affect one location without explaining the entire distribution of offending across society. Avoid making a theory answer every question merely because it has a memorable name.

Have learners write a short chain: proposed condition, mechanism, expected observation and possible alternative explanation. Then identify what evidence could challenge the chain. This keeps theoretical learning connected to investigation. A strong discussion can compare explanations without pretending that every disagreement has already been resolved by one correlation or one striking example.

Previous chapter · Contents · Next chapter

Two variables can move together without one causing the other. A third factor, selection process or measurement practice may influence both. The order in time also matters: a proposed cause must be examined in relation to when the relevant change occurred.

Suppose a fictional service opens in an area with many recorded incidents. A later comparison finds more incidents near the service than elsewhere. That association does not establish that the service caused them. It may have been placed there precisely because the need was already greater.

A credible causal argument explains the comparison, possible confounding and why the chosen design can address the question. Randomised studies, natural experiments and observational analyses each have conditions and limitations. A label alone does not guarantee a sound result.

Connect this chapter with Statistics and the research guide. Ask students to propose two alternative explanations for a pattern before defending a preferred account. The exercise develops restraint without making inquiry impossible: claims can become stronger through better design, clearer assumptions and evidence that distinguishes plausible alternatives.

Previous chapter · Contents · Next chapter

A prevention programme needs a defined purpose, an intended mechanism and an outcome that can be assessed. Counting sessions, posters or patrol hours describes activity. It does not by itself establish a reduction in harm or an improvement in people’s ability to seek help.

Consider who receives the intervention, what comparison is suitable and how long effects are observed. A change may shift incidents to another place, time or method. It may also have benefits or costs that the primary indicator misses.

For a fictional bicycle-security programme, distinguish the number of locks distributed from the experiences of the intended users and the relevant theft outcome. Record implementation: a programme cannot be fairly evaluated as delivered if its essential parts never reached the participants.

An honest evaluation can identify a useful improvement while acknowledging uncertainty about causation. Ask what should happen next: continue, adapt, collect better evidence or stop an ineffective activity. The standard is an accountable decision supported by the results, rather than a retrospective success story built only from the most favourable indicator.

Previous chapter · Contents · Next chapter

Research about crime and harm can involve sensitive information, distress and unequal power. Ethical planning should address the actual study: what information is necessary, how participation works, how confidentiality is handled and what support or safeguarding arrangements apply.

Use appropriate institutional review and qualified supervision where required. A generic promise of complete secrecy can be misleading when specific legal or safeguarding duties apply. Participants need an accurate explanation of the relevant limits through a suitable process.

The Data root connects collection with purpose, access and retention. Research teams should also examine whether their methods create avoidable burdens or make it difficult for some people to participate.

For school learning, use public material and fictional datasets rather than inviting disclosures about identifiable incidents. Students can assess the quality of a research plan without collecting sensitive stories from classmates. This preserves the educational objective while teaching a central responsibility: a valuable research question does not automatically justify every available method of obtaining an answer.

Previous chapter · Contents · Next chapter

A study can become absorbed in categories and institutions while losing sight of the person affected. Their concerns may include safety, information, practical support, privacy, participation or understanding what happens next. These concerns should not be reduced to whether an institution produced a record.

Support requires appropriate expertise and current local routes. The Cybercrime Victim Support branch develops one particular context, including reporting, evidence and recovery. Its specialised advice should be read in relation to the situation and the responsible authorities.

A researcher should distinguish the person’s expressed experience from assumptions made on their behalf. Different people can want different forms of participation or support. A single satisfaction score may not capture whether an essential need remains unresolved.

Use a fictional service journey to examine the difference between referral and useful arrival. Was information understandable? Did the receiving service acknowledge the handoff? Was the person told the next step? These are concrete questions about a response. They help connect institutional activity with the outcome that matters to the person.

Previous chapter · Contents · Next chapter

Forensic work can contribute observations and interpretations relevant to an investigation. Its value depends on suitable methods, traceable handling, appropriate comparison and a clear account of uncertainty. A technical result must remain attached to the question the method can address.

