J Quant Criminol (2015) 31:653–675
DOI 10.1007/s10940-014-9244-3
ORIGINAL PAPER
Unpacking the Victim-Offender Overlap: On Role
Differentiation and Socio-psychological Characteristics
Jean-Louis van Gelder • Margit Averdijk
Manuel Eisner • Denis Ribaud
•
Published online: 21 December 2014
Springer Science+Business Media New York 2014
Abstract
Objectives Provide insight into the victim-offender overlap and role differentiation by
examining to what extent socio-psychological characteristics, risky lifestyles/routine
activities and immersion in a violent subculture explain differences between victims,
offenders and victim-offenders. Specifically, we measure to what extent anxiety and
depression, negative peer relations, dominance, and self-control account for differences in
adolescents’ inclination towards (violent) offending, victimization or both, over and above
risky lifestyles/routine activities or immersion in a violent subculture.
Methods Building on the method proposed by Osgood and Schreck (Criminology
45:273–311, 2007), we use two waves of panel data from the Zurich Project on the Social
Development of Children and Youths, a prospective longitudinal study of adolescents in
Switzerland.
Results Incorporating socio-psychological characteristics provides a more encompassing
view of both the victim-offender overlap and victim versus offender role differentiation
than routine activities/risky lifestyles and subcultural theory alone. Specifically, sociopsychological characteristics in particular differentiate between those who take on predominantly offender roles versus those who are predominantly victims.
Conclusion Unpacking the victim-offender overlap and examining differences in sociopsychological characteristics furthers our understanding of the etiology of the victimoffender overlap.
Jean-Louis van Gelder and Margit Averdijk are both to be regarded as first author of this article.
J.-L. van Gelder (&)
Netherlands Institute for the Study of Crime and Law Enforcement (NSCR), De Boelelaan 1077a,
1081 HV, Amsterdam, The Netherlands
e-mail: jlvangelder@nscr.nl
M. Averdijk D. Ribaud
Swiss Federal Institute of Technology Zurich (ETHZ), Zurich, Switzerland
M. Eisner
University of Cambridge, Cambridge, UK
123
654
J Quant Criminol (2015) 31:653–675
Keywords Victimization Victim-offender overlap Subcultural theory Risky
lifestyles Routine activities
Introduction
Research on the association between victimization and offending has yielded strong correlations between the two (e.g., Berg et al. 2012; Hay and Evans 2006; von Hentig 1948;
Lauritsen and Laub 2007; Lauritsen et al. 1991; Ousey et al. 2011; Jensen and Brownfield
1986; Schreck et al. 2008; Singer 1981, 1986; Wolfgang 1958). As Lauritsen and Laub
(2007) note, little if any research has actually failed to demonstrate the association and it
holds across time, place, subgroups, data-sources and type of crime. Unsurprisingly, it
ranks among the most robust empirical relations in criminology (Reiss and Roth 1993).
However, in their search for common correlates, few studies have explicitly considered
that only part of the offender population also falls victim to crime and that not all victims
also engage in offending. This lack of specificity has implied a restricted ability to account
for unique processes and antecedents of overlap between offenders and victims or lack
thereof (Schreck et al. 2008). That is, a focus restricted to victim-offenders to the neglect of
how victims and offenders differ may mean losing vital information regarding the etiology
of this relation. Additionally, the more individuals tend to adopt one role over the other, the
greater the need for specific and separate theorizing and research to account for both
phenomena (Schreck et al. 2008, p. 874).
Recently, various studies (e.g., Broidy et al. 2006; Daday et al. 2005; Schreck et al.
2008; Mustaine and Tewksbury 2000) have started to address this gap in the literature and
revealed meaningful differences between victims, offenders and victim-offenders. Building
on this research and using an analytical method proposed by Osgood and Schreck (2007;
see also Schreck et al. 2008), the present study examines what factors underlie the tendency to primarily take offender or victim roles in cases of violence. That is, our focus is
not restricted to the victim-offender overlap, but in particular addresses factors that are
associated with people’s tendency towards victimization versus offending.
Extending earlier work, the present study goes beyond using routine activities/risky
lifestyles and subcultural theory as explanatory factors by also examining a series of sociopsychological characteristics, such as anxiety, depression and social isolation, that may
account for differences in offender versus victim role-taking. We hypothesize that these
characteristics can discriminate between those individuals who tend to adopt victim roles
and those who predominantly tend towards offending. In line with earlier work, we expect
that routine activities/risky lifestyles and subcultural theory explanations discriminate in
particular between the group of victim-offenders and their normative peers who have
neither been victimized nor have offended.
To examine these predictions, we use data from the Zurich Project on the Social
Development of Children and Youths (z-proso), a longitudinal study of a sample of urban
Swiss adolescents containing extensive multiwave data on both offending and victimization. Analogous to Schreck et al. (2008), we focus on violent offending and victimization
during adolescence as it has been associated with a variety of important negative life
outcomes such as school failure, substance use, and juvenile arrests, and because identifying risk factors of violent outcomes is critical with regard to adolescent development.
Below, we first briefly discuss the dominant perspectives that have been used to account
for the victim-offender overlap, i.e., routine activities/risky lifestyle theory and subcultural
123
J Quant Criminol (2015) 31:653–675
655
theory. We subsequently provide an individual differences perspective that details how
socio-psychological characteristics are likely to be related to victimization, offending or
both. This is followed by an overview of our research design, method of analysis and
presentation of the results. We conclude with a discussion of how our findings extend
previous efforts and contribute to the literature, and provide suggestions for future research.
Routine Activities/Risky Lifestyle and Subcultural Explanations
for the Victim-Offender Overlap
The first major publication to draw attention to the fact that victims and offenders may
belong to the same group of individuals was von Hentig’s (1948) The Criminal and His
Victim in which he argued that although the ‘‘doer-sufferer relation is put in our codes in
mechanical terms (…), the relationships between the perpetrator and the victim are much
more intricate (…). It may happen that the two distinct categories merge. There are cases in
which they are reversed and in the long chain of causative forces the victim to assume the
role of a determinant’’ (pp. 383–384).
Another early publication drawing attention to the overlap is Wolfgang’s (1958) analysis of incident files of homicides. This study showed that victims and offenders were
often no strangers to each other as killings were frequently the result of domestic quarrels,
altercations over money, or motivated by jealousy or revenge, each of which implicate a
prior social relationship between the parties involved. Importantly, the victims had often
been the first to use physical force against their eventual slayers (Wolfgang 1958).
These early works provided initial support for the idea that victims and offenders are not
as distinct as was generally assumed and while most crime research still tends to be either
offender-focused or victim-focused, it is now commonly understood that offenders and
victims overlap in various important ways (Jennings et al. 2012).
Routine Activities/Risky Lifestyle Theory
The most common theoretical framework to account for the victim-offender overlap is the
routine activities/lifestyle perspective (Cohen and Felson 1979; Hindelang et al. 1978). The
underlying idea is that risky lifestyles (Hindelang et al. 1978) and routine activities (Cohen
and Felson 1979) bring potential victims into contact with motived offenders and expose
them to situations conducive to victimization. In addition, Osgood et al. (1996) found that
unstructured socializing with (deviant) peers in the absence of authority figures also predicts participation in offending. Other studies report similar findings (Anderson and
Hughes 2009; Bernasco et al. 2013; Bernburg and Thorlindsson 2001; Hay and Forrest
2008; Maimon and Browning 2010).
Substance use, e.g., illicit drugs and alcohol consumption, which is also characteristic of
risky lifestyles, is yet another factor related to both victimization (e.g., Felson and Burchfield 2004; Gover 2004; Lauritsen et al. 1992; Malik et al. 1997; Vogel and Himelein
1995) and offending (e.g., Elliott et al. 1985, 1989; Zhang et al. 1997).
Subcultural Theory
An alternative perspective regularly used to account for the relation between victimization
and offending is provided by subcultural theory/subculture of violence explanations, which
posit that violence occurs predominantly among groups that hold norms that support or
123
656
J Quant Criminol (2015) 31:653–675
encourage the use of force to resolve conflicts, such as gangs (Anderson 1999; Berg et al.
