ArticlePDF Available

Abstract

Purpose . This paper is concerned with case linkage, a form of behavioural analysis used to identify crimes committed by the same offender, through their behavioural similarity. Whilst widely practised, relatively little has been published on the process of linking crimes. This review aims to draw together diverse published studies by outlining what the process involves, critically examining its underlying psychological assumptions and reviewing the empirical research conducted on its viability. Methods . Literature searches were completed on the electronic databases, PsychInfo and Criminal Justice Abstracts, to identify theoretical and empirical papers relating to the practice of linking crimes and to behavioural consistency. Results . The available research gives some support to the assumption of consistency in criminals' behaviour. It also suggests that in comparison with intra‐individual variation in behaviour, inter‐individual variation is sufficient for the offences of one offender to be distinguished from those of other offenders. Thus, the two fundamental assumptions underlying the practice of linking crimes, behavioural consistency and inter‐individual variation, are supported. However, not all behaviours show the same degree of consistency, with behaviours that are less situation‐dependent, and hence more offender‐initiated, showing greater consistency. Conclusions . The limited research regarding linking offenders' crimes appears promising at both a theoretical and an empirical level. There is a clear need, however, for replication studies and for research with various types of crime.
Copyright © The British Psychological Society
Reproduction in any form (including the internet) is prohibited without prior permission from the Society
The psychology of linking crimes: A review
of the evidence
Jessica Woodhams*, Clive R. Hollin and Ray Bull
University of Leicester, UK
Purpose. This paper is concerned with case linkage, a form of behavioural analysis
used to identify crimes committed by the same offender, through their behavioural
similarity. Whilst widely practised, relatively little has been published on the process of
linking crimes. This review aims to draw together diverse published studies by outlining
what the process involves, critically examining its underlying psychological assumptions
and reviewing the empirical research conducted on its viability.
Methods. Literature searches were completed on the electronic databases,
PsychInfo and Criminal Justice Abstracts, to identify theoretical and empirical papers
relating to the practice of linking crimes and to behavioural consistency.
Results. The available research gives some support to the assumption of consistency
in criminals’ behaviour. It also suggests that in comparison with intra-individual variation
in behaviour, inter-individual variation is sufficient for the offences of one offender to be
distinguished from those of other offenders. Thus, the two fundamental assumptions
underlying the practice of linking crimes, behavioural consistency and inter-individual
variation, are supported. However, not all behaviours show the same degree of
consistency, with behaviours that are less situation-dependent, and hence more
offender-initiated, showing greater consistency.
Conclusions. The limited research regarding linking offenders’ crimes appears
promising at both a theoretical and an empirical level. There is a clear need, however,
for replication studies and for research with various types of crime.
In recent years, police forces in many countries have moved away from a reactive style
of policing to a more proactive and intelligence-led style. The rationale behind this shift
in approach is to ensure that the police’s limited resources are effectively deployed,
targeting the minority of offenders believed to commit the majority of crime (Innes,
Fielding, & Cope, 2005). Crime analysts, who are usually civilian police personnel, are
employed by the police in numerous countries. The membership of an international
mailing list for crime analysts, Law Enforcement Analyst, illustrates well this
multinational interest with representatives from the USA, UK, Australia, New Zealand,
Canada and the Netherlands, to name a few. The role of crime analysts is to collect
* Correspondence should be addressed to Jessica Woodhams, University of Leicester, UK (e-mail: jaw38@le.ac.uk).
The
British
Psychological
Society
233
Legal and Criminological Psychology (2007), 12, 233–249
q2007 The British Psychological Society
www.bpsjournals.co.uk
DOI:10.1348/135532506X118631
Copyright © The British Psychological Society
Reproduction in any form (including the internet) is prohibited without prior permission from the Society
and analyse information (intelligence) about crime or criminals with the aim of
informing or directing future police action.
Interest in the topic of crime analysis has been recently growing. This growth is
shown by the slowly increasing number of published papers on the topic and in the
development of a small number of university courses specifically in crime analysis.
However, the psychology of crime analysis has received little comprehensive discussion.
This paper aims to address this by discussing the psychology of linking crimes, a process
in which crime analysts can be engaged for much of their time.
Initially, this paper considers the advantages of linking crimes for the police and the
criminal justice system. The steps involved in the linking of crimes are briefly outlined.
Having considered why and how linkage is carried out, theory and research from
personality psychology is reviewed, which informs consistency and inter-individual
variation in behaviour. What this particular approach means for the practice of linking
crimes and how it has been researched is outlined, illustrating how forensic
psychologists have begun to assess both the theoretical underpinning and the practice
of linking crimes. The research conducted thus far is reviewed and finally how research
and practice may progress in the future is considered.
Why is it important to link crimes?
Offenders who have committed two or more crimes of the same type are referred to as
serial offenders and the offences that a serial offender has committed are collectively
referred to as a series. Linking a series of crimes can help the police in three ways. First,
linking allows for the collation and pooling of information from various crime scenes,
potentially increasing the evidence against an offender (Grubin, Kelly, & Brunsdon,
2001). Second, if a collection of crimes is suspected of being committed by the same
individual they can be investigated together, rather than separately, allowing limited
police resources to be used more effectively. Third, evidence of crimes being linked has
been used as similar fact evidence in legal proceedings (Hazelwood & Warren, 2003;
Ormerod & Sturman, 2005).
The most reliable means of identifying crimes committed by the same offender is
by using physical evidence (Grubin et al., 2001), such as DNA profiling. In many
crimes, however, no such physical evidence exists and, when it does, DNA profiling
is both time-consuming and expensive (Davies, 1991). Furthermore, it can be
advantageous for the linking of crimes to be undertaken on a proactive basis, where
crime analysts are actively searching through a large number of unsolved offences
for similar offences that might form part of a series. This proactive action is in
contrast to a reactive approach, where crime analysts try to determine whether a
crime for which the offender is not yet known can be linked to another crime, for
which the offender is already known.
Such case linkage involves ‘identifying behavioural similarities between offences that
point to them being committed by the same perpetrator’ (Woodhams & Grant, 2006).
This process is also known as ‘linkage analysis’ (Hazelwood & Warren, 2003) and
‘comparative case analysis’ (Bennell & Canter, 2002; Merry, 2000). Rather than relying
on similarities in physical evidence, case linkage relies on evidence of behavioural
similarity. Initially, case linkage was limited to use with serious types of crime such as
stranger rape and attempted murder. However, its potential with less serious types of
crime is beginning to be recognized.
234 Jessica Woodhams et al.
Copyright © The British Psychological Society
Reproduction in any form (including the internet) is prohibited without prior permission from the Society
How are crimes linked?
The process of case linkage requires a detailed form of behavioural analysis (Davies,
1991). The first stage involves the analyst studying each crime in detail to gain an
understanding of the offender’s physical and verbal behaviour. The source of this
information is usually the victim’s account of the crime (where the crime did not result
in the victim’s death). This account can be in the form of a statement made to the police
or it can be in the form of an interview transcript. The difficulties associated with using
such sources of information have been discussed in detail elsewhere (Alison, Snook, &
Stein, 2001; Canter & Alison, 2003; Woodhams & Grant, 2006), but include the omission
and distortion of information.
From studying the available material, the analyst seeks to identify the relevant
behaviours in each offence (Grubin et al., 2001). Historically, there has been some
debate as to whether modus operandi (MO) behaviours, those acts necessary to
commit the offence successfully (Davies, 1992; Hazelwood & Warren, 2003), should be
used in the behavioural linking of crimes. Some authors warn that MO is susceptible to
change because it is a learned behaviour and hence should not be relied upon too
heavily in linking crimes (Douglas & Munn, 1992). Douglas and Munn suggest that
ritualistic or fantasy-based behaviours should be analysed instead of MO behaviours;
however, for crimes such as robbery it is questionable whether fantasy-based behaviours
will be displayed, given the nature of the crime. In addition, as discussed below, some
recent empirical studies have indicated that MO behaviours can be used successfully to
link crimes.