The Forensic Scientists branch develops these responsibilities. The Digital Evidence Chain of Custody article focuses on the handling and traceability of digital material. These routes complement criminology without turning a broad social explanation into a forensic conclusion.

Distinguish evidence about a material or association from a complete account of how an event happened. Context, alternative explanations and the relevant legal process still matter. A scientific-sounding statement should not acquire greater scope merely because a reader finds the technical language impressive.

In a fictional classroom file, label the observation, the method’s stated limitation and the conclusion someone wishes to draw. Ask whether the last exceeds the first two. The aim is to develop careful reading, rather than teach students to conduct their own investigation of real people or adjudicate a case from fragments.

Previous chapter · Contents · Next chapter

Reporting, investigation, charging, adjudication and the execution of an order are different activities. Their names and procedures vary by jurisdiction. An outcome at one stage does not automatically describe what occurred at another.

The Justice root and Law root provide the wider routes. For a specific case, current official procedures and qualified advice are necessary. This chapter develops the literacy needed to read public claims without collapsing the whole process into a single label.

Research about an institution should identify the actual decision or service being examined. A shorter processing time may be useful, but speed must be interpreted alongside relevant fairness, accuracy and access questions. A favourable average can conceal a group facing repeated barriers.

Ask learners to draw a simple fictional case journey and mark where each dataset is created. Then compare a count of reports with a count of completed cases. The difference should prompt questions about definitions, timing and process. It should not be automatically described as institutional failure or proof that every uncompleted report lacked substance.

Previous chapter · Contents · Next chapter

Rehabilitation and re-entry concern the conditions in which a person can return to community life and sustain lawful participation. Housing, employment, relationships, learning, health and appropriate supervision may be relevant, depending on the person and the applicable arrangements.

The Prisoner Rehabilitation and Re-entry branch develops the specialist responsibilities. This root asks how programmes are described and evaluated. Attendance, completion and later outcomes answer different questions and should remain separately visible.

Measures of further offending also need definitions. A new arrest, conviction or return to custody is not the same event, and follow-up periods influence comparisons. Participants can differ in ways that affect observed outcomes independently of programme quality.

Use a fictional comparison to ask who entered each programme, who completed it and who remained observable at follow-up. A careful interpretation can recognise meaningful progress while keeping uncertainty about the comparison. The purpose is a fair account of people and evidence, rather than a slogan that either guarantees success or assumes change is impossible.

Previous chapter · Contents · Next chapter

A fictional town introduces a clearer reporting service. Recorded reports rise from 100 to 140 in the next comparable period. A headline declares that the service has made the town less safe. Another declares that the increase proves complete public confidence. Neither conclusion follows from the totals alone.

Begin with what is established: more reports were recorded under the stated system. Inspect changes in definitions, access and recording. Seek suitable evidence about underlying experiences, reporting behaviour and other events during the period. Keep different sources distinct rather than merge incompatible counts.

Now ask which operational decision is needed. The service may require additional capacity because more people are reaching it, even while the cause of the rise remains uncertain. Good reasoning can support that limited decision without inventing an explanation for the entire social pattern.

Have the learner write three sentences: the observation, the unresolved explanation and the next evidence or action. This structure transfers into public discussion. It permits a clear response to a real change while preventing confidence from outrunning what the records establish.

Previous chapter · Contents · Next chapter

A strong learner can define the question, distinguish sources, examine a comparison and explain why a conclusion is or is not supported. They can also keep people and legitimate responsibilities visible when discussing evidence about crime.

Choose a public research summary and trace one claim back to its method. Who was studied? What was measured? Which observations are absent? Does the headline preserve the actual outcome and period? If essential information is missing, identify the omission precisely rather than dismiss the entire subject.

Then change the task. Ask whether the same evidence could support an individual legal decision, a local prevention plan or a comparison between countries. The answer may differ because the claim and consequences differ. This is where learning becomes judgement.

Return to Education: learning that transfers into life. The gold standard is neither automatic suspicion nor automatic trust. It is a reader who can ask a useful question, follow the evidence and choose the appropriate next route. The directory below connects that foundation with research, support, technical evidence and justice.

Previous chapter · Contents · Continue to the subject directory