PBL4810F Research Methods in Criminology & Criminal Justice Exam Notes

PBL4810F Research Methods in Criminology & Criminal Justice is a foundational study area for understanding how crime, justice, and social control are investigated with rigor. It brings together research design, ethics, measurement, sampling, data collection, and analysis, all within the distinctive realities of criminological and criminal justice work. These notes are structured to support exam preparation with clear explanations, practical examples, and the conceptual depth needed for higher education assessment.

1. Research Methods in Criminology and Criminal Justice: Scope, Purpose, and Core Logic

Research methods in criminology and criminal justice are not simply technical tools for collecting information. They are the procedures that allow scholars and practitioners to ask credible questions about crime, offending, victimisation, policing, courts, correctional systems, and community safety. In a field where policy decisions often have immediate consequences for public safety and individual rights, methodological competence is not optional. Poorly designed studies can produce misleading findings, waste public resources, reinforce stereotypes, or justify ineffective interventions. Strong research design, by contrast, enables evidence-based decision-making and allows criminology to move beyond opinion, rhetoric, and anecdote.

At the heart of criminological research is the effort to describe, explain, predict, and sometimes evaluate phenomena. Description answers what is happening, where, when, and to whom. Explanation asks why patterns occur. Prediction estimates likely future outcomes, such as reoffending risk. Evaluation determines whether a programme, policy, or practice works. These purposes often overlap. A study of youth gang involvement might begin descriptively by mapping age, location, and frequency of offending, but quickly move toward explanation by examining family disruption, peer networks, school exclusion, and neighbourhood disadvantage. If the study is well designed, it may also inform intervention by assessing whether mentoring, diversion, or family support reduces participation in gang activity.

Why Research Methods Matter in Criminology

Criminology deals with socially sensitive issues. Crime data are shaped by reporting behaviour, police practices, prosecutorial decisions, and institutional recording systems. A rise in recorded robbery may mean more robbery, better enforcement, changes in public reporting, or altered classification rules. Research methods help separate these possibilities. They also force careful thinking about causality. For example, if a researcher finds that young people in informal settlements are overrepresented in arrest statistics, that finding alone does not prove that residence in such areas causes crime. It may reflect structural inequality, concentrated police surveillance, school dropout, unemployment, or a combination of factors.

Research methods also matter because criminological phenomena are multi-level. Individual behaviour is influenced by family, peers, schools, neighbourhoods, organisations, and broader legal and economic conditions. A method that ignores one level may produce incomplete or distorted conclusions. For instance, studying prison violence only through inmate interviews may miss the impact of staffing shortages, institutional culture, overcrowding, or leadership practices. Similarly, studying victimisation only through official crime reports may overlook hidden harm, especially in domestic violence, sexual offences, and corruption.

A sound research methods course therefore trains students to think critically about:

What counts as evidence
How evidence is generated
How bias enters the research process
How to judge whether findings are trustworthy
How to connect findings to criminal justice policy and practice

Key Concepts in the Research Process

Several recurring concepts anchor the whole field.

Theory is a systematic explanation of relationships among phenomena. In criminology, theories include strain theory, social learning theory, control theory, routine activity theory, labelling theory, and critical perspectives. Research methods test, refine, or challenge these theories.

Variables are measurable characteristics that can vary. Age, income, prior convictions, fear of crime, neighbourhood cohesion, and victimisation frequency are examples. Variables may be independent, dependent, intervening, control, or contextual.

Concepts and constructs are ideas that need operationalisation. “Social disorganisation” is a concept; its construct may be measured using indicators such as residential mobility, unemployment, poverty, and family disruption.

Operationalisation is the process of turning a concept into measurable indicators. This matters because criminological concepts are often abstract. “Repeat victimisation” can be operationalised as being victimised more than once within a defined period, such as 12 months.

Reliability concerns consistency. A reliable measure gives stable results when the underlying condition has not changed.

Validity concerns whether a measure captures what it claims to measure. A reliable measure is not automatically valid. A scale may consistently record police arrest rates, but if arrest rates are heavily shaped by enforcement patterns, they may not validly represent underlying offending.

Generalisation refers to whether findings from a sample can reasonably be extended to a wider population. Criminology often struggles with generalisation because many studies are localised, institution-specific, or based on hard-to-reach populations.