The next stage of the linking process requires the analyst to compare the behaviours
in each crime with the other crimes, noting any similarities and differences. This
comparison is not a simple task, since the analyst must also consider the antecedents for
each behaviour because the context in which a behaviour occurs is important. For
example, the property stolen in two robberies committed by the same offender might
appear to vary if on one occasion cash was stolen but on another jewellery was stolen.
However, when considering the context of the behaviour in the second offence, the
analyst may discover that the offender did demand cash from the victim but on
discovering they did not have any, jewellery was stolen instead. In summary, this stage
requires the analyst determining the ‘degree of match’ between the offences (Grubin
et al., 2001).
Whilst two offences might share a number of similar behaviours, the crime analyst
must also consider whether these similarities could be due to coincidence or other
factors (Grubin et al., 2001). In essence, they must weight the similarities between the
offences by considering the frequency with which each similar behaviour occurs in its
population of crimes. For example, whilst two robberies might share a common feature
of cash being stolen, this is not strong evidence of linkage if cash is the most common
property stolen in robberies. Hazelwood and Warren (2003) make the point that the
analyst must consider the frequency with which the combination of behaviours in the
offence occurs in its population of crimes.
Bennell and Canter (2002) have proposed that the practice of linking crimes shares
similarities with making medical diagnoses in that given two possible decisions for an
analyst to make (linked or unlinked) and two possible realities (actually linked, not
actually linked) there will be four possible outcomes to the task of linking crimes. These
four outcomes are correctly identifying linked crimes and correctly rejecting unlinked
crimes (hits and correct rejections), and mistakenly linking crimes by different offenders
The psychology of linking crimes 235
Copyright © The British Psychological Society
Reproduction in any form (including the internet) is prohibited without prior permission from the Society
or failing to link crimes by the same offender (false alarms and misses). This task has
parallels with other decision-making tasks in forensic psychology, such as seen in risk
assessment with offenders (Quinsey, Harris, Rice, & Cormier, 1998) where, for example,
a false alarm (or false positive) would correspond with awarding a high risk score to a
non-recidivist and a correct rejection (or true negative) would correspond with
awarding a low risk score to a non-recidivist.
Like risk assessment, case linkage can involve actuarial and clinical elements.
The first author has been involved in helping a police force develop a more actuarial
method for linking crimes using a computer program. However, even if such programs
are successful at identifying linked crimes, the ‘clinical’ expertise of the analyst is
still required since such programs are not at a stage where they can take account, for
example, of the potential impact of the context of a behaviour. Rather, computer
programs can only capture whether a behaviour occurred or not. This can be illustrated
by two hypothetical stranger rapes in which the victims were repeatedly punched
in the face. In one case the offender punched the victim each time she physically
resisted him whereas in the other the offender repeatedly punched the victim in a more
gratuitous manner which was unrelated to her resistance. A computer program
evaluating the co-occurrence of punching would find a degree of similarity between
the two rapes, whereas a crime analyst considering the context in which
punching occurred would consider the two offences to be more dissimilar on this
behaviour.
Behavioural consistency and inter-individual variation
Behavioural consistency
Regardless of whether computer programs or individual analysts are involved, the
practice of case linkage rests on the assumption that offenders are consistent in the way
they behave across situations. The question of whether people are consistent in their
behaviour may intuitively seem to have an obvious affirmative answer. However, this
same question has been the focus of research by social and personality psychologists for
many decades and answering it has been far from straightforward.
In the 1970s, the field of personality psychology suffered what has been described as
a paradigm crisis (Mischel & Peake, 1983). This crisis developed because the intuitive
stance that people show consistency in their behaviour across situations seemed in
conflict with the research evidence which was suggesting variability (Mischel, 1968).
In essence, some psychologists held the view that people were consistent in their
behaviour because of their personality traits (the person side of the debate). Others
rejected the position that people were inherently consistent and instead proposed that
the situation affected consistency (the situation side of the debate).
The person–situation debate is now at a stage where the interaction of both the
person and the situation is recognized (Mischel, Shoda, & Mendoza-Denton, 2002;
Shoda & Mischel, 2000). Researchers such as Mischel (1999) suggest that the situation
interacts meaningfully with attributes of the person through a ‘personality system’. For
example, Mischel and Shoda (1995) presented a ‘cognitive-affective personality system’
(CAPS) and similar models have been proposed by others (Greene, 1989; Meyer, 1990).
Essentially, these systems are seen as constructed of mental representations, such as
goals, expectations, beliefs, plans, strategies and memories of people and events
previously encountered, ‘whose activation or inhibition leads to the behaviours
displayed’ (Mischel et al., 2002, p. 53). The features of a situation and how a person
236 Jessica Woodhams et al.
Copyright © The British Psychological Society
Reproduction in any form (including the internet) is prohibited without prior permission from the Society
interprets them will result in the activation of certain cognitions and affects, which in
turn leads to the activation of plans, strategies and ultimately behaviours
through ‘stable associative links’ (Shoda, Lee-Tiernan, & Mischel, 2002). The
behaviour displayed therefore depends on how both the situation and person
variables interact.
The likelihood of activation (or being reminded of a behavioural strategy) is related
to the degree of similarity between the previous situation and the current situation
(Meyer, 1990). Similar situations will, it is argued, elicit similar cognitions, affect and
behaviour (Mischel, 1999). Thus, what makes two situations similar to the individual
concerned is not their physical characteristics but their psychological meaning
(Shoda, 1999). Situations that have been rated as more similar have been shown to
elicit consistent behaviour by research within the personality psychology tradition
(Funder & Colvin, 1991; Furr & Funder, 2003; Shoda, Mischel, & Wright, 1993;
Wright & Mischel, 1987).
As well as external events, internal events, such as fantasy or planning, can trigger
such personality systems (Mischel, 1999). Greene (1989) has proposed that similar goals
will trigger similar behavioural strategies and therefore result in consistent behaviour.
Greene adds that the more frequently a behavioural strategy is activated, the stronger it
becomes and the greater the likelihood that it will be activated in the future. In addition,
the amount of time that has passed since a behavioural strategy was last activated will
affect the likelihood of future activation, given that memory traces are proposed to
decay if they are not activated (Meyer, 1990). Thus, these explanations of behavioural
consistency suggest that the behaviour an individual will display depends on that
person’s current goal, the situational characteristics and the person’s past learning
experiences.
Inter-individual variation
Zayas, Shoda, and Ayduk (2002) state that ‘individuals’ CAPS networks are expected to
differ in the availability of specific cognitions or affects, as well as in the pattern and
strengths of the associations among the cognitions and affects, which determines the
ease with which they are activated’ (p. 856). In the same way that such a system would
predict consistency in an individual’s behaviour across psychologically similar
situations, the same situation might result in different behaviour for another individual
because of their different, distinct personality system.
It therefore appears from personality psychology that people are consistent in their
behaviour; however, the degree of behavioural consistency will depend on the
psychological similarity of the situations. How each person perceives a situation is
likely to vary because each person has his or her own distinct personality system based
on their own individual experiences. Inter-individual variation in behaviour in the
same situations would therefore be expected. Mischel and colleagues (Mischel, 1999;
Wright & Mischel, 1987) have proposed that people have their own distinct ‘if
(situation)–then (behaviour)’ profiles or behavioural signatures, with research giving
support to this view (e.g. Shoda, Mischel, & Wright, 1994). However, Greene (1989)
proposed that ‘because some things work in achieving goals and others do not, people
will tend to establish similar representations of action–outcome relationships’ (p. 202).
Thus, there is unlikely to be infinite variation in behavioural strategies for a particular
task, rather different people will prefer different strategies from a limited number of
possibilities depending on their individual set of competencies.
