SV CRSS EXAM
HC1: Introduction to Criminological Research
1. Criminological research
Criminology:
Interdisciplinary science. Explanatory models: e.g. psychology, sociological,
economy, law, politics.
Analysis on micro, meso and macro level. (individual; organisations; global/national)
Empirical focus: understanding the real world by direct/indirect observations and
experiences
2. Quantitative versus qualitative
Quantitative, dominant model: generalization; testing theories; measuring the size/nature of a
phenomenon; provides broad date (N = big)
Qualitative, less common model: in depth exploration of a phenomenon; developing new
theories; uncovering how and why; provides in depth data (N = small)
Combination strengthens research quality
3. Quantitative research and sampling
Quantitative research (in criminology):
Uses statistical techniques to study crime, criminal behavior, and justice system
responses.
Is systematic: empirical investigation of observable phenomena using numerical data
and objective measurement
Official crime data, non-judicial data, self-report surveys, victim/offender surveys etc.
Is used to:
o Describe crime trends and patterns
o Test criminological theories using empirical data
o Predict crime occurrences and risk factors
o Evaluate the effectiveness of criminal justice policies and interventions
Sampling techniques:
Probability sampling: general trend: Everyone has a known and non-zero chance of
being selected for representative and generalizable results
Simple random sampling: random, e.g. 500 students selected nationwide
Systematic sampling: choosing every xth student from the course
Stratified sampling: separating into categories, e.g. male and female, and taking
random samples from each
Cluster sampling: choosing all the people from a chosen cluster and sampling all of
them
, 4. Quantitative research design
We discuss four main designs:
Descriptive:
Takes ‘snapshot’ of crime patterns or justice system operations
Describes the characteristics of offenders, victims, criminal events or types of crime
Frequency, spatial distribution, population profiling, crime types, time trends, etc.
Used to inform:
o crime mapping (law enforcement)
o resource allocation (policy makers)
o crime reporting (journalists)
o further research design (academics/science)
Example descriptive report: UCR: nationwide, annual statistical report compiled by the FBI,
first established in 1930. It collects and publishes crime data that is reported.
NIBRS: evolution of UCR to replace it, more detailed system. NIBRS is more detailed about
every criminal incident, UCR only reported the most biggest crimes.
Limitations of descriptive research:
Absent of context: causes are not reported, only about things that happen not the
reasons behind them; doesn’t explain the why
Dark number of crime / non-response: crime funnel (trechter); all crime crime
observed crime reported
Complexity of reality becomes oversimplified
Quality depends on accuracy of data sources: if the data is flawed or faulted you don’t
have good findings.
Correlational:
Examines the relationships between two or more variables to determine whether they
are statistically associated, if there is a pattern
Explores risk factors and predictors of criminal behavior
Identify important variables using existing data sets and find relationships between
them, e.g.:
o Patrol frequency <> burglary rate
o Poverty level <> violent crime rate
Example: routine activity approach from Falson:
Used official crime statistics (UCR reports and labor force/census data) Examined:
Independent variables: vacant homes during day, working women, single-person
households; vs:
Dependent variable: property crime rates
Positive correlation found between daytime household absence and property crime
Still has enduring influence on situational crime prevention (target hardening,
surveillance etc.)
,Limitations of correlational research:
Cannot establish causality (other variables may be influencing the result; why of the
correlation is not explained)
Subject to confounding variables (e.g., third variables affecting both)
Spurious correlations can mislead (e.g., ice cream sales and drowning rates both rise
in summer)
Experimental:
Testing causal relationships using randomized control trials
Manipulating one variable (independent) to observe the effect on another
(dependent), while controlling for other factors.
Evaluate whether certain interventions can reduce crime
Identify cause and effect between variables
Evaluate the effectiveness of criminal justice policies
Examples:
Stanford prison experiment: psychological evaluation. Escalated: guard too
aggressive. Failed experiment
Milgram experiment: the shock thing, testing obedience.
Minneapolis domestic violence experiment. 330 police-handled domestic violence
incidents involving misdemeanor assault. Police officers used a random procedure
(lottery-style) to assign one of three responses:
Arrest the suspect
Separate the suspect from the victim for 8 hours
Mediate the situation (advise and calm both parties)
o Most ended up doing arrest: easier. So variables flawed.
o Established evidence-based policing and mandatory arrest for domestic
violence
Limitations of experimental research:
May have low generalizability/external validity
Can be expensive and time-consuming
Informed consent and harm? E.g. if officer was assigned mediate but arrest was more
appropriate, not ethical to not arrest
Difficult to experiment with real world
Longitudinal:
Involves collecting data from the same individuals or groups over time, often across
months or years
Used to track changes, development, and long-term effects related to crime, criminal
behavior, and justice system outcomes.