2012; Berg and Loeber 2011; Cohen 1955; Jacobs and Wright 2006; Lauritsen and Laub
2007; Singer 1981, 1986). According to this perspective, individuals alternate between
offender and victim roles in areas characterized by disorganization and norms of violence
(Schreck et al. 2008).
In early support of this idea, Wolfgang’s (1958) study of homicide in Philadelphia
showed that a quick resort to physical combat is a measure of daring, courage, defense or
status and a cultural means of expression especially for lower-class males. In a similar
vein, Singer (1981, 1986) argued that the association between victimization and offending
is partially rooted in cycles of retaliatory violence that are driven by oppositional conduct
norms. In the US context, subcultures of violence are often interpreted to be neighborhoodrelated and linked to neighborhood disadvantage and disorganization. For example, in a
recent study, Berg et al. (2012) found that the reciprocal relation between victimization and
offending was particularly strong in neighborhoods where a street culture predominates. In
Europe, on the other hand, differences between neighborhoods tend to be less obvious and
neighborhood context tends to exert a much smaller influence on the offending rate of its
residents (Averdijk et al. 2012; Müller 2008). Instead, subcultures of violence are more
related to honor cultures expressed in violence-justifying masculinity norms, which are
closely related to ethnic and socio-economic background (Cohen 1972; Enzmann et al.
2003; Ribeaud and Eisner 2009).
Although most empirical evidence sides with theories that suggest that offending
increases the risk of victimization (Ousey et al. 2011), a negative relation between victimization and offending has also been argued. For example, ethnographic accounts (e.g.,
Anderson 1999; Katz 1988) suggest that the use of violence against others can be used to
gain respect, demonstrate toughness, and avoid subsequent harassment and hence serves as
a deterrent to victimization. Ousey et al. (2011) found evidence for the commonly found
reciprocal positive relation between offending and victimization in a longitudinal model
without controls added to it. However, when controlling for time-stable individual characteristics and dispositions, victimization turned out to be negatively related to later
offending and vice versa. As will be argued in more detail below, we think that it may
precisely be individual characteristics and dispositional factors that can account for differences between victims-offenders, non-offending victims, and non-victimized offenders.
Victims, Offenders and Victim-Offenders
Foreshadowing recent attempts to increase specificity in the victim-offender outcome
variable, von Hentig (1948) argued that not all victims are alike in the sense that certain
groups of victims are passive recipients of violence whereas others actively contribute to
their own misfortunes. Hence, in spite of the fact that victims and offenders often belong to
the same group, victims and offenders should not simply be treated alike in analytic
frameworks.
Recently, several studies have started to examine how victims and offenders differ. For
example, focusing on assault among undergraduate students, Mustaine and Tewksbury
(2000) found that several factors differentiate victims and offenders. Whereas victimization was best predicted by a high exposure to potential offenders or likely criminal events
and, to a lesser extent, by the potential victim’s alcohol use and lifestyle, offending, was
best predicted by demographic characteristics and participation in other illegal activities.
123
J Quant Criminol (2015) 31:653–675
657
Furthermore, Klevens et al. (2002) found that victims tended to avoid risky activities,
whereas victim-offenders did not.
For homicide, Broidy et al. (2006) found that victims with no prior offending history
differed from the offenders on demographic characteristics and social contexts. In contrast,
Daday et al. (2005), comparing victims and offenders of non-lethal violence, found that
both victims and offenders live in socially disorganized neighborhoods and share risky
lifestyles and violent behaviors.
Recently, Schreck et al. (2008) proposed a novel statistical approach to analyze tendencies to gravitate towards either violent offending or victimization. Based on a longitudinal study of US adolescents, they found meaningful variation in the tendency toward
either victimization or offending for age, drinking and attachment to parents. Older participants tended towards a victim role, as did those who got drunk frequently and those who
were more attached to their parents.1 Other variables, such as those reflecting risky lifestyles and emotional distress, were associated with a general exposure to violent
encounters, whether as a victim or as an offender, but not with the differential tendency
towards either victimization or offending.
In sum, recent research suggests that there may exist certain characteristics that predispose people towards offending but not victimization and vice versa. Nevertheless, as
Broidy et al. (2006) argue, while a variety of theories can help make sense of the victimoffender overlap, there is little theoretical discussion of the conditions under which victim
and offender populations diverge and such discussion would be an important step towards
understanding the vulnerabilities that presage victimization, particularly where traditional
measures of structural disadvantage, risky lifestyle and criminal involvement do not appear
to be operative. Below, we explore the possibility that socio-psychological characteristics
can account for these differences.
An Individual Differences Perspective on the Victimization-Offending Nexus
It was again von Hentig (1948) who was among the first to link individual dispositions to
people’s tendency towards victimization by proposing different ‘psychological types of
victim’, such as ‘the depressed’, ‘the wanton’, and ‘the tormentor’. He also suggested that
individual-level variables could explain differences between victims and offenders. In the
present study, we follow von Hentig’s intuition and examine the possibility that specific
socio-psychological characteristics account for differences in people’s inclination towards
offending, victimization or both, over and above risky lifestyles/routine activities or
immersion in a violent subculture. Below, we draw out an individual differences perspective grounded in the idea of violent crime as social interaction.
Violent Crime as Social Interaction
Exceptions aside, violent crime typically implies social interaction, and often also an
interpersonal relationship between the actors that precedes the interaction. As psychological characteristics of individuals influence the onset and development of their social
1
Note that in the publication by Schreck et al. (2008), there is an error as the positive sign of the coefficient
(‘drunk’) in the body text (p. 892) should instead be negative [as it is (correctly) displayed in Table 5 of their
publication] implying that being drunk is related to a tendency towards victimization instead of offending
(Schreck personal communication, September 5, 2013).
123
658
J Quant Criminol (2015) 31:653–675
interactions, and interpersonal behavior more generally, it is plausible that certain types of
characteristics will also have an impact on how violent interactions come about and
develop. If correct, this assumption implies that victims who do not double as offenders
possess certain characteristics or traits that set them apart from the latter. More specifically,
we argue that there is a constellation of different but related individual characteristics and
behaviors that seem to work together to increase people’s risk of victimization. Analogously, those offenders who are able to avoid getting victimized, in spite of their own
engagement in delinquency and hence exposure to risk factors such as those embedded in
risky lifestyles/routine activities and violent subcultures, are likely to be endowed with
different sets of individual qualities than victims and victim-offenders.
Anxiety, Depression and Negative Social Relations
One of the few individual differences variables used in prior research on the victimoffender overlap is self-control. The core idea of self-control theory is that those who lack
it tend to disregard the longer-term consequences of their behavior, which puts them at risk
for crime (Gottfredson and Hirschi 1990). Schreck (1999) reformulated the theory to also
account for victimization by arguing that low self-control produces vulnerability to crime.
For example, the disregard of long-term consequences makes it less likely that people will
take precautions against victimization (Schreck 1999). Several recent empirical studies
support the claim that low self-control is predictive not only of offending but also of
victimization (e.g., Daigle et al. 2008; Schreck 1999; Ousey et al. 2011; Piquero et al.
2005). In the present study, we examine self-control in combination with a larger set of
socio-psychological dispositions. In contrast to self-control, which on the basis of earlier
work we expect to primarily influence the overlap between victims and offenders, these
other socio-psychological characteristics are expected to discriminate in particular between
victims and offenders.
While criminologists examining the victim-offender overlap have mainly focused on
self-control, psychological research on peer victimization has also examined other variables. Importantly, some of this research (e.g., Swearer et al. 2001; Craig 1998) distinguishes between victims, perpetrators and victim-perpetrators demonstrating meaningful
differences between these groups. We think that these findings may extend to general
victimization in meaningful ways and therefore draw from this literature to develop our
individual differences perspective.
As most (violent) crime implies social interaction, it makes sense to assume that victims’ emotional states and behaviors, in particular their internalizing problems, such as
anxiety and depression, influence their risk of victimization. Specifically, youths with
internalizing problems have been shown to display a lack of social competencies and
heightened reassurance seeking, which, in turn, disturb interpersonal relationships
(Rudolph et al. 2008), and puts them at risk for victimization (see also Storch et al. 2005).