The Logic of Scientific Inquiry

Research methods in criminology generally follow a cycle:

Identify a problem or question.
Review what is already known.
Formulate a focused question or hypothesis.
Select a design and sample.
Measure relevant variables.
Collect data ethically and systematically.
Analyse the data.
Interpret results in light of theory and limitations.
Draw conclusions and make recommendations.
Share findings for scrutiny and replication.

This logic applies whether the study is quantitative, qualitative, or mixed methods. The order may shift in exploratory qualitative work, but the need for transparency remains. A study of police-citizen interactions in Johannesburg, for example, may begin with open-ended interviews to understand lived experience, then develop survey questions to test patterns across a larger sample. The sequence differs, but the underlying logic is the same: disciplined inquiry guided by evidence.

Criminology and Criminal Justice as Distinct but Related Fields

Criminology examines crime as a social phenomenon. Criminal justice focuses more directly on institutions such as policing, prosecution, courts, corrections, and community corrections. Research methods are essential in both areas, but the questions differ slightly.

A criminology question might ask:

Why do burglary rates rise in certain neighbourhoods?
How do peer groups influence adolescent offending?
What social factors increase the risk of victimisation?

A criminal justice question might ask:

Does body-worn camera use improve police accountability?
Do diversion programmes reduce re-arrest?
How do parole conditions affect compliance?

These distinctions matter because the same research method can be applied differently depending on the level of analysis. A prison study may investigate inmate misconduct, staff turnover, or sentencing patterns, each requiring distinct data sources and analytic strategies.

Common Challenges in Criminological Research

Criminological research faces particular obstacles:

Hidden populations such as offenders, undocumented migrants, trafficking victims, and corrupt officials
Sensitive topics such as sexual violence, drug use, gang membership, and police abuse
Institutional gatekeeping in prisons, police departments, courts, and schools
Data fragmentation across agencies with different recording systems
Ethical risks involving vulnerability, stigma, and retaliation
Political pressure to produce convenient findings

These challenges require flexibility and methodological care. For example, a researcher studying domestic violence may need to use anonymous surveys, safety protocols, referral pathways, and trauma-informed interviewing. A researcher evaluating a correctional programme may need access permissions, security clearance, and careful handling of confidential records.

Summary Table: Purpose of Research in Criminal Justice

Purpose	Core Question	Example
Description	What is happening?	Recording trends in robbery over five years
Explanation	Why is it happening?	Linking robbery to unemployment and routine activity
Prediction	What is likely to happen next?	Estimating reoffending risk among parolees
Evaluation	Does an intervention work?	Testing a restorative justice programme
Policy development	What should be done?	Designing safer school policing strategies

Research methods make criminology credible because they replace assumption with evidence. They also make criminal justice more accountable by showing whether institutions produce fair and effective outcomes. In examinations, students should repeatedly return to the idea that method is not separate from substance: the quality of the question, design, measurement, and analysis determines the quality of the answer.

2. Research Paradigms, Approaches, and Study Design

A strong understanding of paradigms and design is essential in PBL4810F because methodology begins before data collection. It begins with assumptions about reality, knowledge, and the relationship between the researcher and the researched. These assumptions shape which questions are asked, which methods are chosen, and how results are interpreted. In criminology and criminal justice, where human behaviour, institutions, and social structures intersect, the choice of paradigm can fundamentally influence the direction of a study.

Research Paradigms

A paradigm is a worldview that guides how research is understood and conducted. The main paradigms relevant to criminology and criminal justice include positivism, interpretivism, critical approaches, and pragmatic mixed methods thinking.

Positivism assumes that social reality can be observed, measured, and analysed objectively. It prioritises causal explanation, hypothesis testing, and quantification. A positivist study might measure the relationship between unemployment and property crime across provinces using statistical techniques. The strength of positivism lies in comparability and systematic analysis. Its limitation is that it may simplify complex social meanings.

Interpretivism argues that social reality is constructed through meaning, interaction, and context. Rather than treating behaviour as merely an outcome to be measured, interpretivism seeks to understand how people experience and interpret crime, justice, and authority. A study of first-time offenders’ perceptions of court processes would fit well here. The value of interpretivism is depth and contextual richness. Its limitation is that findings may not easily generalise.