The psychology of linking crimes 237
Copyright © The British Psychological Society
Reproduction in any form (including the internet) is prohibited without prior permission from the Society
Further conditions for consistency
As well as demonstrating evidence of and explaining intra-individual behavioural
consistency and inter-individual behavioural variation, personality theorists have also
investigated factors other than the psychological similarity of situations that affect the
degree of behavioural consistency.
Type of behaviour
Research in personality psychology has indicated that behavioural consistency can vary
depending on the type of behaviour being measured. For example, Furr and Funder
(2003) found greater consistency in ‘automatic’ behaviours, such as laughing and
gesturing, rather than planned behaviours. Greater consistency has been found for
behaviours that are ‘emitted’ by the individual, that are more spontaneous, compared to
behaviours that are ‘elicited’ from the individual, that constitute a response to a stimulus
(Funder & Colvin, 1991). It is thought that this greater consistency is due to emitted
behaviours reflecting the characteristics of the person.
Hettema and Hol (1998) made a distinction between behaviours under primary
control and those under secondary control. Primary control behaviours are those that
involve acting on the environment, reflect the needs or desires of the individual and are
goal-directed. Primary behaviours were found to be more consistent across situations:
Hettema and Van Bakel (1997) summarize this well: ‘Consistency is to be expected,
particularly :::where goal-directed persons meet situations allowing them to transform
the situation in the direction they prefer’ (p. 225).
Expertise
Hettema and Van Bakel (1997) contend that situations conducive to primary control
behaviours are particularly likely to arise where the actor is experienced in the activity,
such as where ‘Experienced professionals meet situations with which they are
professionally familiar’ (p. 225). Thus, experience with a situation or familiarity
increases behavioural consistency. The personality systems that have been proposed
can account for such findings. As noted above, Greene (1989) explained that the more
frequently a behavioural strategy is activated the stronger it becomes and the more likely
it will be activated in the future, encouraging behavioural consistency.
Temporal period
Greater behavioural consistency has been found when behaviour is observed over
shorter time spans (Pervin, 2002). This consistency may arise because less time affords
less opportunity for new learning experiences and the modification of mental
representations in the proposed personality systems. Related to this possibility is the
finding of greater behavioural consistency in adulthood compared to childhood (Pervin,
2002). Greater variation in behaviour during childhood is probably a result of changes in
the personality system during periods of development (Mischel, 1999).
Summary
In summary, research in personality psychology has suggested some basic principles of
behavioural consistency. First, people do show a degree of consistency in their
behaviour but this is most evident when a person encounters psychologically similar
238 Jessica Woodhams et al.
Copyright © The British Psychological Society
Reproduction in any form (including the internet) is prohibited without prior permission from the Society
situations. Second, the proposed individual nature of the theoretical personality
systems, such as CAPS (Mischel & Shoda, 1995) and the associated research base,
suggests that different people will vary in the way they behave in the same situation.
Third, behaviours vary in their degree of consistency. Fourth, behavioural consistency is
proposed to increase with experience in an activity. Finally, the temporal period over
which behaviour is observed can affect the degree of behavioural consistency, with
greater consistency being observed over shorter time-spans and during adulthood
compared to childhood. These five principles have a number of implications for the
effectiveness of case linkage and for research assessing this practice.
Applying personality psychology to the linking of crimes
Behavioural consistency
For crime analysts to be able to identify different crimes committed by the same
perpetrator, criminals must show a degree of consistency in the way that they behave
when committing their crimes. The practice of case linkage therefore assumes
behavioural consistency across crimes, what Canter (1995) has termed the ‘offender
consistency hypothesis’.
The theory and research from personality psychology would suggest support for case
linkage’s assumption of behavioural consistency, providing that an offender’s crimes have
similar psychological meaning to him or her. This position also suggests a greater potential
for case linkage than for some types of offender profiling. Case linkage is more specific in
that it assumes consistency over the same type of situations (i.e. the same type of crime
such as a series of stranger rapes), whereas some predictions made by offender profilers
assume a degree of consistency between the way someone behaves during a crime and
the way they behave in their everyday lives. For example, Canter (2000) stated that crime
‘is an extreme form of non-criminal activity, and is therefore likely to reflect variations that
occur in ordinary day-to-day interpersonal activities’ (p. 39).
Case linkage investigates consistency in behaviour across a series of the same type of
crime. An offender’s crimes might therefore be expected to hold a similar psychological
meaning to him or her. However, in crimes with an interpersonal element (interaction
between the offender and victim), the behaviour of a victim might change the
psychological meaning of a situation for an offender. For example, physical resistance
from a victim of sexual assault might impact on an offender’s sexual fantasy that the
victim is a consenting partner, and hence change the meaning of the situation for him or
her, which could result in behavioural change.
The notion of personality systems suggests that crime analysts should only expect a
degree of consistency in an offender’s behaviour across a series of crimes. As explained
above, such personality theories suggest that the behaviour an individual displays will
depend on their current goal, the situational characteristics and their past learning
experiences. Therefore, changes in offending behaviour could be expected as a result of
learning from previous victims, contact with the police and legal processes, media
publicity, victim resistance and the offender’s perception of the victim (Davies, 1992).
Inter-individual variation
For crime analysts to be able to assign offences to a common offender, criminals must
not only be consistent, they must be consistent in a relatively individualistic manner. If
offenders were consistent but they all behaved in similar ways, it would be impossible to
The psychology of linking crimes 239
Copyright © The British Psychological Society
Reproduction in any form (including the internet) is prohibited without prior permission from the Society
distinguish one individual’s offences from another individual’s. Canter (1995), whilst
not articulating it as an assumption of case linkage, alluded to this situation: ‘The way an
offender commits crimes is characteristic of that individual and distinguishable from the
offence “style” of other offenders committing similar crimes’ (p. 349). Case linkage also
therefore requires inter-individual variation in behaviour, as well as behavioural
consistency.
The theory and research from personality psychology also gives support to the
assumption of inter-individual variation in committing crimes. Personality psychologists
have demonstrated that people have stable but distinctive ways of behaving in response
to situations due to their individual personality systems (e.g. Shoda et al., 1994). These
findings appear very positive for the linking of crimes; however, other findings from
personality psychology suggest a need for some potential modifications to the practice
of case linkage.
Further conditions for consistency
Type of behaviour
Personality psychologists have found that behavioural consistency can vary depending
on the behaviour that is being observed. Thus, behaviours that are emitted and under
primary control show the greatest consistency (Funder & Colvin, 1991; Hettema & Hol,
1998). Some criminal behaviour would seem to reflect such behaviour since it could be
considered manipulative and desire- or goal-driven. It could also be suggested that the
relative imbalance of power seen in some interpersonal crimes, such as rape or robbery,
may provide just the occasion for an offender to manipulate or act on his or her
environment.
Whilst it therefore appears that criminal behaviour has the potential to show
consistency, there is likely to be variation within criminal behaviour with regards to its
consistency. For example, Bennell and Canter (2002) explain that the behaviour in
burglary regarding what property is stolen very much depends on what property
happens to be available to steal. Such theft behaviours would seem to fit within Funder
and Colvin’s (1991) less consistent elicited group of behaviours, as what property is
stolen would usually be stimulus-specific.
Expertise
The effect of expertise on behavioural consistency reported in personality psychology,
with greater behavioural consistency associated with increased expertise, suggests that
linking crimes should be successful if serial offenders can be considered relatively
expert in their type of offending. Consistency in offending behaviour would be
expected to increase the more experienced an offender becomes.