Observe patterns over time
Study causes and consequences of crime
Test life-course theories of criminal behavior
Examine how early life experiences relate to later outcomes (incarceration, offending
etc)
, Example: Cambridge study of delinquent males (N = 411): CSDD followed the same group of
boys from childhood to adulthood, collecting rich data on their lives and criminal behavior
Key variables: family structure, school performance, parental criminality, official/self-
reported offending, employment etc.
The development of offending behavior across the life course
Risk and protective factors for criminal behaviour
The influence of family/peers/community on delinquency
Identified small group of ‘chronic offenders’, early risk factors (poor parental
supervision, low school achievement etc.), and desistance factors (marriage, stable
employment etc.)
Limitations of longitudinal research:
Time-consuming
Requires long-term planning and funding
Participant retention
Ethical concerns around confidentiality in long-term tracking (big brother is watching
you vibe)
Summary
The use of statistical tools and methodologies for analysing crime and developing
strategies for crime reduction.
Essential role in developing criminological approaches and evidence-based policies.
Broad applications in policy and practice (risk assessment, predictive policing, crime
mapping etc)
Data Reliability & Validity, sampling issues, ethical concerns (accurate/unbiased data
collection, representativeness, misleading/manipulated variables)
Part II: Qualitative
5. Main features qualitative research
Main features of qualitative research:
Is explorative: no hypothesis that is tested. Research is done because not much is
known about the topic: exploring
Is interpretive: verstehen: understanding e.g. a point of view, the way people see
reality.
Constructivist: reality and therefor crime is socially constructed, so crime changes
through time because idea of what is crime and what isn’t is constructed
Inductive: data to theory, create theory from the data, start without idea about
something.
Holistic: all aspects (not variables, not a qualitative word) are taken together to
understand a social phenomena
Contextual: research is done in context, we don’t separate them. E.g. location (drugs
in NL vs Ghana)
Cyclical and iterative: repeating your steps, going back to e.g. research question,
research is not linear but cyclical
Primary data: collecting your data first-hand, no secondary data collected by other
people for other purposes
HC1: Introduction to Criminological Research
1. Criminological research
Criminology:
Interdisciplinary science. Explanatory models: e.g. psychology, sociological,
economy, law, politics.
Analysis on micro, meso and macro level. (individual; organisations; global/national)
Empirical focus: understanding the real world by direct/indirect observations and
experiences
2. Quantitative versus qualitative
Quantitative, dominant model: generalization; testing theories; measuring the size/nature of a
phenomenon; provides broad date (N = big)
Qualitative, less common model: in depth exploration of a phenomenon; developing new
theories; uncovering how and why; provides in depth data (N = small)
Combination strengthens research quality
3. Quantitative research and sampling
Quantitative research (in criminology):
Uses statistical techniques to study crime, criminal behavior, and justice system
responses.
Is systematic: empirical investigation of observable phenomena using numerical data
and objective measurement
Official crime data, non-judicial data, self-report surveys, victim/offender surveys etc.
Is used to:
o Describe crime trends and patterns
o Test criminological theories using empirical data
o Predict crime occurrences and risk factors
o Evaluate the effectiveness of criminal justice policies and interventions
Sampling techniques:
Probability sampling: general trend: Everyone has a known and non-zero chance of
being selected for representative and generalizable results
Simple random sampling: random, e.g. 500 students selected nationwide
Systematic sampling: choosing every xth student from the course
Stratified sampling: separating into categories, e.g. male and female, and taking
random samples from each
Cluster sampling: choosing all the people from a chosen cluster and sampling all of
them
, 4. Quantitative research design
We discuss four main designs:
Descriptive:
Takes ‘snapshot’ of crime patterns or justice system operations
Describes the characteristics of offenders, victims, criminal events or types of crime
Frequency, spatial distribution, population profiling, crime types, time trends, etc.
Used to inform:
o crime mapping (law enforcement)
o resource allocation (policy makers)
o crime reporting (journalists)
o further research design (academics/science)
Example descriptive report: UCR: nationwide, annual statistical report compiled by the FBI,
first established in 1930. It collects and publishes crime data that is reported.
NIBRS: evolution of UCR to replace it, more detailed system. NIBRS is more detailed about
every criminal incident, UCR only reported the most biggest crimes.