Critical approaches focus on power, inequality, domination, and structural injustice. Critical criminology examines how laws, policing, prisons, and punishment can reflect social inequalities related to class, race, gender, and political power. A critical study might investigate how stop-and-search practices disproportionately affect marginalised communities. The strength of critical approaches is their sensitivity to structural context and harm. The limitation is that they may be accused of prioritising critique over empirical neutrality, although that criticism itself is often a misunderstanding of the approach.

Pragmatism is often associated with mixed methods. It does not begin with a single view of reality but with the practical question of what method best answers the research problem. If a study of prison rehabilitation needs both statistical outcomes and inmate narratives, pragmatism justifies using surveys and interviews together.

Research Approaches

Research approaches describe the broad logic of inquiry.

Deductive research begins with theory and hypothesis. The researcher identifies a theoretical expectation, then tests it using data. For example, social learning theory predicts that association with delinquent peers increases offending. A deductive study could survey youth and test whether peer delinquency predicts self-reported offending after controlling for age and family background.

Inductive research begins with data and builds toward theory. The researcher explores patterns, themes, or categories and then develops explanations. A qualitative study of women exiting gang life might reveal recurring themes such as motherhood, fear, fatigue, and desire for safety. These findings may help build theory about desistance.

Abductive research moves back and forth between data and explanation. It is common in real-world criminology because researchers often encounter surprising patterns that require revising existing assumptions. For example, a pattern of reduced offending after a prison-based programme may prompt examination of informal peer support, institutional climate, and post-release supervision rather than the programme content alone.

Types of Research Design

Research design is the blueprint for collecting and analysing evidence. The main designs used in criminology and criminal justice include exploratory, descriptive, explanatory, experimental, quasi-experimental, correlational, longitudinal, cross-sectional, case study, and comparative designs.

Exploratory Design

Exploratory studies are used when a problem is not well understood. They help define concepts, generate hypotheses, and identify variables. A researcher exploring cyberfraud among university students might conduct preliminary interviews to understand methods, motivations, and reporting barriers.

Descriptive Design

Descriptive research answers “what is happening” questions. It may document the prevalence of burglary in a district, the demographic profile of prisoners, or the frequency of disciplinary action in a police service. Descriptive studies do not necessarily establish causation, but they provide essential baseline information.

Explanatory Design

Explanatory research seeks cause-and-effect relationships. It asks why one variable influences another. For example, a study may test whether school disengagement predicts juvenile offending and whether parental supervision moderates that relationship.

Experimental Design

Experimental designs involve manipulation of an independent variable and comparison of outcomes between groups. In a true experiment, participants are randomly assigned to treatment and control groups. In criminal justice, experiments are used where ethically and practically possible, such as testing different cognitive-behavioural interventions for offenders.

Quasi-Experimental Design

Quasi-experiments lack random assignment but still compare groups. A researcher may evaluate a new neighbourhood policing model in one district and compare outcomes with a similar district not using the model. Quasi-experiments are common in criminal justice because full randomisation is often impossible.

Correlational Design

Correlational studies examine whether variables are associated. They do not prove causality. For example, a study may find a positive correlation between neighbourhood poverty and violent crime. The next step is to ask whether the association persists after controlling for other influences.

Longitudinal Design

Longitudinal studies follow subjects over time. They are invaluable for understanding trajectories of offending, desistance, victimisation, and institutional experience. A longitudinal study of youth offending might track the same participants over three years to identify factors linked to persistence or cessation.

Cross-Sectional Design

Cross-sectional studies collect data at one point in time. They are efficient and useful for prevalence estimates, but they cannot easily determine temporal order. A survey of university students’ fear of crime in a single semester is cross-sectional.

Case Study Design

A case study investigates one bounded case in depth, such as one prison, one police station, one township, or one policy implementation. Case studies generate rich contextual knowledge and are especially valuable when the case is unusual, critical, or illustrative.

Comparative Design

Comparative designs compare across groups, locations, institutions, or time periods. For example, comparing recidivism rates between offenders participating in restorative justice and those in traditional sentencing can reveal differences in outcomes.

