ArticlePDF Available

Abstract and Figures

Simons and Burt’s (2011) social schematic theory (SST) of crime posits that adverse social factors are associated with offending because they promote a set of social schemas (i.e., a criminogenic knowledge structure) that elevate the probability of situational definitions favorable to crime. The present study extends the SST model by incorporating the role of contexts for action. Further, the study advances tests of the SST by incorporating a measure of criminogenic situational definitions to assess whether such definitions mediate the effects of schemas and contexts on crime. Structural equation models using 10 years of panel data from 582 African American youth provided strong support for the expanded theory. The results suggest that childhood and adolescent social adversity fosters a criminogenic knowledge structure as well as selection into criminogenic activity spaces and risky activities, all of which increase the likelihood of offending largely through situational definitions. Additionally, there was evidence that the criminogenic knowledge structure interacts with settings to amplify the likelihood of situational definitions favorable to crime.
Content may be subject to copyright.
INCORPORATING ROUTINE ACTIVITIES, ACTIVITY SPACES,
AND SITUATIONAL DEFINITIONS INTO THE SOCIAL
SCHEMATIC THEORY OF CRIME*
RONALD L. SIMONS#1, CALLIE H. BURT#2, ASHLEY B. BARR3, MAN-KIT LEI1, and ERIC
STEWART4
1Department of Sociology, University of Georgia
2School of Criminology and Criminal Justice, Arizona State University
3Department of Sociology, University at Buffalo, SUNY
4School of Criminology, Florida State University
# These authors contributed equally to this work.
Abstract
Simons and Burt’s (2011) social schematic theory (SST) of crime posits that adverse social factors
are associated with offending because they promote a set of social schemas (i.e., a criminogenic
knowledge structure) that elevates the probability of situational definitions favorable to crime.
This study extends the SST model by incorporating the role of contexts for action. Furthermore,
the study advances tests of the SST by incorporating a measure of criminogenic situational
definitions to assess whether such definitions mediate the effects of schemas and contexts on
crime. Structural equation models using 10 years of panel data from 582 African American youth
provided strong support for the expanded theory. The results suggest that childhood and
adolescent social adversity fosters a criminogenic knowledge structure as well as selection into
criminogenic activity spaces and risky activities, all of which increase the likelihood of offending
largely through situational definitions. Additionally, evidence shows that the criminogenic
knowledge structure interacts with settings to amplify the likelihood of situational definitions
favorable to crime.
Keywords
criminal propensity; neighborhood; action contexts; routine activities
How do past experiences influence an individual’s propensity to offend? This key
theoretical question drives criminological theorizing. Most theories of individual offending
*Additional supporting information can be found in the listing for this article in the Wiley Online Library at http://
onlinelibrary.wiley.com/doi/10.1111/crim.2014.52.issue-4/issuetoc.
Corresponding Author Contact Sheet: Direct correspondence to Ronald L. Simons, Department of Sociology, University of Georgia,
324 Baldwin Hall, Athens, GA 30602, Phone: 706-424-2626, rsimons@uga.edu.
Co-Author Information: Callie H. Burt: chburt@asu.edu, Ashley B. Barr: abarr@uga.edu, Man-Kit Lei: karlo@uga.edu, Eric Stewart:
estewart2@fsu.edu
HHS Public Access
Author manuscript
Criminology. Author manuscript; available in PMC 2015 November 01.
Published in final edited form as:
Criminology. 2014 November ; 52(4): 655–687. doi:10.1111/1745-9125.12053.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
attempt to identify the elements of criminal propensity and the mechanisms and processes
whereby past experiences give rise to these characteristics. In their recently developed social
schematic theory (SST), Simons and Burt (2011) proposed social schemas to be the key
theoretical mechanisms that account for the effect of past experiences on criminal behavior.