Or, as Slee (1995, p. 57) phrases it, the ‘‘tendency to be victimized may encapsulate
provocative behaviour which elicits aggression from others’’ (see also Felson 1992).
However, the proclivity to be victimized may also be associated with withdrawn behavior,
such as the avoidance of interactions and lack of assertiveness (Storch et al. 2005). From
the offenders’ perspective, the fearfulness, withdrawal and social isolation of potential
targets may trigger negative behavior towards them. Additionally, as anxious and isolated
individuals lack social support structures to help defend them and the social skills to avert
or negotiate an attack, this may lead them to be viewed as easy prey. For example, Egan
and Perry (1998), examining characteristics of third- and seventh-grade students making
123
J Quant Criminol (2015) 31:653–675
659
them vulnerable to victimization, found that behavioral characteristics such as perceived
weakness, manifest anxiety and poor social skills contributed to victimization for children
with low self-regard. The authors argued that these children may contribute to their own
victimization by failing to assert themselves during conflict, which makes them more
vulnerable targets (Egan and Perry 1998). Various longitudinal studies have found that
(young) victims of violence tend to suffer from higher degrees of anxiety and related
internalizing problems and the evidence suggests that these problems indeed precede
victimization (e.g., Fekkes et al. 2006; Hodges and Perry 1999; Kochel et al. 2012).
Although, as noted, similar research in criminology is less prevalent, some criminological
studies have suggested that victimization is related to anxiety and/or depression (BoneyMcCoy and Finkelhor 1996; Silver 2002; Silver et al. 2005).
Swearer et al. (2001) noted that anxious children often have difficulty initiating and
maintaining social and peer relations, which may be the result of, or have an impact upon,
feelings of depression. Additionally, over time these youths may come to view themselves
as deserving of peer attacks which also contributes to symptoms of depression (see also
Craig 1998; Hawker and Boulton 2000; Olweus 1995; Slee 1995). Pagani et al. (2008,
p. 42) argued in this respect that individuals experiencing socio-ecological risks such as
being a victim of bullying tend to be less socially competent in establishing supportive
relationships and avoiding peer rejection, consequently predicating depression.
Although the victim-offender may also possess negative psychological characteristics,
we expect the contrast in socio-psychological differences between groups to be starkest
between victims(-only) and offenders(-only) as opposed to the hybrid victim-offender
group. That is, the offender able to avoid being victimized is likely to possess certain traits
and skills that make him/her confident that he/she will not suffer from the violent behavior
of others. Various studies on bullying have for example found that bullies who did not also
report victimization experience less feelings of anxiety than bully-victims (e.g., Craig
1998; Olweus 1995; Swearer et al. 2001). In criminology, a link has been proposed
between negative emotions and delinquency (Agnew 1992), but most research in this field
has focused on anger and less on feelings of depression. Although some research among
inmates (Silver et al. 2008) and adolescents (Beyers and Loeber 2003; Kandel and Davies
1982) suggests an association between depression and offending, other research has shown
that this association is caused primarily by anger and not by depression. For example,
Broidy (2001) found that a general measure for negative emotions (excluding anger) was
related to less crime, while anger was associated with more crime. Moreover, when
controlling for anger, Sigfusdottir et al. (2004) did not find a significant relation between
depression and delinquency; the effect turned significant when anger was removed from
the analysis.
We therefore hypothesize offenders who are not victimized to suffer less from internalizing problems, such as anxiety and depression, while simultaneously possessing better
social skills. In addition, we therefore hypothesize these adolescents to be more liked, be
more popular and less socially isolated than victims, i.e., to have less negative peer
relations. Furthermore, we expect offenders to be more assertive and dominant compared
to victims and offender-victims, thereby being better able to navigate their way out of
potential conflict situations without getting victimized or to simply be able to dominate
others in these situations. Finally, we expect risky leisure activities and substance use, both
characteristic of risky lifestyles/routine activities, and masculinity norms and membership
of delinquent peer groups, which reflect violent subcultures, and self-control to discriminate in particular between the group of victim-offenders and their normative peers who
have neither been victimized nor have offended.
123
660
J Quant Criminol (2015) 31:653–675
Methods
Participants
The data were drawn from an ongoing combined longitudinal and intervention study, the
Zurich Project on the Social Development of Children and Youths (z-proso) (Eisner et al.
2011). The target population consisted of all 2,520 children who entered the first grade in
one of the 90 public primary schools in the city of Zurich, Switzerland, in 2004. Because
the interventions occurred at the school level, a cluster randomized sampling approach was
used, with schools as the randomization units. The schools were classified by enrollment
size and socioeconomic background of the school district. Subsequently, a stratified sample
of 56 schools was drawn. The final sample consisted of all 1,675 first graders in these
schools, as well as their parents and teachers. At the start of the study, the mean age of the
participants was 7.45 years (SD = 0.39).
The sample was 52 % male. Eleven percent of the children were born outside of Switzerland, and in 46 % of the cases both parents were born outside of Switzerland. In terms of
educational attainment of the parents, 23 % had little to no secondary education, 27 % had
vocational training only, 29 % had attended full-time vocational school or had earned a
baccalaureate degree or advanced vocational diploma and 20 % had a university degree.
Data were based on the two most recent waves (five and six) of the z-proso project,
which will henceforth be referred to as T1 and T2. Predictor variables were collected at T1
by means of both the child and teacher interviews; T2 data regarded the dependent variables and were collected through the child interview only. At T1, when the mean participant age was 13.7 years (SD = 0.37), 82 % of the youths from the original target sample
(N = 1,366) and 76 % of the teachers (N = 1,269) participated. At T2, with a mean age of
15.4 years (SD = 0.36), 86 % of the youths from the original target sample participated
(N = 1,447). The present sample included only those youths who participated in both
waves. Parents were asked to provide passive consent, meaning that they could refuse their
child’s participation by actively notifying the research team and that no parental reaction
was taken to mean that the parents consented to their child’s participation. Questionnaires
were completed in a classroom-setting after school. Participants received 30 Swiss Francs
(approximately 30 USD) for their participation at T1 and 50 Swiss Francs (approximately
50 USD) at T2.
Dependent Variables
Victimization and offending in the preceding 12 months were self-reported by the youths
at T1 and T2 (see descriptions in ‘‘Appendix 1’’). The six victimization items included
robbery, serious assault with a weapon, serious assault without a weapon but with injury,
simple assault, sexual assault and sexual harassment. The six offending items included
threat/extortion, robbery, serious assault with injury, simple assault, sexual assault and
sexual harassment (summary statistics and item parameters appear in ‘‘Appendix 2’’). One
delinquency item (sexual assault) and one victimization item (sexual assault) yielded
prevalence rates of \1 %, which proved to be problematic for the analyses. We therefore
removed these items, resulting in a total of five victimization items and five offending
items included in the analyses.
Six of the items were originally coded as count variables (extortion perpetration, robbery perpetration, serious assault perpetration, robbery victimization, assault victimization
with weapon and assault victimization without a weapon); the other four (simple assault
123
J Quant Criminol (2015) 31:653–675
661
perpetration, simple assault victimization, sexual harassment perpetration and sexual
harassment victimization) were part of a bullying questionnaire and were measured using a
frequency scale from 1 (‘never’) to 6 [‘(almost) every day’].2 Similar to prior studies
(Osgood and Schreck 2007; Schreck et al. 2008), all items were recoded into a dichotomy
of 0 (‘did not experience violence’) and 1 (‘experienced violence’). Although recent
studies have tended to use event rates of victimization or offending (e.g., McGloin et al.
2011; Schreck et al. 2012), we decided against this because the four items that were
measured on the mentioned frequency scale did not include specific crime counts and could
thus not be meaningfully transformed into a count scale. Removing these four items from
the analysis would have meant that a large part of (and the most common) violent experiences by youths would be ignored.