Choosing an Appropriate Design

The best design depends on:

The research question
The type of data available
Ethical constraints
Time and budget
Access to participants or records
Need for depth versus breadth
Desired level of causal inference

A common mistake is to choose a method because it is familiar rather than because it fits the question. If the aim is to understand lived experiences of wrongful arrest, a survey alone may be too blunt. If the aim is to test whether a new risk assessment tool predicts reoffending, interviews alone will not suffice.

Example: Evaluating a Diversion Programme

Consider a youth diversion programme introduced in a metropolitan area. Different designs would answer different questions:

Exploratory: How do youth, parents, and social workers understand the programme?
Descriptive: How many youths were diverted in the first year?
Explanatory: Does programme participation reduce reoffending?
Quasi-experimental: Are reoffending rates lower among diverted youth than among similar youth who were prosecuted?
Longitudinal: Do programme effects persist over two years?
Mixed methods: What outcomes changed, and why did participants experience those changes?

This example shows that design choice shapes the knowledge produced. In examinations, it is useful to state not only what a design is but why it is suitable for a particular criminological problem.

Common Errors in Design Thinking

Students often confuse design with method. Design is the overall strategy; method is the specific technique. A study can be descriptive and use questionnaires, or exploratory and use interviews. Another common error is to assume that quantitative designs are always superior. In criminology, evidence of subjective meaning, institutional culture, or hidden harm may be best captured qualitatively. A final error is ignoring feasibility. A beautiful design is useless if there is no access to participants, no ethical approval, or no budget for data collection.

Understanding paradigms and design is crucial because it gives structure to the entire research process. It ensures that methods are not selected casually, but in line with the question, the context, and the kind of evidence needed.

3. Measurement, Variables, Sampling, and Research Validity

Measurement is the bridge between concepts and evidence. In criminology and criminal justice, this bridge is especially delicate because many key phenomena are complex, hidden, or socially contested. Crime itself is not a single simple object. It can be measured through arrests, convictions, self-reports, victimisation surveys, or incident reports, each of which captures a different part of reality. Research quality depends heavily on how well variables are defined, sampled, and validated.

Variables and Levels of Measurement

A variable is any characteristic that can take different values. Variables may be qualitative or quantitative, and they are central to both descriptive and explanatory research.

The main levels of measurement are:

Level	Description	Example in Criminology
Nominal	Categories with no order	Type of offence, gender, region
Ordinal	Ordered categories without equal spacing	Fear of crime: low, medium, high
Interval	Equal intervals, no true zero	Standardised attitude scale
Ratio	Equal intervals with true zero	Age, income, number of arrests

A ratio variable such as number of prior convictions is especially useful because it supports a wide range of statistical analysis. But not all important criminological concepts are easily counted. Perceived legitimacy of police, trust in courts, or feelings of safety require scales or qualitative interpretation.

Operationalisation

Operationalisation is the process of translating a concept into a measurable form. This is one of the most important methodological skills in criminology.

For example:

Crime severity might be operationalised using offence categories and sentencing guidelines.
Social support might be measured by the number of reliable adults a participant can contact in crisis.
Neighbourhood disorder might be operationalised through physical decay, visible drug activity, vandalism, and noise.
Recidivism might be defined as a new arrest, reconviction, or reincarceration within a specified period.

The key issue is consistency. If a study defines recidivism as reconviction, it must use that definition throughout. Another study may define it as re-arrest, but the two should not be mixed as though they are identical. Since official systems record different stages of the criminal process, operational decisions directly affect findings.

Reliability and Validity

Reliability refers to consistency. If a scale measures fear of crime reliably, people with similar levels of fear should obtain similar scores across repeated measurement under similar conditions.

Validity refers to accuracy or truthfulness of measurement. Several forms are important:

Face validity: Does the measure seem appropriate on the surface?
Content validity: Does it adequately cover the full concept?
Construct validity: Does it behave as theory predicts?
Criterion validity: Does it correlate with an external benchmark?
Internal validity: Are observed effects really caused by the factor being studied?
External validity: Can findings be generalised beyond the sample?

A policing survey may ask whether respondents feel “safe,” but if the question is vague it may have limited validity. Safety may mean protection from crime, freedom from harassment, confidence in police, or general emotional comfort. Strong instruments use carefully worded items and pilot testing.

