Criminology and Safety
Lecture 1
Crime is a social construct (LMSH): legal, moral, social, humanistic
The elements of crime
- The action does harm
- There is social consensus about this
- There is a official societal response
4 pillars of crime reduction (PCCP)
1) Police
2) Court
3) Correction
4) Prevention
3 crime prevention strategies
Developmental – focus on the development of crime in individuals (strain theory)
Situational – focus on the crime itself, taking the opportunities away (situational
development)
Community – societal programs, social conditions (family, friends) (social control)
Types of validity
o Internal: the intervention caused the change
o External: generalizability
o Descriptive: how is everything done, is there missing information
o Statistical conclusion: quality of the statistical analysis
o Construct: measurement
Evidence-based crime prevention can inform policy best, but…
… there are political changes
… politicians want to be re-elected
… costs (you can see the results on the long-time)
… common-sense thinking is that locking-up works best
Article Welsh & Farrington
Statement: qualitative crime research should be used to strike a policy balance between
prevention and control
Prevention (outside the formal justice system) can be seen as an alternative for control to
reduce crime
Evidence based policy as framework (knowledge)
Problems are
- Politicians are afraid they will be seen as weak/soft for not locking up people
- Crime is seen as evil, so locking up is the best decision
- The advantages are not directly presented, this takes a while
,Imposed use – government requires evidence-based research before funding it
Some examples of prevention strategies are: home visiting, day care, parent training,
graduation incentives, supervising delinquent youth
Article Farrington
Validity - correctness of the inference about cause and effect
Types of validity
Internal validity – does the interventions cause the change in the outcomes (control
group to check this)
Statistical conclusion validity – are the intervention and the effect of the
interventions related, what is the effect size
Construct validity – operationalisation of the variables (do you measure what you
need to measure)
External validity – generalizability of the results
Descriptive validity – presentation of the important elements such as design,
sample, operationalisation, reliability, validity, effect size, follow-up measures
The SMS is a methodological quality scale, which is used to set up criteria for inclusion
Level 1. The correlation between an intervention and a measure of crime at a given
time, measured after the introduction of the studied intervention. This design does not
exclude any threat to internal validity, nor does it succeed in establishing causal ordering.
Level 2. Measuring crime before and after the intervention, without a (comparable)
control condition. Although causal ordering is established using a Level 2 design, too many
threats to internal validity remain for Level 2 studies to be considered adequate and
interpretable.
Level 3. Measuring crime before and after the intervention, in an experimental and a
comparable control group (quasi-experimental design). Level 3 designs are considered
the minimally interpretable design because they rule out many threats to internal validity,
including history, instrumentation, and test effects (Welsh & Farrington, 2006). However,
problems with selection effects and regression to the mean remain.
Level 4. Measuring crime before and after the intervention in multiple experimental
and control groups, controlling for other variables influencing violence (e.g. through
matching or statistical control). Compared to level 3 studies, this design has better
statistical control of external influences on outcome and better deals with residual threats to
internal validity.
Level 5. Measuring crime before and after implementation of an intervention where
units are randomly assigned to experimental and control conditions (experimental
design). Because this design has the highest possible internal validity and is therefore the
most convincing method to evaluate behavioral interventions, it is considered the gold
standard in evaluation research.
What works? Two level 3-5 studies that showed effectiveness
, What doesn’t work? Two level 3-5 studies that showed ineffectiveness
What’s promising? One level 3-5 studies that showed effectiveness
What’s unknown? Everything left
Pawson and Tilley had some critic on this, because they say that we need to look more to
the Context, Mechanisms and Outcomes (CMO). They point out that effects can vary in
different contexts
Lecture 2
What pushes people to commit crime?
Social learning
Group dynamics
Strain (stressors)
The assumption is that people are not inherently motivated
Strain theory
Negative stimuli/stressors lead to adaptations to conventional routes (coping)
You adapt with the stressors you encounter in your life and cope with this through crime
Strains Negative emotional states Coping mechanisms
Failure to achieve goals Anger Drug use
Removal positive stimuli Frustration Delinquency
Negative stimuli Fear Aggression
individual and social characteristics
Different people develop different kinds of strain and cope with it differently. A criminal
coping mechanism can occur. The coping mechanism a person choose depends on the
social and individual characteristics
What can prevent people from committing a crime?
Control = constraints on someone’s behaviour
People act in their own self-interest
- Learned or inherent
- Economic, social, psychological benefits
- Adjust perceived beliefs
Networks and control can reduce crime
Social control perspectives Britt and Rocque article for better understanding
Internal – norms, rules of acting, agreement with the moral beliefs
External – relationship with people
Internal/external can be:
Formal – police, law
Informal – peers, parents
So that was it about the offender… now we are looking at the environment