Temporal period
Findings from personality psychology that greater behavioural consistency is observed
over shorter periods of time would suggest that crime analysts should expect greater
behavioural consistency in series of crimes that have occurred in close succession. That
behaviour will be less consistent during periods of personality development has
implications for the linking of juvenile crime. Overall, juvenile offenders might be
expected to be less consistent than adult offenders and therefore the linking of juvenile
240 Jessica Woodhams et al.
Copyright © The British Psychological Society
Reproduction in any form (including the internet) is prohibited without prior permission from the Society
crime may be less accurate. Alternatively, juveniles may show consistency in some
behaviours but not others. For example, during a period of sexual maturation greater
variation in sexual offending behaviours might be expected. In relation to aggressive
behaviour, Shoda et al. (1993) found children to be consistent in this type of behaviour
across psychologically similar situations (similar with regards to their psychological
competency demands). If juveniles are consistent in some behaviours but not others,
case linkage could be used with juvenile crime providing the identified appropriate
behaviours were used.
In summary, the non-forensic findings from personality psychology are relevant to
this specialist area of forensic psychology and suggest avenues for investigation. In
relation to the assumption of behavioural consistency it is suggested that consistency in
criminal behaviour would be expected across crimes, providing the situations are
psychologically similar to the criminal. For the assumption of inter-individual variation
in behaviour, the individual nature of each criminal’s personality system would be
expected to result in different criminals behaving in different ways in the same situation.
Criminals’ behavioural consistency should be expected to vary depending on the type of
behaviour being considered. Criminals would be expected to show greater consistency
in their criminal behaviour the more experienced they become in offending. Juvenile
criminals might be expected to show less consistency in their criminal behaviour
compared to adult offenders. Finally, crime analysts would be more likely to observe
greater consistency in criminal behaviour over shorter time periods of offending.
Forensic psychologists have recently begun to investigate the validity of such
predictions for criminal behaviour and the linking of crimes.
Empirical evidence for the psychology of linking crimes
Behavioural consistency
Not all forensic studies that have considered whether offenders show consistency in
aspects of their offending behaviour have been conducted with the aim of assessing case
linkage. Nevertheless, such studies do show some limited evidence of behavioural
consistency. For example, researchers have reported stability in aspects of sex
offenders’ behaviours regarding the type of victim and type of sex offence (Sjostedt,
Langstrom, Sturidsson, & Grann, 2004; Soothill, Francis, Sanderson, & Ackerley, 2000).
These studies are limited, however, in their applicability to the practice of linking crimes
because most of the behaviours investigated are too broad and are unlikely to possess
much discriminatory power (e.g. type of sex offence committed, such as rape vs.
indecent assault or contact vs. non-contact offence).
Knight, Warren, Reboussin, and Soley (1998) investigated the behavioural
consistency of serial rapists. They found some behaviours, such as the presence of a
firearm and the cutting or slashing of clothing, to be more consistent than others, but
applying these findings to linking crimes is problematic. It becomes apparent from
Knight et al. that some behaviours which would be expected to occur in a rape have not
been investigated, for example, more minor sexual acts and escape behaviours. In
addition, some of the coding categories used by Knight et al. appear to overlap, such as
‘planning of rape’ and ‘extensive planning’.
In relation to serial homicide offenders, Salfati and Bateman (2005) have demonstrated
stability in whether offenders’ homicides reflect either an expressive or instrumental
behavioural theme. Expressive homicides are described as being more emotionally
The psychology of linking crimes 241
Copyright © The British Psychological Society
Reproduction in any form (including the internet) is prohibited without prior permission from the Society
driven, as represented by behaviours such as the victim being tortured, whereas
instrumental homicides are those where the victim is used to ‘satisfy an ulterior aim such
as sex’ (p. 133). Similarly, Yokota and Canter (2004) identified four themes/types of
burglary and found most burglars to be relatively consistent in the type of burglary
(residential, public, industrial/storage and commercial) they committed. How applicable
and useful the findings are from these studies for the practice of case linkage is
questionable. Measuring behavioural consistency at the thematic level is potentially too
general to be of use in distinguishing offenders from one another. In addition, in Salfati
and Bateman’s study of serial homicide, when more stringent methods of classifying
offenders as expressive or instrumental were used, between 36 and 59% of offences could
not be classified. The inability to classify all offences into a behavioural theme would be a
significant problem for the linking of crimes as some crimes would have to be excluded
from the process due to being unclassifiable and linked crimes could be missed.
Research with the specific aim of assessing the linking of crimes began with a study
of residential burglaries reported by Green, Booth, and Biderman (1976). They
investigated whether 15 burglaries known to have been committed by three lone
offenders (five burglaries each) could be correctly linked using cluster analysis. They
considered six modus operandi categories: location of entry, side of entry, location on
block, method of opening, day of week, type and value of property taken. The cluster
analysis allocated all but one of the offences to the correct offender on the basis of
behavioural similarity. However, it should be noted that the three series of burglaries
were specifically selected because their MOs were distinctive from one another.
Eight other studies were identified by the authors which have been conducted since
1976 assessing whether crimes could be linked on the basis of behavioural similarity
(Bennell & Canter, 2002; Bennell & Jones, 2005; Canter et al., 1991; Grubin et al., 2001;
Hammond, 1990 as cited in Canter, 1995; Santtila, Fritzon, & Tamelander, 2005; Santtila,
Junkkila, & Sandnabba, 2005; Woodhams & Toye, 2007). These studies have sampled
arson, residential and commercial burglary, commercial robbery and sex offences, and
they have all indicated that offenders are consistent to some degree in their offending
behaviour, therefore supporting the first assumption of case linkage.
Most studies of behavioural consistency have used a coefficient of similarity, called
Jaccard’s coefficient (Bennell & Canter, 2002; Bennell & Jones, 2005; Canter et al.,
1991;
1
Woodhams & Toye, 2007). This statistic measures the similarity between two
events or pairs of crimes, and its value ranges from 0 to 1, with 0 indicating no similarity
and 1 indicating a perfect match (Bennell & Canter, 2002). The studies using Jaccard’s
coefficient have compared the similarity scores of linked pairs of crimes, i.e. those
known to have been committed by the same offender, with the scores for unlinked
crime pairs, i.e. those known to have been committed by different offenders. The
former were found to be more similar than the latter, providing support for the
assumption of consistency in criminal behaviour.
Inter-individual variation
In relation to the second assumption of case linkage, inter-individual variation, seven
studies have gone on to determine, within a crime type, whether offences committed by
one offender can be distinguished from offences committed by different offenders.
1
It could not be determined from their report exactly which similarity coefficient was used. However, it is likely that this was
Jaccard’s coefficient given its use in subsequent analyses by the principal author.
242 Jessica Woodhams et al.
Copyright © The British Psychological Society
Reproduction in any form (including the internet) is prohibited without prior permission from the Society
As outlined above, theories from personality psychology would predict that criminals
have distinctive personality systems which should produce inter-individual variation in
behaviour.
The methods used by researchers to investigate this assumption have varied: for
example, Canter et al. (1991) assessed whether offences could be linked by plotting four
offenders’ series of three stranger sex offences on to a multiple scalogram analysis
(MSA), which correctly clustered the offences committed by the same offender.
Grubin et al. (2001) also investigated the possibility of linking stranger sex offences
using behavioural similarity, but adopted a different approach to Canter et al. (1991).
Using similarity within four behavioural domain types, control, style, sex and escape, a
computer algorithm selected from the overall data set of offences, the 10% of offences
that were the most similar to the offence in question. It was then determined how many
offences by the same offender were identified in this top 10% and this figure was
compared with the number that would have been expected by chance. For all but two
of the 81 series in the data set, the number of linked offences featuring within the top
10% were significantly more than the number expected by chance.
Santtila, Junkkila et al. (2005) used a different approach. Prior to assessing linkage,
they purposefully selected those behaviours that showed greatest consistency within
the known rapists’ series. When making their selection they also considered whether the
behaviours differentiated between offenders successfully. Typically studies assessing the
ability to link crimes have used all the offence behaviours displayed by the offenders
except those behaviours whose occurrence is very common or very rare in the sample
of offences. Santtila, Junkkila et al. then assessed case linkage in two ways. First, the
offences from each series were plotted using multidimensional scaling analysis, where
similarity on the selected offence variables was indicated by their spatial proximity. For
each offence, the five offences closest to it were analysed to determine if they contained
another crime committed by the same offender and for over 40% of the offences this was
the case. (Nearly 60% of the time a crime from the same series would have been found
within the closest 10 offences.) Second, for each offence the proportion of behaviours
from each of four behavioural themes identified, i.e. expressive involvement, deceptive
involvement, sexual hostility and physical hostility, was used as the basis of linkage.