Measurement Error and Bias

Measurement error occurs when recorded values differ from the true values. In criminology, error may arise from memory problems, social desirability, interviewer influence, ambiguous definitions, or administrative inconsistencies.

Examples include:

Victims underreporting domestic violence due to fear or shame
Offenders minimising drug use or violent behaviour
Police data varying according to shifts, supervisors, or recording practices
Survey respondents misunderstanding legal terms
Prison records containing incomplete disciplinary information

Bias is a systematic error rather than random noise. If a victimisation survey consistently misses households without stable addresses, it may underestimate harm among highly mobile or homeless populations. If a police dataset reflects enforcement concentration in poor areas, it may exaggerate crime there relative to wealthier areas with lower scrutiny.

Sampling in Criminal Justice Research

Sampling is the process of selecting a subset of a population for study. Since criminological populations are often large, dispersed, or difficult to reach, sampling strategy is critical.

Probability Sampling

Probability sampling gives each member of the population a known chance of selection. It supports generalisation and statistical inference.

Common types include:

Simple random sampling
Systematic sampling
Stratified sampling
Cluster sampling

Stratified sampling is particularly useful when the population has important subgroups, such as gender, race, institution type, or offence category. A researcher studying prisoners may stratify by security level to ensure representation from remand and sentenced facilities.

Non-Probability Sampling

Non-probability sampling is used when probability sampling is impractical. It includes:

Convenience sampling
Purposive sampling
Snowball sampling
Quota sampling

Snowball sampling is often used for hidden populations such as drug users, gang affiliates, or undocumented victims. However, it can overrepresent tightly connected networks and underrepresent isolated individuals.

Sample Size and Representativeness

A large sample is not automatically a good sample. Representativeness matters more than size alone. A sample of 200 carefully selected offenders may be more useful than 2,000 conveniently available respondents if the smaller sample better reflects the target population.

Sample size depends on:

Population variability
Desired precision
Available resources
Statistical power
Expected attrition in longitudinal studies

Attrition is especially important in criminological panel studies because participants may move, withdraw, be re-incarcerated, or become uncontactable. Researchers must anticipate loss over time and plan accordingly.

Internal and External Validity

Internal validity asks whether the observed outcome is caused by the independent variable rather than confounding factors. Suppose a rehabilitation programme shows reduced reoffending. Was the programme effective, or were participants already more motivated to change? If the more motivated offenders self-selected into the programme, selection bias threatens internal validity.

External validity asks whether findings apply elsewhere. A study conducted in one urban correctional centre may not generalise to rural areas, different provinces, or other countries. Criminal justice systems differ in legal frameworks, institutional capacity, and social context.

Common Threats to Validity

Threats include:

Selection bias
History effects
Maturation
Testing effects
Instrumentation changes
Attrition
Regression to the mean
Confounding variables

For example, if a youth intervention study measures behaviour before and after a programme, improvement might be due to natural maturation rather than the programme itself. If a prison introduces a new disciplinary policy at the same time as a rehabilitation initiative, the two effects may be difficult to separate.

Practical Example: Measuring Fear of Crime

A useful illustration is fear of crime among university students in South Africa. Fear can be measured through:

Likert-scale questions on worry about robbery, assault, sexual violence, and kidnapping
Behavioural indicators such as avoiding night travel
Emotional indicators such as stress or hypervigilance
Perceived risk at home, on campus, and on public transport

A strong study would not use one item alone. It would test multiple items, check internal consistency, and perhaps compare survey results with qualitative interviews. This improves both reliability and validity.

Measurement and sampling are often treated as technical details, but they are actually central to whether a study can be trusted. If the wrong variable is measured, or the wrong sample selected, even the most sophisticated analysis cannot rescue the study.

4. Research Methods: Quantitative, Qualitative, and Mixed Approaches

One of the most important skills in PBL4810F is being able to distinguish between quantitative, qualitative, and mixed methods research, and to know when each is appropriate. In criminology and criminal justice, no single approach can answer every question. A statistical trend may show what is happening across a population, while interviews reveal how people experience the system. The strongest research often integrates both.