Limitations of descriptive research:
Absent of context: causes are not reported, only about things that happen not the
reasons behind them; doesn’t explain the why
Dark number of crime / non-response: crime funnel (trechter); all crime crime
observed crime reported
Complexity of reality becomes oversimplified
Quality depends on accuracy of data sources: if the data is flawed or faulted you don’t
have good findings.
Correlational:
Examines the relationships between two or more variables to determine whether they
are statistically associated, if there is a pattern
Explores risk factors and predictors of criminal behavior
Identify important variables using existing data sets and find relationships between
them, e.g.:
o Patrol frequency <> burglary rate
o Poverty level <> violent crime rate
Example: routine activity approach from Falson:
Used official crime statistics (UCR reports and labor force/census data) Examined:
Independent variables: vacant homes during day, working women, single-person
households; vs:
Dependent variable: property crime rates
Positive correlation found between daytime household absence and property crime
Still has enduring influence on situational crime prevention (target hardening,
surveillance etc.)
,Limitations of correlational research:
Cannot establish causality (other variables may be influencing the result; why of the
correlation is not explained)
Subject to confounding variables (e.g., third variables affecting both)
Spurious correlations can mislead (e.g., ice cream sales and drowning rates both rise
in summer)
Experimental:
Testing causal relationships using randomized control trials
Manipulating one variable (independent) to observe the effect on another
(dependent), while controlling for other factors.
Evaluate whether certain interventions can reduce crime
Identify cause and effect between variables
Evaluate the effectiveness of criminal justice policies
Examples:
Stanford prison experiment: psychological evaluation. Escalated: guard too
aggressive. Failed experiment
Milgram experiment: the shock thing, testing obedience.
Minneapolis domestic violence experiment. 330 police-handled domestic violence
incidents involving misdemeanor assault. Police officers used a random procedure
(lottery-style) to assign one of three responses:
Arrest the suspect
Separate the suspect from the victim for 8 hours
Mediate the situation (advise and calm both parties)
o Most ended up doing arrest: easier. So variables flawed.
o Established evidence-based policing and mandatory arrest for domestic
violence
Limitations of experimental research:
May have low generalizability/external validity
Can be expensive and time-consuming
Informed consent and harm? E.g. if officer was assigned mediate but arrest was more
appropriate, not ethical to not arrest
Difficult to experiment with real world
Longitudinal:
Involves collecting data from the same individuals or groups over time, often across
months or years
Used to track changes, development, and long-term effects related to crime, criminal
behavior, and justice system outcomes.
Observe patterns over time
Study causes and consequences of crime
Test life-course theories of criminal behavior
Examine how early life experiences relate to later outcomes (incarceration, offending
etc)
, Example: Cambridge study of delinquent males (N = 411): CSDD followed the same group of
boys from childhood to adulthood, collecting rich data on their lives and criminal behavior
Key variables: family structure, school performance, parental criminality, official/self-
reported offending, employment etc.
The development of offending behavior across the life course
Risk and protective factors for criminal behaviour
The influence of family/peers/community on delinquency
Identified small group of ‘chronic offenders’, early risk factors (poor parental
supervision, low school achievement etc.), and desistance factors (marriage, stable
employment etc.)
Limitations of longitudinal research:
Time-consuming
Requires long-term planning and funding
Participant retention
Ethical concerns around confidentiality in long-term tracking (big brother is watching
you vibe)
Summary
The use of statistical tools and methodologies for analysing crime and developing
strategies for crime reduction.
Essential role in developing criminological approaches and evidence-based policies.
Broad applications in policy and practice (risk assessment, predictive policing, crime
mapping etc)
Data Reliability & Validity, sampling issues, ethical concerns (accurate/unbiased data
collection, representativeness, misleading/manipulated variables)
Part II: Qualitative
5. Main features qualitative research
Main features of qualitative research:
Is explorative: no hypothesis that is tested. Research is done because not much is
known about the topic: exploring
Is interpretive: verstehen: understanding e.g. a point of view, the way people see
reality.
Constructivist: reality and therefor crime is socially constructed, so crime changes
through time because idea of what is crime and what isn’t is constructed
Inductive: data to theory, create theory from the data, start without idea about
something.
Holistic: all aspects (not variables, not a qualitative word) are taken together to
understand a social phenomena
Contextual: research is done in context, we don’t separate them. E.g. location (drugs
in NL vs Ghana)
Cyclical and iterative: repeating your steps, going back to e.g. research question,
research is not linear but cyclical
Primary data: collecting your data first-hand, no secondary data collected by other
people for other purposes