Discriminant function analysis was used to determine whether by using this information
each offence could be allocated to its offender. The percentage of cases correctly
classified was greater than chance and when considering the top 10 most likely
offenders to which each offence might belong, for 86% of the offences the true offender
was within the top 10.
Santtila, Fritzon et al.’s (2005) study of arson took a similar approach to Santtila,
Junkkila et al. (2005). A principal components analysis (PCA) of crime scene behaviour
variables was conducted to identify six themes of behaviour. Summary scores from the
PCA were entered into a discriminant function analysis to assess whether each case
could be allocated to its correct series. (The offences each offender had committed were
already known.) One-third of the cases were allocated to the correct series and for
approximately one-half of the sample, the series to which the case belonged was within
the 10 most probable series.
The remaining studies have adopted similar approaches in assessing the practice of
case linkage (Bennell & Canter, 2002; Bennell & Jones, 2005; Woodhams & Toye, 2007).
In these studies the behaviours displayed in the offences were allocated to behavioural
domains reflecting collections of behaviours serving a similar purpose. Bennell and
Canter’s and Bennell and Jones’ studies of burglary used the domains of target selection,
The psychology of linking crimes 243
Copyright © The British Psychological Society
Reproduction in any form (including the internet) is prohibited without prior permission from the Society
entry behaviours and property stolen. Woodhams & Toye’s study of commercial robbery
used the domains of target selection, planning and control behaviours. In all three
studies, similarity was measured for behavioural domains using Jaccard’s coefficient and
logistic regression analyses and ROC analyses were used to determine how accurately
linked pairs of offences were differentiated from unlinked pairs on the basis of similarity
within behavioural domains. All three studies showed that linked pairs could be
differentiated from unlinked pairs on the basis of behavioural similarity. However, the
accuracy with which this was achieved varied depending on the behavioural domain
used.
Type of behaviour
Bennell and Canter (2002) and Bennell and Jones (2005) when assessing the more
typical MO-type behaviours for burglary (such as entry behaviours) found little
difference in their ability to distinguish linked and unlinked crimes. However,
Woodhams and Toye (2007) found that control behaviours in the offence of commercial
robbery, those used to control the victim (e.g. use of weapon), outperformed both
planning and target selection behaviours in the accurate identification of linked and
unlinked offences. This result parallels a similar finding by Grubin et al. that, for stranger
sex offences, control and escape behaviours showed greater consistency than sexual or
style behaviours. Grubin et al. suggested two reasons for this finding. First, these
particular domains may capture behaviours that are more inherent to the offender, or as
Bennell and Canter suggest, behaviours that are less situation-dependent. This point
relates to findings noted above, where personality psychologists have found greater
behavioural consistency for behaviours emitted by the individual (Funder & Colvin,
1991). Second, behaviours such as control and escape behaviours are less dependent on
accurate victim recall (Grubin et al., 2001). A further possibility is that the police, when
interviewing a victim or writing a statement, are aiming to demonstrate that an offence
actually occurred. In both robbery and sexual assault, the police will seek to establish
how the victim was controlled and how the suspect made his or her escape because this
could have implications for defining an offence (for example, as robbery rather than
theft), for the conduct of the current investigation and for the likelihood of obtaining a
conviction. It follows that more details may therefore be available about such
behaviours.
Geographical proximity of crimes can also be considered as an offence behaviour
since offenders can usually choose where to commit their crimes. The geographical
distance between burglaries was the best predictor of linkage as reported by both
Bennell and Canter (2002) and Bennell and Jones (2005). Woodhams and Toye (2007)
also assessed distance between commercial robberies as a predictor of linkage. Unlike
Bennell and colleagues, distance did not outperform the other predictors, rather
similarity in control-type behaviours performed marginally better than distance between
crimes. Control-type behaviours could not be included as a predictor in Bennell and
colleagues’ studies since such behaviours are unlikely to be observed in burglary (the
type of offence they were investigating).
If, as it seems from the research conducted thus far, the consistency of criminal
behaviour is greater if a behaviour is emitted rather than elicited, case linkage could
potentially be more successful with some crime types compared to others. With
interpersonal-type crimes it is likely that the offender will be responding to his or her
victim’s behaviour and these responses would constitute an elicited rather than an
244 Jessica Woodhams et al.
Copyright © The British Psychological Society
Reproduction in any form (including the internet) is prohibited without prior permission from the Society
emitted behaviour. In such crimes, the crime analyst will therefore need to take into
account the victim’s behaviour.
If, as suggested from the research, some criminal behaviours show greater
consistency than others, and are therefore more accurate predictors of linkage, this has
implications for the data collation role of crime analysts. The data collation aspect of
their role is particularly time-consuming (Innes et al., 2005) and being able to prioritize
certain data for collation would be advantageous (Bennell & Canter, 2002). In addition,
the process of linking crimes can put considerable cognitive load on the analyst
(Santtila, Korpela, & Ha¨kka¨nen, 2004) so that focusing on a smaller number of
behaviours thought to be accurate predictors of linkage could reduce this load. In
addition, crime analysts would want to avoid using inaccurate predictors of linkage
which could result in false alarms or misses.
Variation in the accuracy of predictors across different types of crime suggests that
the best predictors for each crime type remain to be established. However, the accuracy
of distance between crimes as a predictor of linkage seems to generalize across two
crime types, commercial robbery and burglary, indicating that further studies will be
required to determine if this is a predictor which should be prioritized. An additional
avenue for future research would be to assess whether offender speech can also be used
to determine linkage, with its potential use in case linkage having long been recognized
(Davies, 1992).
Expertise
Whether expertise in a crime type increases behavioural consistency has received limited
research attention. Grubin et al. (2001) discuss the effect of expertise on behavioural
consistency, in relation to length of the series. They found the likelihood of a stranger sex
offender displaying an identical domain type across a series (e.g. control type 1, control
type 1, control type 1) to decrease with increasing series length. Nevertheless, they did
note a high degree of consistency in series of five sexual assaults; however, this was
limited to only certain types of behaviour (such as control behaviours), rather than others
(such as sexual behaviours). It is possible that, for sexual offences, whilst fantasy-based
sexual behaviours might evolve across a series, offenders develop expertise and show
more consistency in their control behaviours. This possibility requires investigation.
However, there is an inherent difficulty in investigating expertise in criminal
behaviour. Typically, researchers study samples of solved crimes with apprehended
offenders. If offender A has been convicted of four crimes in the sample and offender B has
committed two crimes in the sample, offender A would appear to have greater expertise
and hence would be predicted to show greater consistency in his or her crimes. However,
it is quite possible that offender A has been convicted for all the crimes that he or she has
committed, whereas offender B has 10 undetected crimes. In such a situation, it is
offender B who has greater expertise than offender A. Researchers investigating the effect
of criminal expertise on behavioural consistency may therefore wish to consider taking a
within-offender approach, comparing the behavioural consistency of an offender earlier
in his or her series to behavioural consistency later in the series.
Temporal period
With regards to whether crimes committed by an offender in greater temporal
proximity show greater behavioural consistency, Grubin et al. (2001) found that
The psychology of linking crimes 245
Copyright © The British Psychological Society
Reproduction in any form (including the internet) is prohibited without prior permission from the Society
the inclusion of temporal proximity increased the hit rate (identifying linked crimes),
but at the same time it increased the miss rate (failure to identify linked crimes). This
topic has received limited attention thus far and is therefore an area for future research.