Quantitative Research

Quantitative research uses numerical data to identify patterns, test hypotheses, and estimate relationships between variables. It is especially useful when the researcher wants breadth, comparability, and statistical inference.

Common quantitative methods include:

Surveys
Official record analysis
Structured observation
Experiments and quasi-experiments
Secondary data analysis

Quantitative data are useful for questions like:

What percentage of victims report to police?
Which factors predict recidivism?
Does a policing intervention reduce calls for service?
How do arrest rates vary by district?

A major advantage of quantitative research is standardisation. Everyone answers the same items or is measured in the same way, which supports comparison. Quantitative data can also identify patterns not obvious from individual cases. For instance, a dataset of 10,000 court cases may reveal sentencing disparities by offence type, previous convictions, or legal representation.

However, quantitative research can flatten complexity. Numbers may not capture fear, humiliation, informal coping, or institutional culture. A re-arrest statistic does not explain how a person experienced supervision or why compliance broke down. For that, qualitative insight may be necessary.

Qualitative Research

Qualitative research aims to understand meaning, experience, context, and process. It is especially valuable when the phenomenon is complex, underexplored, or sensitive.

Common qualitative methods include:

In-depth interviews
Focus groups
Participant observation
Ethnography
Document analysis
Case studies

Qualitative research is ideal for questions like:

How do ex-offenders describe the stigma of reintegration?
Why do survivors of gender-based violence delay reporting?
How do police officers interpret “professionalism” in patrol work?
How do prisoners make sense of rehabilitation programmes?

Qualitative data are rich because they capture voice, interpretation, contradiction, and nuance. A participant may describe a prison programme as “helpful but humiliating,” a response that would be hard to reduce to a single numerical score. Qualitative methods also uncover processes, not just outcomes. They can show how decisions are made, how trust is built, and how institutional rules are interpreted in practice.

The limitation is that qualitative research usually uses smaller samples and is not designed for statistical generalisation. Yet this is not a weakness when the goal is depth rather than breadth. A carefully selected sample of 30 police officers may reveal organisational patterns that a large survey misses.

Mixed Methods Research

Mixed methods combines quantitative and qualitative approaches in one study. In criminology, this is especially powerful because crime and justice are both measurable and deeply social.

Mixed methods may be used to:

Explain statistical trends with interviews
Validate survey findings with observations
Develop survey instruments from qualitative exploration
Evaluate outcomes and understand implementation

There are different mixed methods designs:

Convergent design: Quantitative and qualitative data are collected separately and then compared.
Explanatory sequential design: Quantitative results are collected first, followed by qualitative data to explain them.
Exploratory sequential design: Qualitative data are collected first, followed by quantitative testing.
Embedded design: One method is primary and the other supports it.

Example: Domestic Violence Research

Domestic violence illustrates why mixed methods can be superior. A survey may show the prevalence of abuse, reporting rates, and service use. Interviews may reveal fear, financial dependence, cultural pressure, and institutional distrust. Combined, these methods provide a fuller picture than either alone.

If a survey shows that only 18% of survivors report to police, that figure is important. But interviews may reveal that many survivors fear retaliation, worry about children, or expect poor treatment at the station. The numbers show the scale of the issue; the narratives explain the mechanism.

Data Collection Techniques

Surveys

Surveys are efficient for large samples. They can be administered face-to-face, by telephone, online, or on paper. Good survey design requires clear wording, logical sequencing, and pilot testing.

Interviews

Interviews range from structured to unstructured. Semi-structured interviews are common in criminology because they balance consistency with flexibility. They allow the researcher to cover key topics while still exploring unexpected issues.

Focus Groups

Focus groups gather several participants together to discuss a topic. They are useful for exploring shared norms and collective perceptions, but confidentiality concerns must be managed carefully.

Observation

Observation allows the researcher to see behaviour in context. It is useful in policing, courtrooms, prisons, and public spaces. Structured observation uses a checklist, while participant observation involves deeper immersion in the setting.

Documentary Analysis

Documents such as police directives, court judgments, policy reports, disciplinary records, and media coverage can be analysed to understand institutional practice and discourse.