We assessed the analytical properties of these items in two ways. First, we assessed
whether the items represented an underlying tendency towards ‘violent encounters’ (i.e.,
the victim-offender overlap) by running a correlation analysis and a principal components
analysis (PCA) (see Tables 1, 2). The results indicated that most of the victimization and
offending items displayed a consistent pattern of positive correlations with each other. An
exception was sexual harassment victimization, which displayed some low or even negative correlations. A subsequent PCA based on tetrachoric correlations showed high
positive loadings on the first factor. Only the loadings for sexual harassment victimization
were not satisfactory. Subsequent analyses revealed that these anomalies were due to the
gendered nature of sexual harassment victimization. We therefore added gender as a
predictor for the sexual harassment dummy indicator and also controlled for gender in our
regression analyses. Second, the correlations provided an indication that the victimization
and offending items displayed distinctiveness because the victimization items were more
strongly associated with each other than with the offending items, and vice versa.
Predictors
Risky Lifestyle
3
Risky leisure activities were measured with eight items referring to unstructured out-ofhome leisure activities with friends without supervision by parents (e.g., ‘‘hang around and
have fun with friends at the train station, shopping mall, or park’’; a = .83). Answers were
given on a 6-point scale from 1 (‘never’) to 6 [‘(almost) everyday’].
Substance use was assessed with four items that measured the frequency of tobacco,
alcohol, strong liquor and marijuana consumption (a = .81). Answers were given on a
5-point scale from 1 (‘never’) to 5 (‘daily’).
Subculture of Violence
Three items measuring masculinity norms assessed the extent to which youths endorsed
violence as a necessary means to defend themselves or those around them (e.g., ‘‘A real
2
Because all victimization questions were asked in relation to violence among youths, they can be expected
to mainly tap into victimization by other youths. The same was not the case for the offending items.
However, a follow-up question to the offending item on serious assault with injury revealed that 91 % of
offenses were committed against persons between 10 and 18 years of age, suggesting that these incidents
primarily occurred between youths too.
3
When applicable, items of the predictor and outcome variables appear in ‘‘Appendix 1’’.
123
123
.42 [.52]
5. Robbery
.39 [.32]
.35 [.34]
.09 [-.01]
.41 [.32]
7. Serious assault without weapon
8. Serious assault with weapon
9. Sexual harassment
10. Robbery
T2 correlations displayed in brackets
.76 [.72]
6. Simple assault
Victimization
.50 [.46]
.57 [.62]
4. Extortion
.58 [.63]
3. Sexual harassment
2. Serious assault
1. Simple assault
Offending
1
.26 [.28]
.20 [.05]
.38 [.34]
.37 [.42]
.44 [.40]
.54 [.42]
.62 [.57]
.42 [.27]
2
.12 [.18]
.47 [.44]
.37 [.18]
.32 [.28]
.30 [.25]
.45 [.37]
.58 [.42]
3
.21 [.30]
.17 [-.07]
.41 [.18]
.45 [.18]
.46 [.12]
.79 [.85]
4
Table 1 Tetrachoric correlations between offending and victimization items for T1 and T2
.36 [.31]
.04 [.04]
.45 [.10]
.31 [.25]
.28 [.27]
5
.41 [.39]
.22 [.15]
.38 [.61]
.58 [.56]
6
.43 [.43]
.19 [.21]
.55 [.62]
7
.40 [.25]
.26 [.25]
8
.07 [-.08]
9
10
662
J Quant Criminol (2015) 31:653–675
J Quant Criminol (2015) 31:653–675
Table 2 Factor loading onefactor solution principal
components analysis
663
T1
T2
1. Simple assault
0.82
0.85
2. Serious assault
0.68
0.74
3. Sexual harassment
0.66
0.51
4. Extortion
0.83
0.72
5. Robbery
0.73
0.69
6. Simple assault
0.73
0.73
7. Serious assault without weapon
0.66
0.61
8. Serious assault with weapon
0.61
0.59
Offending
Victimization
9. Sexual harassment
0.31
0.19
10. Robbery
0.49
0.54
4.45
4.09
Eigenvalue
man must defend himself’’; a = .69; derived from Nisbett and Cohen 1996). Answers were
given on a 4-point scale from 1 (‘entirely incorrect) to 4 (‘entirely correct).
Gang membership was coded ‘‘1’’ if the respondent was part of a group of friends that
was involved in at least one of nine delinquent activities (threatening, assaulting, or
fighting with other people; theft or burglary; robbery; extortion; drug dealing; carrying
weapons; vandalism; substance use; other illegal activities) and ‘‘0’’ otherwise.
Socio-psychological Characteristics
Anxiety and depression were measured through the Social Behavior Questionnaire (SBQ;
Tremblay et al. 1991). The scale included eight items ranging from 1 (‘never’) to 5 (‘very
often’) (e.g., ‘‘I was sad without knowing why’’; a = .83).
Our measure for self-control included 10 items measured on a 4-point Likert scale from
1 (‘does not apply at all’) to 4 (‘very much applies’) (e.g., I act spontaneously, without
thinking too much; a = .83), adapted from Grasmick et al. (1993) (see Ribeaud and Eisner
2006).
Dominance towards others was measured through a one-item measure filled out by the
teachers. Answers were recorded on a 5-point scale from 1 (‘does not apply at all’) to 5
(‘very much applies’).
We included two additional variables that measured negative peer relations. Isolation
and popularity were each rated by the teachers on a 5-point scale from 1 (‘does not apply at
all’) to 5 (‘very much applies’). After reverse-coding the popularity item, a composite scale
consisting of both items was computed (a = .73).
Control Variables
We controlled for gender (‘‘0’’ is female, ‘‘1’’ is male), ethnicity (with ‘‘0’’ signifying at
least one Swiss parent, and ‘‘1’’ two non-Swiss parents), and socio-economic status (SES).
The latter was based on coding the caregiver’s current profession (Elias and Birch 1994).
This code was subsequently transformed into an International Socio-Economic Index of
123
664
J Quant Criminol (2015) 31:653–675
occupational status (ISEI) score (Ganzeboom et al. 1992). The final SES score was based
on the highest ISEI score of the two caregivers.
Analytic Strategy
We used the statistical approach proposed by Osgood and Schreck (2007) to examine
specialization in offending, and which was later applied to the victim-offender overlap by
Schreck et al. (2008). Because this method has been extensively described in these two
publications and also in several subsequent studies (e.g., McGloin et al. 2011; Schreck
et al. 2012), we restrict ourselves to a summary here.
The approach is grounded in item response theory (IRT), which provides a framework
for modeling the relations between individual test items and the latent constructs the items
are intended to measure (e.g., victimization, offending). Different from the usual practice
of summing items to represent a construct, IRT estimates an individual’s most likely
position on the latent trait given his responses on the test items. The latent trait captures the
construct on a continuous scale with equal intervals that is free from measurement error
(Osgood et al. 2002). This is particularly beneficial for the study of self-report data on
offending and victimization, because summative measures of these phenomena are typically skewed and overemphasize the less serious and less important forms of crime
(Osgood et al. 2002). The method is further based on Raudenbush et al. (2003), who
developed a multivariate, multilevel IRT framework.
The model we used in the present study consisted of two levels, the first being the
response of the respondent to each of the ten victimization and offending items (i.e., the
IRT measurement model) and the second being the respondent.4 The probability that an
individual endorsed a particular item was modeled as a function of three factors. The first
was a latent overall propensity for a combined tendency to be involved in both offending
and victimization, overall termed ‘‘violent encounters’’ (b0j), which was based on the
respondent’s item responses and which varied randomly across individuals. The second
factor, and the one central to our research question, was the latent variable for role differentiation towards offending versus victimization (Diff). Diff returned a positive value for
offending items and a negative value for victimization items. A predominantly offender
role therefore yielded a positive value on b1j whereas a predominantly victim role yielded a
negative value. This variable varied randomly across individuals. Differentiation was
completely separated from and thus not confounded with the overall propensity for violent
encounters by group mean centering the item scores within individuals. The third factor
regarded the severity of the crime-type (i.e., item difficulty) and thus reflected the base rate
of an offense (b1j). Rarer offenses will have lower values than more common offenses.
Thus, a series of dummy variables with a ‘‘1’’ for the relevant item and a ‘‘0’’ otherwise
was included and indicated which item reflected which answer (with one of the dummies
excluded as the reference category). Models were estimated using HLM7 (Raudenbush
et al. 2011).