With regard to whether the developmental stage of the offender affects behavioural
consistency, research is under way at the University of Leicester investigating the
consistency of juvenile serial stranger sex offenders’ offending behaviour. Cross-
validation research will however be needed with other samples of sex offenders and
investigation will be required with other crime types.
Conclusions
In conclusion, case linkage is a form of behavioural analysis involving crime analysts
making predictions about whether offences have a common offender based on their
assessment of behavioural similarity across crimes. Analysts are therefore making
predictions about the stability of an offender’s behaviour across situations, a task which
personality psychologists have been studying for decades. The theories underlying case
linkage can therefore be grounded in personality psychology in contrast to other
forensic psychological practices, such as offender profiling, that have been criticized for
not explicitly considering the assumptions on which they rest (e.g. Alison, Bennell,
Mokros, & Ormerod, 2002). In contrast, the preliminary research conducted thus far on
case linkage is directly testing the validity of the assumptions of behavioural consistency
and inter-individual behavioural variation.
In both empirical research and in an applied context, behavioural similarity has been
used successfully to identify serial crime. Rather than being limited to use with
serious crime, case linkage is now being applied to volume crime, such as burglary and
robbery. The findings of Bennell and Canter (2002), Bennell and Jones (2005) and
Woodhams and Toye (2007) suggest that earlier concerns about using MO behaviours in
linking crimes may have been unfounded but they should be regarded with caution until
they have been sufficiently cross-validated. There are a number of factors which can
affect the consistency of an offender’s behaviour which mean that it is highly unlikely
that the identification of serial crime in this way will ever be perfect. Fantasy can evolve
and the offender’s mood, the timing and the external circumstances of an offence can
influence the behaviour an offender displays at the crime scene (Hazelwood & Warren,
2003). As noted by Grubin et al. (2001), a consequence of incorrectly linking crimes is
the inhibition of police investigations. Therefore case linkage evidence should be
presented in a suitably cautious manner.
Comparisons between studies assessing linkage are currently difficult to make due to
the variation in statistical methods used in assessing linkage accuracy. These various
methods require comparison to determine which method is the most appropriate
(Bennell & Jones, 2005). The approach initially adopted by Craig Bennell (Bennell, 2002;
Bennell & Canter, 2002) would seem particularly appropriate because it allows for the
consideration of all four decision outcomes associated with this process. In addition,
ROC analysis is being used more often in other areas of forensic psychology to
determine decision-making accuracy (see Quinsey et al., 1998 for some examples) and
this statistical technique clearly has other advantages for assessing the linking of crimes
that are outlined elsewhere (Bennell, 2002; Bennell & Canter, 2002; Bennell & Jones,
2005). For some measures of similarity, the shared absence of a behaviour would
contribute to how similar two crimes were determined to be. The absence of
246 Jessica Woodhams et al.
Copyright © The British Psychological Society
Reproduction in any form (including the internet) is prohibited without prior permission from the Society
a behaviour in a victim’s account does not mean that the behaviour did not occur
(Bennell & Canter, 2002), rather the victim may fail to remember that a behaviour did
occur, or the interviewer may fail to elicit or record this information. When choosing a
measure of similarity, researchers would be advised to consider whether the measure of
similarity selected includes joint absences of behaviour in its calculation. Jaccard’s
coefficient is one such measure that does not take account of joint absences in this way.
Thus far, the research conducted assessing case linkage has taken a retrospective
approach, assessing linkage with crimes that have already been solved. As noted by
Bennell and Canter (2002), there is an inherent difficulty with this because behavioural
similarity might be one reason why the offences have been linked and solved. One
further avenue for future research is to take a more prospective approach to assessing
case linkage.
Research in personality psychology has indicated that people have stable but
distinctive ways of reacting to situations that can be understood as ‘if (situation)–then
(behaviour)’ contingencies (Mischel, 1999). Personality psychologists have called for
researchers to begin considering behavioural consistency in this way (Shoda, 1999).
Whilst crime analysts will consider the context in which a behaviour has occurred when
conducting case linkage, at present, research in linking crimes has been limited to
considering the ‘then’ part of this equation. In other words, researchers have coded for
the absence or presence of distinct behaviours in an offence rather than considering the
context in which an offence behaviour is or is not displayed. One potential obstacle to
progressing case linkage research in this way is whether there is the detail in victim
accounts to always determine the ‘if’ part of the contingency. With this potential
obstacle in mind it is reassuring that the research completed thus far suggests that
crimes can be linked with some accuracy based only on the ‘then’ part of the equation.
Accuracy might, however, be improved further if researchers can consider the whole if
(situation)–then (behaviour) equation.
Research in personality and forensic psychology is also suggesting that some
modifications may be required to the case linkage process to ensure its appropriateness
for different types of crime and criminal. For example, it is suggested that some
behaviours may be more accurate predictors of linkage than others and hence could be
prioritized in database population and analysis. However, it seems likely that the best
predictors will vary with crime type and further research is needed in this area before it
can inform the practice of case linkage.
References
Alison, L., Bennell, C., Mokros, A., & Ormerod, D. (2002). The personality paradox in offender
profiling: A theoretical review of the processes involved in deriving background
characteristics from crime scene actions. Psychology, Public Policy and Law,8, 115–135.
Alison, L. J., Snook, B., & Stein, K. (2001). Unobtrusive measurement: Using police information for
forensic research. Qualitative Research,1, 241–254.
Bennell, C. (2002). Behavioural consistency and discrimination in serial burglary. Unpublished
doctoral dissertation, University of Liverpool, Liverpool.
Bennell, C., & Canter, D. V. (2002). Linking commercial burglaries by modus operandi: Tests using
regression and ROC analysis. Science and Justice,42, 1–12.
Bennell, C., & Jones, N. (2005). Between a ROC and a hard place. International Journal of
Investigative Psychology and Offender Profiling,2, 23–41.
The psychology of linking crimes 247
Copyright © The British Psychological Society
Reproduction in any form (including the internet) is prohibited without prior permission from the Society
Canter, D. (1995). Psychology of offender profiling. In R. Bull & D. Carson (Eds.), Handbook of
psychology in legal contexts (pp. 343–355). Chichester: Wiley.
Canter, D. (2000). Offender profiling and criminal differentiation. Legal and Criminological
Psychology,5, 23–46.
Canter, D., & Alison, L. J. (2003). Converting evidence into data: Use of law enforcement archives
as unobtrusive measurement. Qualitative Report,8, 151–176.
Canter, D., Heritage, R., Wilson, M., Davies, A., Kirby, S., Holden, R., et al. (1991). A facet approach
to offender profiling: Volume 1. Final unpublished report to the Home Office.
Davies, A. (1991). The use of DNA profiling and behavioural science in the investigation of sexual
offences. Medicine, Science and Law,31, 95–101.
Davies, A. (1992). Rapists’ behaviour: A three-aspect model as a basis for analysis and the
identification of serial crime. Forensic Science International,55, 173–194.
Douglas, J. E., & Munn, C. (1992). Violent crime scene analysis: Modus operandi, signature and
staging. FBI Law Enforcement Bulletin,61, 1–10.
Funder, D. C., & Colvin, C. R. (1991). Explorations in behavioral consistency: Properties of
persons, situations and behaviors. Journal of Personality and Social Psychology,60,
773–794.
Furr, R. M., & Funder, D. C. (2003). Situational similarity and behavioral consistency: Subjective,
objective, variable-centered, and person-centered approaches. Journal of Research in
Personality,38, 421–447.
Green, E. J., Booth, C. E., & Biderman, M. D. (1976). Cluster analysis of burglary M/Os. Journal of
Police Science and Administration,4, 382–388.
Greene, J. O. (1989). The stability of nonverbal behaviour: An action-production approach to
problems of cross-situational consistency and discriminativeness. Journal of Language and
Social Psychology,8, 193–220.
Grubin, D., Kelly, P., & Brunsdon, C. (2001). Linking serious sexual assaults through behaviour:
HORS 215. London: Home Office Research Development and Statistics Directorate.