SST emphasizes the role of several criminogenic social environments in shaping social
schemas. In doing so, it integrates findings from a variety of traditions in criminology that
evince the importance of social adversity in shaping criminality, including those related to
neighborhood conditions, parenting, and racial discrimination (e.g., Agnew, 2006; Loeber
and Farrington, 2000; Sampson and Laub, 1993; Tittle, 1995; Unnever and Gabbidon,
2011). What unites these seemingly disparate social influences, according to SST, is that all
teach similar lessons about the future, social norms, and the nature of people and
relationships. As such, learning is central, and SST can be thought of as a life-course
learning theory. However, SST departs from the dominant learning theory in criminology in
several ways. Primary among these is its focus on the content of learning rather than on its
form. Whereas Akers’s (1985) social learning theory emphasizes operant learning
principles, SST shifts the focus to the messages or tenets implicit in the repeated patterns of
interaction that occur in an individual’s social environment. Simons and Burt (2011) argued
that criminogenic conditions such as harsh parenting, racial discrimination, and community
disadvantage promote social schemas involving a hostile view of people and relationships, a
preference for immediate rewards, and a cynical view of conventional norms. Furthermore,
they posited that these three schemas are interconnected and combine to form a
criminogenic knowledge structure (CKS) that gives rise to situational interpretations
legitimating or compelling criminal and antisocial behavior.
In their initial test of the theory, Simons and Burt (2011) found strong support for the SST
model, as the identified social factors strongly influenced individuals’ social schemas, which
in turn increased the likelihood of offending. Indeed, with one exception, the effects of all of
the social factors they examined as well as of sex/gender and prior offending were fully
mediated by the CKS. Additional support for the theory was provided by Simons and Barr
(2012), who reported that much of the effect of romantic relationships on desistance is
explained by a reduction in the CKS. In addition, Burt and Simons (2013) showed that racial
discrimination increased the likelihood of offending through the CKS and that a resilience
factor, racial socialization, reduced offending through its effect on the CKS.
Thus, the initial support for SST is strong and promising. This work can be extended in two
clear ways. First, SST proposes that the CKS increases an individual’s probability of
engaging in crime by making it more likely that situations will be perceived as justifying or
requiring acts of law violation. As a result of data limitations, prior tests of SST were unable
to test the idea that the CKS influenced offending through definitions of the situation. With
the addition of a measure of criminogenic situational definitions in the most recent wave of
the Family and Community Health Study (FACHS), we can test the idea that criminogenic
situational definitions are the mechanism through which CKS increases the likelihood of
offending. This is the first aim of the current study.
SIMONS et al. Page 2
Criminology. Author manuscript; available in PMC 2015 November 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
In addition, in their initial presentation of the theory, Simons and Burt (2011) focused on the
role of social environments as contexts for learning and development. As Wikström et al.
(2012; Wikström and Sampson, 2003) have noted, however, contexts are not only sites for
development but also sites for action. Individuals bring their social schemas into various
contexts, but schemas alone are not sufficient to motivate action. Actions, including crime,
result from the combination of individual characteristics and situational cues. Moreover,
individuals are not randomly placed in various contexts, but they actively seek out certain
contexts consistent their aims and preferences. Consistent with recent work, rather than
viewing selection as a nuisance in modeling, we view it as an important mechanism and
causal force (e.g., Wikström et al., 2012; Sampson, 2012). Thus, we examine whether
individuals’ CKSs influence their likelihood of offending in part by influencing the contexts
in which they choose to spend their leisure time (selection). In addition, we explore the idea
that an individual’s CKS interacts with criminogenic contexts to amplify the likelihood of
criminogenic situational definitions and, in turn, criminal behavior. This idea, as will be
elaborated on in the following section, is that individuals with high CKSs are more likely to
respond to situational inducements with crime than those with lower criminal propensity.
In sum, the purpose of this article is both to elaborate SST and to test this elaboration in a
theory-sensitive research design. In doing so, we incorporate the role of social contexts as
sites both for development and for action into the theoretical model. In addition, we test
whether situational definitions serve as the mechanisms through which social schemas and
contexts influence criminal behavior. In the following pages, we discuss SST, focusing extra
attention on the elaborated role of context as a site for action, drawing especially on insights
and findings from situational action theory (Wikström et al., 2012), crime pattern theory
(Brantingham and Brantingham, 1984), and routine activities theory (Cohen and Felson,
1979). We then test this model using several waves of panel data from a sample of several
hundred African American young adults from the FACHS. Given its inclusion of measures
of both developmental and interactional contexts, as well as a host of other strengths
including its longitudinal design and measures designed to test SST, the FACHS is
particularly well suited for evaluating the elaborated SST model under consideration.
