Science for Everybody

Bullet-point answer points

• Builds a profile from evidence from the particular crime or crimes.
• Does not begin with predetermined offender categories.
• May use behavioural or geographical patterns.

Full model answer

The data-driven approach builds an offender profile from evidence gathered from the particular crime or series of crimes. Investigators work from behavioural and geographical patterns towards hypotheses about the offender rather than beginning with a predetermined offender type.

Common pitfalls and how to avoid them

Do not simply state that it “uses data”. Explain that investigators build upwards from crime evidence. Also recognise that data-driven is the current AQA terminology for the approach previously called bottom-up.

Bullet-point answer points

• Associated particularly with David Canter.
• Applies psychological theory and research to investigations.
• Uses information about offender behaviour.
• Considers behavioural consistency.
• Considers interpersonal coherence.
• Can use statistical techniques such as SSA.
• May assist case linkage.

Full model answer

Investigative psychology is associated particularly with David Canter and applies psychological research to criminal investigations. Investigators analyse behaviours displayed during offences and look for behavioural consistency and interpersonal coherence. Statistical techniques such as Smallest Space Analysis can identify relationships between behaviours. This information may help investigators link crimes and generate hypotheses about an unknown offender.

Common pitfalls and how to avoid them

Do not describe geographical profiling alone. Mention behaviour, psychological principles and preferably a statistical technique such as Smallest Space Analysis.

Bullet-point answer points

• Plot all offence locations.
• Identify the main geographical cluster.
• Consider a possible centre of gravity.
• The five clustered offences could indicate marauder behaviour.
• The offender’s base might lie within the general crime region.
• Do not assume the mathematical centre is the home.
• Compare the distant burglary behaviourally before linking it.
• It could indicate travel, expanding range or a different offender.

Full model answer

Investigators could plot the six burglary locations and identify the main offence region formed by the five clustered crimes. If the offender is a marauder, their home or another important base may lie within this general region. However, police should not assume that the exact centre is their home because roads and other environmental factors influence movement. They should also compare the behaviour in the sixth burglary because its greater distance may indicate an expanding crime range, commuter behaviour or a different offender.

Common pitfalls and how to avoid them

Use the details from the scenario. Simply defining geographical profiling will not gain full AO2 credit.

Do not write that the offender definitely lives in the middle.

Bullet-point answer points

• Offenders may repeat behaviours across crimes.
• Examples include victim approach, language, control or forensic precautions.
• Investigators compare behavioural combinations.
• Unusual repeated behaviours may support case linkage.
• Linked offences allow evidence to be combined.
• Patterns may then become clearer.
• Consistency is probabilistic rather than absolute.

Full model answer

Behavioural consistency refers to the idea that an offender may repeat some behaviours across different offences. For example, they may approach victims using similar language, use the same method of control and take similar precautions to avoid detection. If several crimes contain the same unusual combination of behaviours, investigators may hypothesise that one person committed them. Linking the crimes allows police to combine witness statements, forensic information and geographical evidence from several investigations. However, behavioural consistency should not be treated as proof because offenders can change their behaviour over time.

Common pitfalls and how to avoid them

Do not claim that offenders behave identically every time.

For higher marks, explain the consequence: case linkage allows investigators to combine information from different crimes.

Bullet-point answer points

• Data-driven profiling begins with evidence from the particular crime.
• Typology profiling begins with pre-existing offender categories.
• Data-driven profiling works bottom-up.
• Typology profiling works top-down.
• Data-driven profiling uses investigative psychology and geographical profiling.
• Statistical analysis may be used.
• Typology approaches classify behaviour into offender types.
• Data-driven profiling is less reliant on rigid categories.
• Both produce hypotheses rather than proof of guilt.

Full model answer

The data-driven approach begins with evidence from a particular crime and works upwards towards conclusions about the offender. Investigators may analyse behavioural consistency using investigative psychology or examine crime locations through geographical profiling. Statistical methods can also identify relationships between behavioural variables.

In contrast, the typology approach begins with existing offender categories and attempts to decide which category best fits the crime. It therefore works from general categories down towards an individual offender.

One advantage of the data-driven approach is that it does not require offenders to fit rigid types. This may better reflect the complexity of human behaviour. Its use of statistical and geographical evidence can also make some conclusions more objective. However, neither approach can independently establish guilt. Both should provide investigative hypotheses that require supporting evidence.

Common pitfalls and how to avoid them

Do not write two completely separate descriptions without making comparisons.

Use comparison language such as:

• whereas;
• in contrast;
• both;
• however.

Bullet-point answer points

AO1

• Previously called the bottom-up approach.
• Builds hypotheses from crime evidence.
• Associated particularly with David Canter.
• Investigative psychology analyses offender behaviour.
• Behavioural consistency can assist case linkage.
• Interpersonal coherence links victim interactions with wider interpersonal behaviour.
• SSA identifies statistical relationships between behavioural variables.
• Canter and Heritage analysed 66 offences from 27 offenders.
• Geographical profiling analyses offence locations.
• Mental maps influence location choice.
• Least-effort principle.
• Distance decay.
• Buffer zones.
• Circle hypothesis.
• Marauders and commuters.
• Canter and Larkin investigated 45 serial sexual offenders.