Hazelwood, R. R., & Warren, J. I. (2003). Linkage analysis: Modus operandi, ritual and signature in
serial sexual crime. Aggression and Violent Behavior,8, 587–598.
Hettema, J., & Hol, D. P. (1998). Primary control and the consistency of interpersonal behaviour
across different situations. European Journal of Personality,12, 231–247.
Hettema, J., & Van Bakel, A. P. (1997). Cross-situational consistency in a mastery condition.
Journal of Research in Personality,31, 222–239.
Innes, M., Fielding, N., & Cope, N. (2005). The appliance of science? The theory and practice of
criminal intelligence analysis. British Journal of Criminology,45, 39–57.
Knight, R. A., Warren, J. I., Reboussin, R., & Soley, B. J. (1998). Predicting rapist type from crime-
scene variables. Criminal Justice and Behavior,25, 46–81.
Merry, S. (2000). Crime analysis: Principles for analysing everyday serial crime. In D. Canter &
L. Alison (Eds.), Profiling property crime (pp. 297–318). Aldershot: Ashgate.
Meyer, J. R. (1990). Cognitive processes underlying the retrieval of compliance-gaining strategies:
An implicit rules model. In J. P. Dillard (Ed.), Seeking compliance: The production
of interpersonal influence messages (pp. 57–73). Scottsdale, AZ: Gorsuch Scarisbrick
Publishers.
Mischel, W. (1968). Personality and assessment. New York: Wiley.
Mischel, W. (1999). Personality coherence and dispositions in a cognitive-affective personality
system (CAPS) approach. In D. Cervone & Y. Shoda (Eds.), The coherence of personality:
Social-cognitive bases of consistency, variability and organisation (pp. 37–60). London:
Guilford Press.
Mischel, W., & Peake, P. K. (1983). Analyzing the construction of consistency in personality. In
M. M. Page (Ed.), Nebraska Symposium on Motivation, 1982: Personality – Current theory
and research (Vol. 30, pp. 233–262). Lincoln, NE: University of Nebraska Press.
248 Jessica Woodhams et al.
Copyright © The British Psychological Society
Reproduction in any form (including the internet) is prohibited without prior permission from the Society
Mischel, W., & Shoda, Y. (1995). A cognitive-affective system theory of personality:
Reconceptualizing situations, dispositions, dynamics and invariance in personality structure.
Psychological Review,102, 246–268.
Mischel, W., Shoda, Y., & Mendoza-Denton, R. (2002). Situation-behavior profiles as a locus of
consistency in personality. Current Directions in Psychological Science,11, 50–54.
Ormerod, D., & Sturman, J. (2005). Working with the courts: Advice for expert witnesses. In
L. Alison (Ed.), The forensic psychologists’ casebook: Psychological profiling and criminal
investigation (pp. 170–193). Cullompton, Devon, UK: Willan.
Pervin, L. A. (2002). Current controversies and issues in personality (3rd ed.). New York: Wiley.
Quinsey, V. L., Harris, G. T., Rice, M. E., & Cormier, C. A. (1998). Violent offenders: Appraising and
managing risk. Washington, DC: American Psychological Association.
Salfati, C. G., & Bateman, A. L. (2005). Serial homicide: An investigation of behavioural
consistency. Journal of Investigative Psychology and Offender Profiling,2, 121–144.
Santtila, P., Fritzon, K., & Tamelander, A. L. (2005). Linking arson incidents on the basis of crime
scene behavior. Journal of Police and Criminal Psychology,19, 1–16.
Santtila, P., Junkkila, J., & Sandnabba, N. K. (2005). Behavioural linking of stranger rapes. Journal
of Investigative Psychology and Offender Profiling,2, 87–103.
Santtila, P., Korpela, S., & Ha¨kka¨nen, H. (2004). Expertise and decision making in the linking of car
crime series. Psychology, Crime, and Law,10, 97–112.
Shoda, Y. (1999). A unified framework for the study of behavioral consistency: Bridging person
£situation interaction and the consistency paradox. European Journal of Personality,13,
361–387.
Shoda, Y., Lee-Tiernan, S., & Mischel, W. (2002). Personality as a dynamical system: Emergence of
stability and distinctiveness from intra- and interpersonal interactions. Personality and Social
Psychology Review,6, 316–325.
Shoda, Y., & Mischel, W. (2000). Reconciling contextualism within the core assumptions of
personality psychology. European Journal of Personality,14, 407–428.
Shoda, Y., Mischel, W., & Wright, J. C. (1993). The role of situational demands and cognitive
competencies in behavior organisation and personality coherence. Journal of Personality and
Social Psychology,65, 1023–1035.
Shoda, Y., Mischel, W., & Wright, J. C. (1994). Intraindividual stability in the organization and
patterning of behavior: Incorporating psychological situations into the idiographic analysis of
personality. Journal of Personality and Social Psychology,67, 674–687.
Sjostedt, G., Langstrom, N., Sturidsson, K., & Grann, M. (2004). Stability of modus operandi in
sexual offending. Criminal Justice and Behavior,31, 609–623.
Soothill, K., Francis, B., Sanderson, B., & Ackerley, E. (2000). Sex offenders: Specialists, generalists –
or both? A 32 year criminological study. British Journal of Criminology,40, 56–67.
Woodhams, J., & Grant, T. (2006). Developing a categorisation system for rapists’ speech.
Psychology, Crime, and Law, 12, 245–260.
Woodhams, J., & Toye, K. (2007). An empirical test of the assumptions of case linkage and offender
profiling with serial commercial robberies. Psychology, Public Policy and Law,13, 59–85.
Wright, J. C., & Mischel, W. (1987). A conditional approach to dispositional constructs: The local
predictability of social behavior. Journal of Personality and Social Psychology,53,
1159–1177.
Yokota, K., & Canter, D. (2004). Burglars’ specialisation: Development of a thematic approach in
investigative psychology. Behaviormetrika,31, 153–167.
Zayas, V., Shoda, Y., & Ayduk, O. N. (2002). Personality in context: An interpersonal systems
perspective. Journal of Personality and Social Psychology,70, 851–900.
Received 20 January 2006; revised version received 19 May 2006
The psychology of linking crimes 249
... Crime Linkage (CL) [1] is a multidisciplinary field that has garnered significant attention from sociologists, psychologists, statisticians, and computer scientists [2,3,4,5,6,7]. In essence, they aim to connect pairs of crimes by creating crime link associations [3]. ...
... Crime Linkage (CL) [1] is a multidisciplinary field that has garnered significant attention from sociologists, psychologists, statisticians, and computer scientists [2,3,4,5,6,7]. In essence, they aim to connect pairs of crimes by creating crime link associations [3]. The CL analysis is typically based on Modus Operandi (MO) or criminal behavior, which can be inferred from evidence found at crime scenes or other source of police data [8]. ...
... Previous literature surveys have not included machine learning approaches [3,4,5], thus, our aim is to provide an extension of research assessment on the theme. When analyzing the studies collected, we observed a general framework in the CL process, which will be also discussed in this paper. ...
Preprint
Full-text available
Crime linkage is the process of analyzing criminal behavior data to determine whether a pair or group of crime cases are connected or belong to a series of offenses. This domain has been extensively studied by researchers in sociology, psychology, and statistics. More recently, it has drawn interest from computer scientists, especially with advances in artificial intelligence. Despite this, the literature indicates that work in this latter discipline is still in its early stages. This study aims to understand the challenges faced by machine learning approaches in crime linkage and to support foundational knowledge for future data-driven methods. To achieve this goal, we conducted a comprehensive survey of the main literature on the topic and developed a general framework for crime linkage processes, thoroughly describing each step. Our goal was to unify insights from diverse fields into a shared terminology to enhance the research landscape for those intrigued by this subject.