SOCIAL SCHEMAS AND SITUATIONAL DEFINITIONS
SST starts with the assumption that humans adapt to survive in their environments, and a
significant part of this adaptation is cognitive. The theory assumes, consistent with a
burgeoning body of work on human morality, that humans are born with innate capacities to
be fair, cooperative, and sympathetic (e.g., Alexander, 1987; De Waal, 2006; Hauser, 2006)
as well as to be selfish, egoistic, and sometimes aggressive (Shermer, 2004; Smith, 2007).
Rather than being born good, bad, or as empty vessels into which society pours its views of
morality, SST assumes that we are born with the capacity (i.e., the wiring) to adapt our
orientation to fit our environment. Humans have evolved to survive in a variety of contexts,
which vary in the degree to which they are supportive and predictable versus hostile and
dangerous and, thus, require different orientations and competencies (Belsky, Scholomer,
and Ellis, 2012; Ellis et al., 2012). The emphasis is on the fact that individuals adapt to
survive, not necessarily to thrive, in the contexts in which they find themselves and that
egoistic, unkind, and criminal behavior can be incited by such adaptations. Given these
SIMONS et al. Page 3
Criminology. Author manuscript; available in PMC 2015 November 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
assumptions, the theory’s focus is on the role of social environments—especially persistent
and memorable ones—in blunting humans’ innate capacity to be sympathetic, fair, and
cooperative into an orientation that that is cautious, self-defensive, selfish, and even hostile.
From this perspective, offenders do not engage in criminal behavior despite their “morality”
or commitment to conventional norms. Rather, individuals offend because their
interpretations of situations shaped by past experiences lead them to believe that criminal
acts are required or justified by the exigencies of the situation. This perspective is supported
by evidence that most individuals do not believe that their illegal actions are evil or
immoral, but they consider their (mis)deeds to have been compelled by the situation (e.g.,
Black, 1998; Katz, 1988). Thus, crimes result when individuals come to define situations as
requiring or justifying aggression, cheating, or coercion. Undergirded by these insights, SST
aims to explain the process—the underlying mechanisms—that explain individual
differences in situational definitions compelling or justifying crime, which are referred to as
criminogenic situational definitions.
Drawing on insights from information processing theories in cognitive psychology (e.g.,
Dodge and Pettit, 2003) and social learning theories in criminology (e.g., Akers, 1985),
Simons and Burt (2011) emphasized the primary role of persistent or memorable social
experiences in shaping situational definitions. As noted, the emphasis in SST is on the
content of learning. Whereas the dominant learning theory in criminology, Akers’s (1985)
social learning theory, focuses on the process of learning, SST focuses more on the lessons
inherent in the reoccurring interactions that comprise an individual’s existence. As Simons
and Burt (2011) proposed, these lessons are stored as social schemas, which are cognitive
representations of the patterns and messages from past interactions, and these schemas link
social stimuli to future behavior through their effects on situational definitions (see Crick
and Dodge, 1994). Individuals’ social schemas are abstract principles and dispositions that
specify the meaning and salience of various social stimuli and the probable consequences of
various action alternatives (Baldwin, 1992; Crick and Dodge, 1994). Social schemas make
defining and responding to situations both more efficient and more successful as they
suggest which cues are worth noticing, what they mean, what responses are expected or
necessary, and the likely outcomes of various lines of action. Importantly, social schemas
are durable and transposable (Bourdieu, 1990; Sallaz and Zavisca, 2007), although they are
malleable in response to changes in recurring patterns of interaction (Mickelson, Kessler,
and Shaver, 1997; Simons and Burt, 2011).
Focusing on criminal acts, Simons and Burt (2011) proposed that various social factors
identified in past research as strong predictors of criminal behavior increase an individual’s
propensity for crime because they foster social schemas that increase the likelihood of
situational definitions conducive to crime. Offenders are more likely than their conventional
counterparts to have experienced social environmental difficulties and challenges, such as
those related to community disadvantage, inept parenting, criminal victimization, and racial
discrimination. What unites these seemingly disparate social factors is that all impart
messages about the supportiveness and predictability versus the hostility and
unpredictability of the world in a manner that fosters schemas related to crime. The latter
environments nurture a view of the world as harsh and dangerous, where delayed rewards do
SIMONS et al. Page 4
Criminology. Author manuscript; available in PMC 2015 November 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
not predictably materialize, people are untrustworthy, and social rules and punishments do
not apply to everyone equally. According to SST, these fundamental lessons are internalized
as criminogenic social schemas that promote criminogenic definitions of the situation.
