16.1.2 The Bottom-Up Approach to Offender Profiling

• The bottom-up approach (BUA) to offender profiling contrasts to the top-down approach as developed by the FBI in the USA as it is ‘data-driven’ 
• ‘Data-driven’ means that the BUA does not begin with the assumption that each crime will fit into a typology; instead it uses the crime scene as the basis for the creation of a profile
• The BUA is inductive as it uses information already present and draws ideas, theories and conclusions from it i.e. it aims to develop a theory from the data presented
• The BUA starts the process of profiling using small, possibly seemingly irrelevant details from the crime scene and uses them to create the ‘bigger picture’, making no assumptions as to the offender 
• The BUA relies on computational/statistical analysis and database records in the collection of crime scene minutiae
• The BUA is popularly known as the British approach to profiling as it was devised by a British researcher, Professor David Canter
• Canter’s methods involve the cross-referencing of crime-scene details to determine the composition of the offender profile via use of statistical methods,  known as investigative psychology (IP)
• The statistical system used in the BUA is known as ‘smallest space analysis’ using evidence from the crime scene which renders a correlation of the behaviours that occurred most frequently across offences

• IP may be broken down into the following components:
  • Interpersonal coherence - this has some relevance to the top down approach concept of modus operandi i.e. how the offender behaved at the crime scene, their treatment of the victim before, during and after the offence (this is based on the idea that ‘how someone does anything is how they do everything’ - crime scene behaviours will reflect everyday behaviours)
  • The significance of time and place - where was the crime committed - town, city, village, indoors/outdoors? When was it committed - at night, in the early morning, in summer/winter?
  • Forensic awareness - has the offender been careful to leave no trace evidence (e.g. DNA, fingerprints etc), have they avoided CCTV etc?

The BUA focuses on practical issues such as the location of crimes which may be linked.

• Geographical profiling (GP), also known as ‘crime mapping’ is a key aspect of the BUA, first developed by Rosso (1997)
• GP is used to analyse the geographical locations of offences which appear to be linked i.e. committed by the same offender
• GP can be used alongside IP to develop a profile as together these two systems form a complete picture of the likely offender (or offenders - some serial crimes have been carried out by more than one person as was the case with John Duffy, see below)
• GP operates along the assumption that serial offenders fall into two categories: marauders or commuters
• Marauders commit crimes within their neighbourhood, not far from where they live/work possibly as this is where they feel safe plus they are more likely to know the escape routes to exit the scene as quickly as possible
• Commuters commit crimes away from their neighbourhood which may be as a result of travelling a lot for work or as a way of avoiding detection
• GP uses the ‘circle theory’ which is applied by drawing a circle around the seemingly linked offences: the offender is thought to live somewhere inside the circle

Circle theory places the offender at the heart of their crimes.

• One of the first cases of the successful use of GP was in the identification of the ‘Railway Rapist’ John Duffy 
• Duffy had committed several rapes and murders around railway stations in North London in the 1980s
• David Canter was asked to analyse the crimes and to come up with a profile of the likely offender
• Canter studied the location of the crimes using circle theory: this helped him to identify where the attacker lived and the resulting pattern suggested that the offender was a marauder 
• Here is Canter’s profile alongside facts about Duffy which emerged once he had been arrested: 

• Years later (2001) it emerged that Duffy had a partner in these crimes - David Mulcahy - which is something that the police had suspected but which they had no proof of until Duffy confessed this detail in prison

Research which investigates the bottom-up approach to offender profiling

• Canter (1996) - spatial activity based on 45 males convicted of sexual assaults showed that 87% of the sample were marauders, using their home base to carry out the attacks plus they travelled the same distance from their home  to carry out each offence 
• Beauregard (2009) - recommends that IP be used to identify sex offenders as it has good application to these type of offences

Evaluation of the bottom-up approach to offender profiling

Strengths

• Much research - particularly by Canter (alone or with colleagues) supports the effectiveness of the BUA in identifying and (more importantly) apprehending dangerous offenders
• The BUA takes a more objective approach than the TDA (which uses as its basis the results of interviews conducted over 50 years ago) with its use of statistical methods which means that it is more reliable than other approaches to profiling 

Weaknesses

• Kocsis et al. (2002) tested the profiling skills of various police professionals compared to a sample of Chemistry students: the Chemistry students produced the most accurate profiles (interestingly the more experienced the police professionals were, the more inaccurate their profiles were) hence profiling may involve little more than guesswork
• When profiling goes wrong it can be catastrophic: British psychologist Paul Britton’s profile of the killer of Rachel Nickell completely derailed the police investigation and resulted in the murderer going on to claim more victims