4
The level 1 model is defined as (see Osgood and Schreck 2007):
I
P
Log½odds Yij ¼ 1 ¼ b0j þ b1j Diff þ bij Dij ð1Þ
i¼2
The level 2 model is defined as:
b0j ¼ c00 þ c01 X1j þ c02 X2j þ þ u0j
b1j ¼ c11 X1j þ c12 X2j þ . . . þ u1j ð3Þ
bij ¼ ci0 ð4Þ.
123
ð2Þ
J Quant Criminol (2015) 31:653–675
665
Results
We first examined the precision of our measures for violent encounters and role differentiation by means of their reliability. The results are displayed in Table 3. For overall
violent encounters, the reliability was .73 at T1 and .71 at T2, which can be regarded as
moderate. The reliability for role differentiation was much lower at .43 and .45, which is in
line with prior research (Schreck et al. 2008) and to be expected given that most
respondents reported none or few violent encounters, thus providing limited information
about the contrast between offending and victimization. Our latent variable approach
accounts for this limited reliability (see Osgood and Schrec 2007).
Next, we assessed whether the tendency towards offending versus victimization was
greater than could be expected by chance. To this end, we estimated the full variance of the
latent variables for role differentiation in both waves (i.e., using models where explanatory
level 2 variables were omitted). The approximate significance of the variance was
examined using z tests, which were obtained by dividing the variance estimates by their
standard errors (see Table 3). This yielded values of 11.0 (=3.30/.30) at T1 and 11.5
(=4.59/.40) at T2. Given that the critical value for significant role differentiation at
a = .001 equals 3.3, this means that there was a highly significant tendency towards role
differentiation in both waves. The variance for role differentiation was comparable to, or
even larger than, the variance for overall violent encounters, suggesting that role differentiation contributed considerably to the respondents’ violence profiles. Consistent with
the findings reported by Schreck et al. (2008), the variance for role differentiation
increased somewhat over age.
Table 4 provides a more intuitive description of the magnitude of the tendency towards
role differentiation. It displays the observed distribution of offending and victimization for
three groups of respondents: those who tended towards offending (defined by scores at
least 1 standard deviation above the mean on role differentiation), those who tended
towards victimization (scores at least 1 standard deviation below the mean on role differentiation) and those with a mixed profile (scores within 1 standard deviation from the
mean). Analogous to Schreck et al. (2008), we only included those youths who were the
most useful for observing role differentiation towards a predominant offender or victim
role; i.e., youths who endorsed between three and seven of the ten violent encounters
(18 % of the T1 sample and 14 % of the T2 sample). The results show that those who
tended towards offending had committed about half of the offenses described by the items
across both waves, while they had experienced about one-fifth of the different types of
victimization. Those with a tendency towards victimization had committed one-sixth of the
offense items, while they had experienced two-thirds of the victimization items. Both
patterns differed from respondents with a mixed profile, who had committed about onethird of the different types of offenses and had experienced a little over 40 % of the
Table 3 Reliability and variance of overall violent encounters
and role differentiation
Violent encounters
Role differentiation
T1
T1
T2
T2
Reliability
.73
.71
.43
.45
Variance (s)
3.73 (.22)
4.11 (.24)
3.30 (.30)
4.59 (.40)
No. of respondents
1,046
1,046
1,046
1,046
123
666
J Quant Criminol (2015) 31:653–675
Table 4 Observed distribution of offending and victimization, by role differentiation
Role
Observed distribution of violent encounters
Offender
Victim
Total
n
T1
Offender ([?1 SD)
.48
.21
.35
76
Mixed ([-1 SD and \?1 SD)
.31
.45
.38
117
Victim (\-1 SD)
.15
.67
.41
54
T2
Offender ([?1 SD)
.53
.17
.35
59
Mixed ([-1 SD and \?1 SD)
.30
.42
.36
92
Victim (\-1 SD)
.15
.66
.40
50
SD standard deviation
victimization types. These findings suggest that, in line with our expectations, there are
sizable differences between those with a tendency towards victimization and those with a
tendency towards offending.
In a subsequent step, we examined the extent to which role differentiation was stable
over the two waves. To this end, we included violent encounters and role differentiation at
both waves in our model and estimated the correlations among these latent variables
(Table 5). The stability in role differentiation was substantial across both waves (r = .72)
and larger than the stability in violent encounters (r = .57).
In the last step of our analysis, we predicted violent encounters and role differentiation
by our variables for routine activities/lifestyle, violent subcultures and socio-psychological
dispositions. The results are shown in Table 6. Recall that our predictors were measured at
T1 and our dependent variables at T2 to ensure correct measurement of the theorized
temporal ordering. As prior research has focused on routine activities/risky lifestyles and
violent subcultures as the primary correlates, our first model was restricted to these two
perspectives. Analogous to previous findings, our results showed that risky leisure activities, substance use and violent subcultures in the form of masculinity norms endorsement
and delinquent peer-group membership were significantly related to violent encounters. In
contrast, the only variable to predict role differentiation in this model was masculinity
norms with higher scores on this variable leading to offending role taking. Furthermore,
being male predicted violent encounters and lower SES predicted a tendency towards
Table 5 Correlations among
violent encounters and role differentiation across two timepoints
Violent encounters
Role differentiation
T1
T1
T2
Violent encounters
T1
T2
.57
Role differentiation
123
T1
.07
.15
T2
.18
.21
.72
T2
J Quant Criminol (2015) 31:653–675
667
Table 6 Regression results of violent encounters and role differentiation on explanatory variables
T1 predictors
Model 1
Model 2
T2 violent
encounters
T2 role
differentiation
T2 violent
encounters
T2 role
differentiation
c
SE
c
c
SE
c
SE
SE
Routine activities and subculture
Risky leisure activities
.10**
.03
-.06
.05
.05
.03
-.09
.05
Substance use
.23**
.05
.10
.06
.17**
.04
.11
.07
Masculinity norms
.19**
.04
.31**
.06
.11*
.04
.29**
.07
Member of delinquent peer group
.20*
.09
.19
.13
.17
.09
.16
.13
.06
Socio-psychological characteristics
Anxiety and depression
.23**
.04
-.24**
Low self-control
.40**
.07
.09
.10
Negative peer relations
.04
.03
.03
.05
Dominant
.03
.03
.16**
.06
Control variables
Male
.56**
.06
.15
.08
.66**
.06
.03
.08
Non-Swiss
-.07
.06
-.07
.08
-.04
.06
-.08
.08
Socio-economic status
.00
.00
-.01*
.00
.00
.00
-.01*
.00
No. of respondents
1,046
1,046
c is the HLM population average estimate and SE its robust standard error
SE standard error
p \ .10; * p \ .05; ** p \ .01 (two-tailed tests)
victimization. In other words, the results for Model 1 roughly replicate those of prior
research.
In our second and final model, we went beyond the usual routine activities/lifestyle and
violent subcultures explanations and also assessed the predictive value of socio-psychological dispositions. Suffering from anxiety and depression and having low self-control
were associated with having experienced more violent encounters and thus were predictive
of the victim-offender overlap. Interestingly, and in line with our hypothesis, suffering
from anxiety and depression was also associated with a tendency towards victimization. In
addition, being dominant towards others was associated with a tendency towards offending. In contrast to our hypothesis, positive peer relations were not associated with role
differentiation. Importantly, including the socio-psychological characteristics led to lower
gamma’s of the routine activities and subculture variables, and for the effects of risky
leisure activities and delinquent peer group-membership on violent encounters to nonsignificance. This finding suggests that the effects of routine activities and violent subcultures may operate on offending and victimization via socio-psychological characteristics. In other words, as we hypothesized, the inclusion of socio-psychological variables
beyond the commonly used routine activities and subculture of violence variables extends
our knowledge of the factors underlying youths’ tendency to enter into violent encounters
and which role(s) they assume in them.