Integrating insights from prominent theories of crime and research evidence, SST identifies
three key criminogenic schemas: hostile views of relationships (Anderson, 1999; Dodge,
2006), immediate gratification (or discounting the future; e.g., Gottfredson and Hirschi,
1990; Wikström and Trieber, 2007; Wilson and Herrnstein, 1985), and disengagement from
conventional norms (e.g., Akers, 1985; Hirschi, 1969). Simons and Burt (2011: 556-61)
argued that because these schemas are rooted in the same set of unpredictable and harsh
social conditions, which convey similar messages about the value of delayed gratification,
the nature of relationships and benevolence of others, and the wisdom of following
conventional norms, these three schemas are mutually reinforcing and operate in tandem.
Specifically, SST proposes that these schemas coalesce into a higher order knowledge
structure that incites criminogenic situational definitions: “It is not any one schema that
predicts an individual’s actions in a situation; rather, it is the dynamic interplay of the
constellation of relevant schemas that is important” (Simons and Burt, 2011: 561). This
higher order knowledge structure, referred to as a CKS, exists on a continuum. Individuals at
the low end presumably have experienced consistently supportive, predictable, fair
environments, in general, and thus, they have developed benign views of relationships and
recognize the value of delayed gratification as well as the wisdom of following conventional
rules. At the other (high) end are individuals who have learned to view the world as harsh,
unpredictable, unforgiving, and unjust, and thus, they are more likely to define situations as
justifying or requiring criminal behavior (Burt and Simons, 2013).1
In sum, in their initial presentation of SST, Simons and Burt (2011) proposed that social
experiences influence individuals’ criminality through the messages they convey about the
hostility, predictability, and fairness of the world that are stored in a CKS. Central to this
argument is the role of criminogenic situational definitions, which are proposed to mediate
the effects of the CKS on crime. Although in their initial test of SST, Simons and Burt
(2011) found strong support for their theoretical model, the mediating role of criminogenic
situational definitions was assumed but not tested because of data limitations (the absence of
a measure). Thus, they found that adverse community conditions, racial discrimination, and
harsh parenting increased the risk of crime by increasing affiliation with delinquent peers
and the CKS. The effect of affiliation with delinquent peers on offending was, like sex/
gender and prior delinquency, fully mediated by the CKS. Indeed, with the exception of a
small but significant direct effect of racial discrimination on offending, the effect of all of
the social factors was fully mediated by the CKS. In addition, Simons and Burt (2011) also
showed that changes in exposure to social conditions were associated with changes in the
social schemas, supporting the notion that schemas are durable and transposable but also
malleable in response to changes in social conditions. Two other studies that tested SST also
1There is a shared aspect to schemas as individuals who inhabit similar social positions will have analogous experiences and thus
develop comparable constellations of schemas. The consequence of the similar experiences and comparable schemas among
individuals is similar interpretations of social interactions, expectations, and lines of action (Simons and Burt, 2011). This aspect of
shared experiences shaping similar worldviews is central to cultural sociology and, in particular, the constructs of cultural frames (e.g.,
Lamont and Small, 2008; Snow and Benford, 1992) and cognitive landscapes (Sampson and Wilson, 1995).
SIMONS et al. Page 5
Criminology. Author manuscript; available in PMC 2015 November 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
have provided support for the theory. As noted, SST asserts that a persistently supportive
environment can reduce an individual’s CKS. Consistent with this idea, Simons and Barr
(2014) showed that much of the effect of supportive romantic relationships on criminal
desistance was explained by a reduction in the CKS. A second study by Burt and Simons
(2013) found that the bulk of the effect of racial discrimination on increased offending was
through the CKS and that racial socialization provided resilience to the criminogenic effects
of racial discrimination by buffering the effect of racial discrimination on the CKS as well as
the effect of the CKS on crime.
Although these initial tests of SST provided clear support for much of the theoretical model,
the role of a key mechanism in the theory—criminogenic situational definitions—has not
been tested. Addressing this gap is the first goal of the current study. Similar to Wikström et
al.’s (2012) theory, SST considers two types of situational definitions to be criminogenic.