Strengths and Weaknesses of Each Approach

Approach	Strengths	Limitations
Quantitative	Comparison, generalisation, measurement, testing	May oversimplify meaning and context
Qualitative	Depth, context, process, voice	Smaller samples, limited statistical generalisation
Mixed methods	Completeness, triangulation, flexibility	Time-consuming, requires skill in both traditions

Triangulation and Complementarity

Triangulation involves using more than one method or data source to study the same issue. It can improve confidence in findings. If a prison study shows through official records, interviews, and observation that overcrowding is linked to conflict, the convergence of evidence strengthens the conclusion. Complementarity means each method addresses different parts of the research problem. The survey may establish scale, while the interview explains experience.

Matching Method to Research Question

The best method depends on the question:

How many? → quantitative
How often? → quantitative
Why? → often mixed or qualitative, sometimes quantitative
How do people experience it? → qualitative
Does it work? → quantitative or mixed
How is it implemented? → qualitative or mixed

Students should be careful not to treat one method as universally superior. In criminal justice, where institutions shape human lives, robust evidence often requires both measurement and interpretation. The ability to justify a method choice is a key exam skill.

5. Ethics, Data Analysis, and Writing Up Research in Criminology and Criminal Justice

Ethics, analysis, and reporting bring the research process to completion, but they are not merely the final stage. Ethical thinking begins with the research question and continues through sampling, data collection, analysis, and dissemination. Likewise, analysis is shaped by the earlier choices made in design and measurement. A well-conducted criminological study is not only methodologically sound; it is also ethically defensible and clearly communicated.

Ethical Principles in Criminal Justice Research

Research ethics in criminology are particularly important because participants may be vulnerable, stigmatised, detained, traumatised, or otherwise constrained. The main ethical principles are:

Respect for persons and autonomy
Informed consent
Voluntary participation
Beneficence
Non-maleficence
Justice
Confidentiality and privacy
Integrity and transparency

These principles require careful application. For example, interviewing prisoners about staff misconduct can expose them to retaliation. In such a case, anonymity, secure data storage, careful scheduling, and institutional permissions become essential. Similarly, interviewing survivors of sexual violence requires trauma-informed practice, the ability to stop the interview, and referrals to support services if needed.

Informed Consent

Informed consent means participants understand:

The purpose of the study
What participation involves
Possible risks and benefits
Their right to refuse or withdraw
How data will be used and protected

Consent is not just a signature. It must be meaningful. In a correctional setting, participants may feel coerced by authority. Researchers must therefore stress that refusal will not affect parole, treatment, privileges, or institutional standing. This is especially important when the researcher is linked to the criminal justice system.

Confidentiality and Anonymity

Confidentiality means protecting information shared by participants. Anonymity means the identity of the participant is not known even to the researcher. In criminology, anonymity is difficult in small institutions or rare cases, so confidentiality strategies become crucial.

Measures include:

Removing names and identifying details
Using pseudonyms
Storing files securely
Restricting access to data
Reporting findings in aggregated form
Avoiding descriptions that reveal identity indirectly

A study of senior police management in one district cannot always guarantee full anonymity if the organisation is small and the role is identifiable. The researcher must therefore decide carefully what can be reported without exposing participants.

Ethical Challenges Specific to Criminology

Criminological research often involves:

Illegal or harmful behaviour
Ongoing victimisation
Risk of disclosure of crimes
Imbalances of power
Access to institutions with security restrictions
Potential legal obligations to report certain disclosures

Researchers must know the limits of confidentiality. If a participant discloses imminent harm to a child or a credible threat to life, the researcher may have a duty to act according to law and ethics protocols.

Data Analysis: Quantitative

Quantitative analysis begins with cleaning and organising data. Missing values, outliers, coding errors, and inconsistent categories must be checked before interpreting results. Basic statistical analysis may include:

Frequencies and percentages
Measures of central tendency
Cross-tabulations
Correlation
T-tests and chi-square tests
Regression analysis
Analysis of variance
Time-series or panel analysis in advanced studies

A simple example is a table showing the age distribution of offenders in a sample:

18–24: 32%
25–34: 41%
35–44: 17%
45+: 10%

Such a table is descriptive. If the study then tests whether age predicts violent offending, the analysis becomes explanatory. Regression can help identify which factors remain significant when other variables are controlled. For example, if peer delinquency, unemployment, and prior victimisation all correlate with offending, regression may reveal that peer delinquency is the strongest predictor once the others are included.