123
668
J Quant Criminol (2015) 31:653–675
Discussion
We argued that disentangling victim and offender roles is important for advancing our
understanding of the etiology of the victim-offender overlap. Previous efforts have tended
to emphasize the ways in which victims and offenders are alike, somewhat to the neglect of
what sets them apart. Furthermore, most research has tried to explain the association
between victimization and offending through demographic characteristics, routine activities/risky lifestyles and subcultures of violence, and much less through individual qualities
that may determine the onset, development and outcome of violent interactions. However,
as Lauritsen and Laub (2007:62) argued, in order to identify the mechanisms underlying
the relation between victimization and offending, it is imperative to go a step further and
examine types of heterogeneity that are not captured by demographic and neighborhood
characteristics, family and peer factors, lifestyle activities and subcultural norms. In this
article, we did so by examining a series of socio-psychological characteristics in conjunction with the routine activity/lifestyle perspective and subcultural notions to explain
differences between those who take on predominantly offender roles versus those who are
predominantly victims.
We hypothesized that the socio-psychological factors we examined in particular would
account for a differential tendency towards offending or victimization and as such extend
routine activities/risky lifestyle and subcultural explanations. The results show that this
was the case as the socio-psychological variables predicted the tendency to take on victim
roles versus offender roles over and above routine activities/risky lifestyles and subculture
of violence perspectives. Moreover, while not anticipated and exceeding our expectations,
the results also indicated that certain psychological variables, i.e., anxiety and depression,
also discriminate between the victim-offender group and those who did not offend and
were also not victimized, implying that here too psychological variables can contribute to
our understanding of the etiology of the victim-offender overlap in important ways. Furthermore, and in line with expectations, low self-control also predicted the overlap versus
the non-involved. These findings warrant the observation that there are important differences between individuals who have been victimized without having offended, those that
have offended without having been victimized, and those that have engaged in both or in
neither.
Using psychological constructs to account for victimization could be interpreted as
blaming the victim, but such interpretation would be erroneous. Arguing along similar
lines as Schreck (1999), we interpret our results as indicating that certain psychological
characteristics increase people’s risk to be targeted for crime and hence render them
vulnerable and make them preferential targets. It would therefore be more productive to
use these findings for designing policy and interventions that aid these individuals to
reduce their likelihood of victimization. Schreck et al. (2008, p. 894) noted that programs
for violence reduction and prevention are often based on the idea that victims and
offenders belong to distinct populations and address one group or the other but not both,
and thereby ignore the fact that a large share of those who frequently encounter violence
take on roles as victims as well as offenders. The findings of the present study underscore that differentiation between those who encounter violence as both offender and as
victim, those that primarily fall victim to crime, and those that primarily offend is
important. As our results indicate, the latter group differs from the former on a range of
psychological variables, which hints at the fact that interventions tailored specifically to
each group are likely to be more successful than interventions that are indiscriminate in
this respect.
123
J Quant Criminol (2015) 31:653–675
669
When interpreting the results a number of considerations and limitations should be
borne in mind. Our variable for peer relations, which was not significant, was measured
through teacher evaluations. These have been shown to have different views on youths’
social relations than the youths themselves (Averdijk et al. 2013). Although teachers’
views are valid in and of themselves, future research should investigate whether the views
of different informants yield different results for this measure. Furthermore, we note that
our operationalization of socio-psychological characteristics should not be interpreted as
being encompassing. Future research should explore other potentially relevant characteristics, such as personality dimensions which have already been shown to be consistent
predictors of delinquency, such as agreeableness, conscientiousness and honesty-humility
(Miller and Lynam 2001; Van Gelder and De Vries 2012), as well as trait level emotions
incorporated in general strain theory such as anger and frustration (see Agnew 1992).
Individual differences in the experience of strain-related emotions such as anger and
frustration may for example explain why some people resort to violent of offending after
having been victimized while others do not (see also Agnew 2002). It should also be noted
that our sample consisted of youths and that most of the offending incidents occurred
between youths. Furthermore, most youths in our sample reported few violent encounters,
which limits the information that can be derived from the contrast between offending and
victimization. Our findings, therefore, do not necessarily apply to older age groups and/or
more delinquent groups. Hence, future research should replicate our results among older
offenders to examine to what extent they can be generalized to these populations.
Finally, we acknowledge that the use of a self-reported delinquency measure carries as a
limitation that it is prone to bias, such as recollection errors and over- and underreporting
(e.g., Huizinga and Elliott 1986). Nonetheless, we think that for the purposes of the present
study self-report measures carry various advantages over alternative methods that outweigh
the shortcomings. Specifically, self-report methods allow for the detection of forms of
crime that are not picked up by official statistics because they are not reported (Krohn et al.
2010).
Conclusion
It has become commonplace for researchers studying the victim-offender overlap to
lament the fact that most studies have treated victims and offenders as separate categories in spite of the strong empirical association between victimization and offending.
However, it was over 60 years ago that von Hentig not only pointed out the association
between victimization and offending but also went a step further by noting that certain
victims who also engage in offending may differ from victims who do not in important
ways. It was again von Hentig (1948) who hinted that individual characteristics, such as
anxiety or depression, could influence people’s risk of victimization. Ahead of his time,
he argued that examining individual traits and psychological characteristics could
advance our understanding of the victim-offender overlap. In another early publication,
Wolfgang (1958, p. 4) added: ‘‘As personality and social environment are inseparable, so
must the bio-psychological and sociological approaches to homicide and other problems
also be interdependent’’.
Schreck et al. (2008, p. 873) argued that the more people adopt one role versus the
other, the greater the need for specific theorizing and research to account for both phenomena: ‘‘As evidence of similarities between victims and offenders has mounted, it may
be time to step back and evaluate the extent to which offenders and victims differ as well.
123
670
J Quant Criminol (2015) 31:653–675
The greater these differences, the more justification for specialized theories of criminality
and victimization.’’ We believe that the results of the present study emphasize the need for
this type of specialization and can guide future efforts in the direction of individual-level
variables to complement and extend research frameworks using routine activities/risky
lifestyle and subcultural theory variables.
Appendix 1: Specification of Variables
Victimization
Serious victimization questionnaire In the past 12 months, so since July 2010, has one of
the following things happened to you? And if yes, how many times since July 2010?
• Someone took something from you while using violence or threatening with violence,
for example your purse, bike or money [Robbery].
• Someone deliberately injured you with a weapon (e.g., a knife) or with an object (e.g., a
stick) or through kicking you with heavy shoes [Serious assault with weapon].
• Someone hit you so seriously, that you got injured (e.g., a bleeding wound or a black
eye). However, no weapon or object was used [Serious assault without weapon].
Bullying questionnaire How many times since July 2010 have other youths:
• hit, bit or kicked you or pulled your hair? [Simple assault]
• sexually harassed you (e.g., hit on you, groped you)? [Sexual harassment]
Offending
Offending questionnaire Since July 2010, have you ever:
• threatened anyone with violence to obtain money or goods? [Extortion]
• taken money or things from anyone while using violence? [Robbery]
• hit, kicked or cut anyone deliberately while injuring him/her? [Serious assault] (followup question: If yes, how many times since July 2010?)
Bullying questionnaire How many times since July 2010 have you:
• hit, bit or kicked another youth, or pulled his/her hair? [Simple assault]
• sexually harassed another youth (e.g., angemacht, begrapscht)? [Sexual harassment]
Risky Lifestyle
How often do you do the following things?
• Meet friends at night and do something together.
• Meet friends at someone’s home without the presence of adults.
• Hang around at the youth center with friends, without taking part in organized
activities.
• Go to a party in the evening with friends.
• Meet with friends at a café or a restaurant (e.g., Starbucks, McDonalds).
• Go out with friends to a bar or a club at night.
123
J Quant Criminol (2015) 31:653–675
671
• Hang around in a park, at the train station or in a shopping mall and have fun with
friends in the afternoon.
• Hang around in a park, at the train station or in a shopping mall and have fun with
friends in the evening.
Masculinity Norms
• A man is allowed to use violence when he is insulted.
• A real man is ready to use violence when someone says bad things about his family.
• A real man is strong and protects his family.
Anxiety and Depression
Please indicate how you felt in the past month.
•
•
•
•
•
•
•
•
I
I
I
I
I
I
I
I
cried.
was fearful for no particular reason.
was unhappy.
felt lonely.
could not fall asleep at night.
was sad without knowing why.
was bored.
was worried.
Self-Control
•
•
•
•
•
•
•
•
•
I act spontaneously, without thinking too much.
I try to get what I want, even if this causes problems for others.