AO3

• Scientific use of empirical data.
• Canter and Heritage provide research support.
• Geographical research supports spatial consistency.
• John Duffy demonstrates real-world application.
• Flexible compared with rigid typologies.
• Supports case linkage.
• Solved-crime database bias.
• Restricted samples.
• Cultural bias.
• Behaviour may change over time.
• Geographical models can oversimplify movement.
• Potential ethical concerns and confirmation bias.
• Classic studies may have reduced temporal validity.
• Profiles should assist rather than replace conventional evidence.

Full model answer

The data-driven approach to offender profiling was traditionally known as the bottom-up approach. Rather than beginning with a predetermined offender category, investigators start with information from a particular crime or series of crimes and develop hypotheses about the offender.

One element is investigative psychology, associated particularly with David Canter. Investigative psychology attempts to apply scientific psychological research to police investigations. One important concept is behavioural consistency. This suggests that an offender may repeat certain behaviours across crimes, such as victim approach, language or methods of control. Similar combinations of behaviour may therefore help investigators establish case linkage.

Another concept is interpersonal coherence. This proposes that the way an offender interacts with a victim may provide information about the offender’s wider interpersonal behaviour. However, these are probabilistic predictions rather than certain conclusions.

Canter and Heritage (1990) provide research support. They analysed 66 sexual assaults committed by 27 offenders and identified 33 behavioural variables from victim accounts. Smallest Space Analysis was used to examine which behaviours occurred together. They identified systematic relationships between behaviours, supporting the possibility that crime-scene behaviour can be analysed scientifically.

The second major component is geographical profiling. Investigators plot locations associated with linked crimes and attempt to identify spatial patterns. The approach assumes that offenders develop mental maps based on places they know and therefore do not choose crime locations completely randomly.

Distance decay predicts that offending generally becomes less likely as distance from the offender’s base increases. However, offenders may avoid locations immediately around their home because of the risk of recognition, producing a buffer zone.

Canter and Larkin distinguished between marauders and commuters. A marauder commits offences around a base that lies within the overall crime region. In contrast, a commuter travels away from their normal area before committing offences. Canter and Larkin studied 45 serial sexual offenders and found considerable evidence of marauding behaviour.

One strength of the data-driven approach is its scientific basis. Investigative psychology can use measurable behaviours and statistical procedures rather than depending entirely on the subjective judgement of a profiler. Canter and Heritage’s use of Smallest Space Analysis demonstrates that behavioural relationships can be objectively investigated. Therefore, the approach may have greater scientific credibility than profiling based largely on intuition.

There is also research support for geographical profiling. Canter and Larkin identified systematic relationships between offenders’ homes and offence locations. This supports the claim that criminal behaviour often demonstrates spatial consistency.

Furthermore, the approach has useful real-world applications. David Canter contributed to the investigation of John Duffy by analysing both behavioural and geographical information. His predictions included Duffy’s likely geographical area and knowledge of the railway network. Therefore, even when profiling cannot identify an offender directly, it may help police prioritise suspects and locations.

However, the approach has methodological limitations. Much profiling research relies on solved crimes because researchers need to know the identity and characteristics of the offender. Offenders whose crimes remain unsolved may behave differently. Therefore, the databases used to generate profiling principles may be biased towards offenders who are easier to detect.

There are also sample and cultural biases. Classic studies often investigated male serial sexual offenders in particular countries. Travel behaviour depends on factors such as road networks, public transport and population density. Consequently, geographical patterns found in British cities cannot automatically be generalised to every offender.

A further criticism is that behaviour may not remain completely consistent. Offenders can change their methods, learn from experience or react differently to individual victims. Similarly, geographical range may change when an offender moves home or becomes more confident.

There are also ethical risks. Many innocent people could fit a statistical profile. If investigators treat the profile too confidently, they may develop tunnel vision and ignore evidence that contradicts their preferred suspect.

Finally, classic profiling research may have reduced temporal validity. CCTV, smartphones, digital communications and improved forensic techniques have changed both offending and police investigation. Therefore, older profiling principles should continue to be tested against contemporary data.

Overall, the data-driven approach represents an important attempt to make offender profiling more scientific. Investigative psychology and geographical profiling can identify useful patterns and improve investigative decision-making. However, profiling provides probabilities rather than certainties and should therefore complement forensic and conventional police evidence rather than replace it.

Common pitfalls and how to avoid them

• Cover both investigative psychology and geographical profiling.
• Include named research.
• Develop evaluation rather than listing isolated criticisms.
• Link evidence back to profiling.
• Do not claim profiling identifies offenders with certainty.
• Use data-driven approach as the current AQA terminology while recognising that textbooks may still refer to the bottom-up approach.