Interpreting Statistical Results

Students often make the mistake of focusing only on significance. Statistical significance does not automatically imply practical importance. A tiny effect may be statistically significant in a very large sample, yet have little policy value. Researchers should consider:

Effect size
Direction of the relationship
Confidence intervals
Practical implications
Potential confounders

In criminal justice, practical importance is often crucial. A programme that reduces reoffending by 2% may be statistically interesting but operationally modest unless it is inexpensive and scalable. A programme that reduces violence in a remand centre by 15% may be far more significant in practice.

Data Analysis: Qualitative

Qualitative analysis usually involves coding and thematic interpretation. The researcher reads transcripts or field notes, identifies recurring ideas, and groups them into categories or themes. The process may be:

Familiarisation with the data
Initial coding
Searching for themes
Reviewing themes
Defining and naming themes
Writing the interpretation

For example, interviews with parolees may produce themes such as “pressure to find work,” “stigma from employers,” “police surveillance,” and “family responsibility.” These themes can be linked to desistance theory and reintegration policy.

Good qualitative analysis is not impressionistic. It should be systematic, transparent, and grounded in evidence from the data. Quotations are used to demonstrate themes, but they must be chosen carefully and interpreted in context.

Writing Up Research

A well-written research report usually includes:

Title
Abstract or executive summary
Introduction and problem statement
Literature review
Methodology
Findings or results
Discussion
Conclusion and recommendations
References

The methodology section is especially important in research methods courses. It should explain:

Research design and rationale
Population and sample
Sampling method
Data collection methods
Instruments or interview guides
Ethical procedures
Data analysis techniques
Limitations

The findings section should present results clearly and logically, without overclaiming. The discussion should interpret the findings in relation to theory, previous studies, and policy relevance. Conclusions should be concise and grounded in evidence.

Example of a Strong Criminology Research Report Structure

Suppose a student studies “Factors affecting reporting of domestic violence in rural communities.” A strong report might proceed as follows:

Introduce the problem and its social significance
Review literature on underreporting, patriarchy, service access, and fear
Explain that a mixed methods design was used
State that 120 survey respondents and 20 interview participants were selected
Present survey statistics on reporting patterns
Present interview themes on fear, dependence, and mistrust
Discuss how structural barriers and interpersonal dynamics interact
Recommend community-based support and accessible reporting channels

This structure shows how method and substance work together.

Common Exam Mistakes

Students frequently lose marks by:

Confusing reliability with validity
Treating correlation as causation
Ignoring sampling limitations
Misstating the purpose of qualitative research
Describing ethics too generally without criminological examples
Failing to link findings back to theory
Overlooking how institutions shape official data
Writing methodology as a list rather than a reasoned justification

Final Exam Preparation Points

To perform well in PBL4810F, it helps to remember the following:

Research methods are the foundation of credible criminology.
The research question determines the most suitable design and method.
Measurement must be reliable, valid, and context-sensitive.
Sampling affects generalisability and bias.
Quantitative and qualitative methods answer different but complementary questions.
Ethics is central in criminal justice research because of vulnerability and power imbalance.
Data analysis must be interpreted carefully, with attention to limitations.
Good research writing is clear, structured, and evidence-based.

Consolidated Revision Table

Topic	Key Idea	Why It Matters
Paradigms	Different worldviews shape research	Guides what counts as knowledge
Design	Overall strategy for inquiry	Ensures method fits question
Measurement	Turning concepts into variables	Determines accuracy of evidence
Sampling	Selecting participants or cases	Affects representativeness
Quantitative methods	Numerical analysis	Useful for testing and comparison
Qualitative methods	Meaning and experience	Useful for depth and context
Ethics	Protection of participants	Essential in vulnerable settings
Analysis	Organising and interpreting data	Produces findings and conclusions

Research methods in criminology and criminal justice are ultimately about disciplined thinking. They teach students to ask better questions, examine evidence critically, and recognise the limits of what data can prove. In a field shaped by inequality, harm, state power, and public concern, that discipline is indispensable. A student who understands methodology is better prepared not only for exams, but also for responsible professional practice in policing, corrections, policy analysis, victim support, and community safety work.