I enjoy doing dangerous things, just because it is fun.
If I don’t get what I want fast, I get angry.
I enjoy going out and doing something rather than reading and thinking.
I don’t care if others are upset about something that I did.
I lose control pretty easily.
If I can, I like to do something with my hand rather than with my head.
I always do whichever I like doing in that moment, without considering the
consequences.
• Excitement and adventure are more important to me than security.
Appendix 2
See Table 7.
123
672
J Quant Criminol (2015) 31:653–675
Table 7 Offending and victimization: summary statistics and item parameters
T1
% Yes
T2
ci0
SE
% Yes
ci0
SE
Offending
Simple assault (reference)
29
[-.82]
.06
26
[-1.07]
.06
Serious assault
10
-1.12
.07
10
-.97
.06
Sexual harassment
7
-1.34
.07
6
-1.32
.07
Extortion
1
-2.56
.07
2
-2.00
.06
Robbery
2
-2.28
.07
2
-2.02
.06
Victimization
Simple assault
26
-0.18
.06
20
-0.24
.06
Serious assault without weapon
10
-1.18
.08
7
-1.18
.07
Serious assault with weapon
7
-1.51
.08
5
-1.48
.07
Sexual harassment
19
0.14
.09
21
.57
.10
Sexual harassment by gender
n.a.
-1.61
.12
n.a.
-1.96
.13
Robbery
8
-1.36
.08
4
-1.51
.07
No. of respondents
1,046
1,046
c is the HLM population average estimate and SE its robust standard error
SE standard error
References
Agnew R (1992) Foundation for a general strain theory of crime and delinquency. Criminology 30:47–88
Agnew R (2002) Experienced, vicarious, and anticipated strain: an exploratory study on physical victimization and delinquency. Justice Q 19:603–632
Anderson E (1999) Code of the street: decency, violence, and the moral life of the Inner City. W. W. Norton,
New York
Anderson AL, Hughes LA (2009) Exposure to situations conducive to delinquent behaviour. J Res Crime
Delinq 46:5–24
Averdijk M, Elffers H, Ruiter S (2012) Disentangling context effects on criminal careers during adolescence
and into early adulthood. In: Loeber R, Hoeve M, Slot NW, van der Laan P (eds) Persisters and
desisters in crime from adolescence into adulthood: explanation, prevention and punishment. Ashgate,
Aldershot, pp 51–76
Averdijk M, Eisner M, Ribeaud D (2013) Method effects in survey questions about peer victimization. In:
Ruiter S, Bernasco W, Huisman W, Bruinsma G (eds) Eenvoud en verscheidenheid: liber amicorum
voor Henk Elffers [Simplicity and Diversity: Festschrift for Henk Elffers]. NSCR and Afdeling
Criminologie Vrije Universiteit Amsterdam, Amsterdam, pp 425–439
Berg MT, Loeber R (2011) Examining the neighbourhood context of the violent offending-victimization
relationship: a prospective investigation. J Quant Criminol 27:427–451
Berg MT, Stewart EA, Schreck CJ, Simons RL (2012) The victim-offender overlap in context: examining
the role of neighbourhood street culture. Criminology 50:359–390
Bernasco W, Ruiter S, Bruinsma GJN, Pauwels LJR, Weerman FM (2013) Situational causes of offending: a
fixed-effects analysis of space-time budget data. Criminology 51:895–926
Bernburg JG, Thorlindsson T (2001) Routine activities in social context: a closer look at the role of
opportunity in deviant behaviour. Justice Q 18:543–567
Beyers JM, Loeber R (2003) Untangling developmental relations between depressed mood and delinquency
in male adolescents. J Abnorm Child Psychol 31:247–266
Boney-McCoy Sue, Finkelhor David (1996) Is youth victimization related to trauma symptoms and
depression after controlling for prior symptoms and family relationships? A longitudinal, prospective
study. J Consult Clin Psychol 64:1406–1416
123
J Quant Criminol (2015) 31:653–675
673
Broidy LM (2001) A test of general strain theory. Criminology 39:9–36
Broidy LM, Daday JK, Crandall CS, Sklar DP, Jost PF (2006) Exploring demographic, structural, and
behavioral overlap among homicide offenders and victims. Homicide Stud 10:155–180
Cohen AK (1955) Delinquent boys: the culture of the gang. The Free Press, Glencoe
Cohen P (1972) Working-class youth cultures in East London. Working papers in cultural studies, Birmingham, 2
Cohen LE, Felson M (1979) Social change and crime rate trends: a routine activities approach. Am Sociol
Rev 44:588–608
Craig WM (1998) The relationship among bullying, victimization, depression, anxiety, and aggression in
elementary school children. J Pers Individ Dif 24:123–130
Daday JK, Broidy LM, Crandall CS, Sklar DP (2005) Individual, neighborhood, and situational factors
associated with violent victimization and offending. Crim Justice Stud 18:215–235
Daigle LE, Beaver KM, Hartman JL (2008) A life-course approach to the study of victimization and
offending behaviors. Vict Offenders 3:365–390
Egan SK, Perry DG (1998) Does low self-regard invite victimization? Dev Psychol 34:299–309
Eisner MP, Malti T, Ribeaud D (2011) Large-scale criminological field experiments. In: Gadd S, Karstedt S,
Messner S (eds) Sage handbook of criminological research methods. Sage, London, pp 410–424
Elias Peter, Birch Margaret (1994) Establishment of community-wide occupational statistics: ISCO 88
(COM)—a guide for users. Institute of Employment Research, University of Warwick, Coventry
Elliott DS, Huizinga D, Ageton SS (1985) Explaining delinquency and drug use. Sage, Beverly Hills
Elliott DS, Huizinga D, Menard S (1989) Multiple problem youth: delinquency, substance use, and mental
health problems. Springer, New York
Enzmann D, Brettfeld K, Wetzels P (2003) Männlichkeitsnormen und die Kultur der Ehre. Kölner Z Soz
Sozialpsychol 43:264–287
Fekkes M, Pijpers FIM, Miranda FA, Vogels T, Verloove-Vanhorick SP (2006) Do bullied children get ill,
or do ill children get bullied? A prospective cohort study on the relationship between bullying and
health-related symptoms. Pediatrics 117:1568–1574
Felson RB (1992) ‘‘Kick ‘em when they’re down’’. Explanations of the relationship between stress and
interpersonal aggression and violence. Soc Q 33:1–16
Felson RB, Burchfield KB (2004) Alcohol and the risk of physical and sexual assault victimization.
Criminology 42:837–859
Ganzeboom HBG, Degraaf PM, Treiman DJ, De Leeuw J (1992) A standard international socio-economic
index of occupational status. Soc Sci Res 21:1–56
Gottfredson M, Hirschi T (1990) A general theory of crime. Stanford University Press, Palo Alto
Gover AR (2004) Risky lifestyles and dating violence: a theoretical test of violent victimization. J Crim
Justice 32:171–180
Grasmick HG, Tittle CR, Bursik RJ Jr, Arneklev BJ (1993) Testing the core empirical implications of
Gottfredson and Hirschi’s general theory of crime. J Res Crime Delinq 30:5–29
Hawker DSJ, Boulton MJ (2000) Twenty years research on peer victimization and psychosocial maladjustment: a meta-analytic review of cross-sectional studies. J Child Psychol Psychiatry 41:441–455
Hay C, Evans MM (2006) Violent victimization and involvement in delinquency: examining predictions
from general strain theory. J Crim Justice 34:261–274
Hay C, Forrest W (2008) Self-control theory and the concept of opportunity: the case for a more systematic
union. Criminology 46:1039–1072
Hindelang MJ, Gottfredson MR, Garofalo J (1978) Victims of personal crime: an empirical foundation for a
theory of personal victimization. Ballinger, Cambridge
Hodges EVE, Perry DG (1999) Personal and interpersonal antecedents and consequences of victimization by
peers. J Pers Soc Psychol 76:677–685
Huizinga D, Elliott DS (1986) Reassessing the reliability and validity of self-report delinquency measures.