... Accurately linking crimes can improve police investigations in numerous ways. For instance, linking crimes can provide additional data for investigative techniques such as criminal and geographic profiling, while also allowing for more efficient use of police resources (Woodhams et al., 2007). Furthermore, prosecutors have used crime linkage analysis in criminal trials as inculpatory evidence to establish the guilt of an offender, and as an aggravating factor to help secure more appropriate sentences for serial offenders (Labuschagne, 2015). ...
Article
Deciding whether two crimes have been committed by the same offender or different offenders is an important investigative task. Crime linkage researchers commonly use receiver operating characteristic (ROC) analysis to assess the accuracy of linkage decisions. Accuracy metrics derived from ROC analysis—such as the area under the curve (AUC)—offer certain advantages, but also have limitations. This paper describes the benefits that crime linkage researchers attribute to the AUC. We also discuss several limitations in crime linkage papers that rely on the AUC. We end by presenting suggestions for researchers who use ROC analysis to report on crime linkage. These suggestions aim to enhance the information presented to readers, derive more meaningful conclusions from analyses, and propose more informed recommendations for practitioners involved in crime linkage tasks. Our reflections may also benefit researchers from other areas of psychology who use ROC analysis in a wide range of prediction tasks.
... Behavioural crime linkage (BCL) uses the principles of behavioural consistency and distinctiveness to identify patterns of behaviour across a series of crimes, which can then be attributed to a serial offender. Research has shown that this type of behavioural analysis can be used to successfully link multiple crimes committed by the same offender [9]. Therefore, we studied whether incorporating questions about the offender's behaviour increases the amount of information gathered from witnesses about offences, and the offender's behaviour in particular, which in turn can be used to facilitate the application of BCL. ...
Article
Full-text available
Police interviews gather detailed information from witnesses about the perpetrator that is crucial for solving crimes. Research has established that interviewing witnesses immediately after the crime maintains memory accuracy over time. However, in some contexts, such as in conflict settings and low-income countries, witness interviews occur after long delays, which decreases survivors’ access to vital services and justice. We investigated whether an immediate interview via a mobile phone application (SV_CaseStudy Mobile Application, hereafter MobApp) developed by the Kenyan Survivors of Sexual Violence Network preserves people’s memory accuracy over time. Participants (N = 90) viewed a mock burglary and were then interviewed either immediately using MobApp or MobApp+ (which included additional questions about the offender’s behaviour) and again one week later (n = 60), or solely after a one-week delay (n = 30). We found that memory accuracy one week later was higher for participants immediately interviewed with MobApp or MobApp+ compared to those interviewed solely after a one-week delay. Additionally, memory accuracy was maintained for those interviewed with the mobile application across the one-week period. These findings indicate that the mobile phone application is promising for preserving memory accuracy in contexts where crimes are reported to the police after a delay.
... Notwithstanding the growing attention to predictive policing, this research has rarely addressed the possibility of identifying potential offenders and, more specifically, co-offending in crime, that is, when two or more individuals participate in the same crime [4]. Yet, similar problems are central in another research stream focusing on crime linkage, a process of associating two or more crimes based on evidence and offender behaviour [16,17,18]. However, this approach has largely relied on qualitative, case-specific information (e.g., identifying the modus operandi of one specific offender and linking it to crime events) and has rarely developed into more general methods [19]. ...
Article
Applications of Artificial intelligence (AI) and soft computing techniques are rapidly growing biomedical research fields. This review article summarizes the most recent advances in detecting and classifying human gait parameters developed over the last forty years. The study explains various initiatives that can help increase the role of soft computing techniques in the health sector in diagnosing and treating diseases much faster. The proposed overview describes the conventional to recent advancements in the subject, including detecting neurodegenerative diseases by applying multimodal medical picture fusion, transfer learning, and deep learning. The techniques in the current paper have been validated with the Gait in Neurodegenerative Disease Database for raw sensor force data collected for left and right foot gait signals for neurodegenerative disease and healthy subjects, which are available online, i.e., PhysioNet.
Article
While the time dimension is part of the definition of serial crime, little attention has been given to understanding offending timelines and the impact of patterns of shorter or longer delays between crimes on the evolution of series on the ability to link crimes. Given the scarcity of research focusing on the timelines of serial offenders, the current study aimed to (1) determine whether series can be classified based on their timeline trajectories (e.g., progressively shorter intervals or progressively longer), and (2) determine whether these timeline trajectories correlate with trajectories of behavioral consistency and change across series. Data included 43 homicide series encompassing 216 crime scenes. Tra-jectories were determined based on whether intervals increased, decreased, or remained consistent across series, and whether timeline trajectories and behavioral trajectories align. Results revealed that distinct timeline trajectories can reliably distinguish between series. Results also suggested a complex interaction between time between crimes and how this relates to behavioral consistency and change trajectories, thus, suggesting that time is a useful, but potentially separate dimension in the linkage process.
Article
Full-text available
E tanulmányban arra a kérdésre keressük a választ, mivel tehető könnyebbé, biztosabbá, megbízhatóbbá a bűncselekmények összekapcsolása. Az egyik megoldást a profilalkotással, konzisztenciával kapcsolatos kutatások eredményei jelenthetik, másik oldalról pedig az elemző- értékelő munka igénybevétele, szoftveres támogatása és kutatásokra támaszkodó fejlesztése. Minél megbízhatóbb kutatási eredményeket tudnak alapul venni a rendvédelmi szervek és az elemző egységek, annál nagyobb hatékonysággal végezhetik munkájukat a sorozatjelleg megállapításakor és vizsgálatakor. Tanulmányunkban áttekintjük a bűncselekmények összekapcsolásával kapcsolatos legfontosabb kutatási eredményeket, valamint betekintést nyújtunk gyakorlati bevezetésük lehetőségeibe. Publikációnk végén pedig felvetünk néhány lehetséges kutatási témát, amelyekkel modernizálható lenne a hazai elemző-értékelő tevékenység, illetve az elektronikus Modus Operandi Nyilvántartás.
Article
Full-text available
In nomothetic analyses, the cross-situational consistency of individual differences in social behavior, assessed in vivo in a camp setting, depended on the similarity in the psychological features of situations. As predicted by the social–cognitive theory of personality, idiographic analyses revealed that individuals were characterized by stable profiles of if…then…, situation–behavior relationships that formed “behavioral signatures” of personality (e.g., he aggresses when warned by adults but complies when threatened by peers. Thus, the intraindividual organization of behavior variation across situations was enduring but discriminatively patterned, visible as distinctive profiles of situation–behavior relationships. Implications were examined for an idiographic reconceptualization of personality coherence and its behavioral expressions in relation to the psychological ingredients of situations.
Article
Full-text available
Little is known about the stability of modus operandi (MO) in sexual offending. The authors studied a cohort of all sexual offenders released from prison into the Swedish community during the years from 1993 to 1997 (N = 1,303) and analyzed sexual reoffenders’ MO in terms of victim choice, offense nature, and severity, comparing prior offenses with those registered during an average 6-year follow-up. Stability in MO, explored with Cohen’s Kappa and Odds Ratios (ORs) as measures of agreement across registered sexual offenses, was high, specifically with respect to victim choice. Results are discussed in relation to sexual deviance and opportunity structure. The authors argue that assessment and management of sexual recidivism risk might benefit from information on offense MO. Furthermore, the results could inform police investigative strategies, such as linking multiple offenses committed by an unidentified offender.
Article
The Metropolitan Police Forensic Science Laboratory has been using DNA profiling in crime cases since 1987 and has established an index of results from personal samples, together with those from stains from serious unsolved cases. Despite the small numbers recorded as yet, series of unsolved sexual crimes have been detected, new cases have been added to established series and suspects have been nominated for several rapes. The laboratory also has a Sexual Assault Index which is used to identify linked cases in the absence of DNA. This is done using behavioural factors, which inevitably leads to the acquisition of knowledge about the perpetrator and is part of the process known as offender profiling. The effect of the use of DNA profiling and behavioural science on some aspects of forensic science is described.