J Quant Criminol 2:293–327
Jacobs BA, Wright R (2006) Street justice: retaliation in the criminal underworld. Cambridge University
Press, Cambridge
Jennings WG, Piquero AR, Reingle JM (2012) On the overlap between victimization and offending: a
review of the literature. Aggress Violent Behav 17:16–26
Jensen GF, Brownfield D (1986) Gender, lifestyles, and victimization: beyond routine activity. Violence
Vict 1:85–99
Kandel DB, Davies M (1982) Epidemiology of depressive mood in adolescents. An empirical study. Arch
Gen Psychiatry 39:1205–1212
Katz Jack (1988) Seductions of crime. Moral and sensual attractions of doing evil. Basic Books, New York
123
674
J Quant Criminol (2015) 31:653–675
Klevens J, Duque LF, Ramı́rez C (2002) The victim-perpetrator overlap and routine activities: results from a
cross-sectional study in Bogotá, Colombia. J Interpers Violence 17:206–216
Kochel KP, Ladd GW, Rudolph KD (2012) Longitudinal associations among youth depressive symptoms,
peer victimization, and low peer acceptance: an interpersonal process perspective. Child Dev
83:637–650
Krohn MD, Thornberry TP, Gibson CL, Baldwin JM (2010) The development and impact of self-report
measures of crime and delinquency. J Quant Criminol 26:509–525
Lauritsen JL, Laub JH (2007) Understanding the link between victimization and offending: New reflections
on an old idea. In: Hough M, Maxfield M (eds) Surveying crime in the 21st century, vol XXII.
Criminal Justice Press, New York
Lauritsen JL, Sampson RJ, Laub JH (1991) The link between offending and victimization among adolescents. Criminology 29:265–291
Lauritsen JL, Laub JH, Sampson RJ (1992) Conventional and delinquent activities: implications for the
prevention of violent victimization among adolescents. Violence Vict 7:91–108
Maimon D, Browning CR (2010) Unstructured socializing, collective efficacy, and violent behavior among
urban youth. Criminology 48:443–474
Malik S, Sorenson SB, Aneshensel CS (1997) Community and dating violence among adolescents: perpetration and victimization. J Adolesc Health 21:291–302
McGloin JM, Schreck CJ, Stewart EA, Ousey GA (2011) Predicting the violent offender: the discriminant
validity of the subculture of violence. Criminology 49:767–794
Miller Joshua D, Lynam Donald (2001) Structural models of personality and their relation to antisocial
behavior: a meta-analytic review. Criminology 39:765–798
Müller B (2008) Soziale Kohäsion in der Nachbarschaft. Eine empirische Analyse von Einflussfaktoren auf
Quartier- und Individualebene [Social cohesion in the neighborhood. An empirical analysis of predictors on the neighborhood and individual level]. Master‘s Thesis, University of Zurich, Zurich
Mustaine EE, Tewksbury R (2000) Comparing the lifestyles of victims, offenders, and victim-offenders: a
routine activity assessment of similarities and differences for criminal incident participants. Sociol
Focus 33:339–362
Nisbett RE, Cohen D (1996) Culture of honor: the psychology of violence in the South. New directions in
social psychology. Westview Press, Boulder
Olweus D (1995) Bullying or peer abuse at school: facts and intervention. Curr Dir Psychol Sci 4:196–200
Osgood DW, Schreck CJ (2007) A new method for studying the extent, stability, and predictors of individual
specialization in violence. Criminology 45:273–311
Osgood DW, Wilson JK, O’Malley PM, Bachman JG, Johnston LD (1996) Routine activities and individual
deviant behavior. Am Sociol Rev 61:635–655
Osgood DW, McMorris BJ, Potenza MT (2002) Analyzing multiple-item measures of crime and deviance I:
item response theory scaling. J Quant Criminol 18:267–296
Ousey GC, Wilcox P, Fisher BS (2011) Something old, something new: revisiting competing hypothesis of
the victimization-offending relationship among adolescents. J Quant Criminol 27:53–84
Pagani LS, Japel C, Vaillancourt T, Côté S, Tremblay RE (2008) Links between life course trajectories of
family dysfunction and anxiety during middle childhood. J Abnorm Child Psychol 36:41–53
Piquero A, MacDonald J, Dobrin A, Daigle LE, Cullen FT (2005) Self-control, violent offending, and
homicide victimization: assessing the general theory of crime. J Quant Criminol 21:55–71
Raudenbush SW, Bryk AS, Cheong YF, Congdon RT Jr, du Toit M (2011) HLM7. Hierarchical linear and
nonlinear modeling. Scientific Software International, Lincolnwood
Raudenbush SW, Johnson C, Sampson RJ (2003) A multivariate, multilevel Rasch model with application to
self-reported criminal behavior. Sociol Methodol 33:169–211
Reiss AJ, Roth JA (1993) Understanding and preventing violence. National Academy Press, Washington
Ribeaud D, Eisner M (2006) The ‘drug–crime link’ from a self-control perspective: an empirical test in a
Swiss youth sample. Eur J Criminol 3:33–67
Ribeaud D, Eisner M (2009) Entwicklung von Gewalterfahrungen Jugendlicher im Kanton Zürich. Sauerländer, Oberentfelden
Rudolph KD, Flynn M, Abaied JL (2008) A developmental perspective on interpersonal theories of youth
depression. In: Abela JRZ, Hankin BL (eds) Handbook of depression in children and adolescents.
Guildford, New York, pp 79–102
Schreck CJ (1999) Criminal victimization and low self-control: an extension and test of a general theory of
crime. Justice Q 16:633–654
Schreck CJ, Stewart EA, Osgood DW (2008) A reappraisal of the overlap of violent offenders and victims.
Criminology 46:871–906
123
J Quant Criminol (2015) 31:653–675
675
Schreck CJ, Ousey GC, Fisher BS, Wilcox P (2012) Examining what makes violent crime victims unique:
extending statistical methods for studying specialization to the analysis of crime victims. J Quant
Criminol 28:651–671
Sigfusdottir ID, Farkas G, Silver E (2004) The role of depressed mood and anger in the relationship between
family conflict and delinquent behavior. J Youth Adolesc 33:509–522
Silver E (2002) Mental disorder and violent victimization: the mediating role of involvement in conflicted
social relationships. Criminology 40:191–212
Silver E, Arseneault L, Langley J, Caspi A, Moffitt TE (2005) Mental disorder and violent victimization in a
total birth cohort. Am J Public Health 95:2015–2021
Silver E, Felson RB, Vaneseltine M (2008) The relationship between mental health problems and violence
among criminal offenders. Crim Justice Behav 35:405–426
Singer S (1981) Homogeneous victim-offender populations: a review and some research implications.
J Crim Law Criminol 72:779–788
Singer S (1986) Victims of serious violence and their criminal behavior: subcultural theory and beyond.
Violence Vict 1:61–70
Slee PT (1995) Peer victimization and its relationship to depression among Australian primary school
students. Personality Individ Differ 18:57–62
Storch EA, Masia-Warner C, Crisp H, Klein R (2005) Peer victimization and social anxiety in adolescence:
a prospective study. Aggressive Behave 31:437–452
Swearer SM, Song SY, Cary PT, Eagle JW, Mickelson WT (2001) Psychosocial correlates in bullying and
victimization: the relationship between depression, anxiety, and bully/victim status. J Emot Abuse
2:95–121
Tremblay RE, Loeber R, Gagnon C, Charlebois P, Larivée S, LeBlanc M (1991) Disruptive boys with stable
and unstable high fighting behavior patterns during junior elementary school. J Abnorm Child Psychol
19:285–300
Van Gelder JL, De Vries RE (2012) Traits and states: integrating personality and affect into a model of
criminal decision making. Criminology 50:637–671
Vogel RE, Himelein MJ (1995) Dating and sexual victimization: an analysis of risk factors among precollege women. J Crim Justice 23:153–162
von Hentig Hans (1948) The criminal and his victim. Yale University Press, New Haven
Wolfgang ME (1958) Patterns in criminal homicide. University of Pennsylvania Press, Philadelphia
Zhang L, Wieczorek WF, Welte JW (1997) The impact of age of onset of substance use on delinquency.
J Res Crime Delinq 34:253–268
123
Reproduced with permission of the copyright owner. Further reproduction prohibited without
permission.
Purchase answer to see full
attachment