The first involves perceptions of provocation or threat. Crime is more likely to be viewed as
justified or necessary when individuals perceive danger to their persons or property
(physical threats) as well threats to their status, self- and social-esteem, or reputation (social
and psychological threats). The latter includes, for example, perceptions of disrespect, which
may be indirectly related to physical safety in some milieus (e.g., Anderson, 1999). These
definitions might involve, for example, a perceived threat, slight, or injustice that requires a
forceful reaction.
The second category of criminogenic situational definitions involves perceptions of
opportunity. Individuals may, for example, discern an opportunity for a quick reward or an
immediate benefit that is justified or excused by their views of the harshness and
unpredictability of the social and physical world. Such perceptions tend to be associated
with positive affect and excitement and to increase the chances of engaging in a criminal act
to satisfy a need or want (Wikström et al., 2012). When situational cues are interpreted in
such a way that the actor sees a justified or compelling opportunity to get over on someone
or to obtain a valued resource by bending the rules a bit, crime is more likely. Questions
tapping into criminogenic situational definitions were incorporated into the most recent
wave of the FACHS data to test SST. As such, we provide the first test of the full SST
model with criminogenic situational definitions in the context of a theory-sensitive research
design.
The SST model also can be extended in another important way. As scholars have pointed
out, social environments are not only contexts for learning and development but also
“contexts for action” (Wikström and Sampson, 2003). Contexts for action arguments are
concerned with the immediate effects of context: the way that the characteristics of an area
influence the behavior of the actors operating within it (Wikström et al., 2012). Although
social schemas may specify the import and meaning of stimuli and the intentions and
probable actions of actors in various situations, such schemas have no effect on their own
but only operate in response to situational stimuli. In the initial presentation of SST, Simons
and Burt (2011) only theorized about the role of contexts for development. In this article, we
will elaborate the model to include contexts for action, drawing on Brantingham and
Brantingham’s (1984) crime pattern theory, Wikström’s (2006; Wikström et al., 2012)
SIMONS et al. Page 6
Criminology. Author manuscript; available in PMC 2015 November 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
situational action theory, and routine activities theory (Cohen and Felson, 1979; Osgood et
al., 1996). This is followed by a test of the elaborated SST.
SOCIAL ENVIRONMENTS AS CONTEXTS FOR ACTION
Given the predominance of contexts for development theorizing in criminology, contextual
research has centered almost exclusively on residential neighborhoods (but see, Bernasco et
al., 2013; Wikström et al., 2010). Although this focus may be warranted during the crucial
formative years, it ignores the fact that adolescents and adults spend considerable amounts
of leisure time outside of their residential neighborhoods (e.g., Brantingham and
Brantingham, 2008; Weerman et al., 2013; Wikström et al., 2012). Context for action
theorizing, in particular, requires a shift in focus toward broader contexts with the
recognition that as children age they gain increasing mobility and freedom, “exerting greater
agency in their selection of social environments and greater autonomy in interacting with
them” (Wikström et al., 2012: 46; see also Osgood, Anderson, and Shaffer, 2005). Thus,
context in people’s lives as it relates to crime is not limited to residential neighborhoods, but
it requires a wider consideration of patterns of movement across space at different times, and
this is particularly true of increasingly mobile adolescents and young adults who move about
in space to hang out with their peers.
Brantingham and Brantingham’s (1984, 2008) crime pattern theory provides a useful
framework for viewing context outside of the residential neighborhood. Their theory
highlights routine patterns of travel across space and time. Individuals have a range of daily
activities that are concentrated around various “activity nodes,” such as home, school, work,
entertainment, and shopping, and they develop “routine movement patterns,” which include
the usual path between these activity nodes. Brantingham and Brantingham (2008: 84)
introduced the concept of activity spaces, defined as the “set of normal nodes and the normal
paths between them.” This concept links the individual to the contexts he or she consistently
spends time in, many of which are outside their residential neighborhood. In doing so, it
facilitates the recognition that individuals living in the same residential neighborhood often
spend much of their daily routines in very different settings and that individuals from
different residential neighborhoods can share settings as a result of overlapping activity
spaces (Brantingham and Brantingham, 1984; Wikström et al., 2012).
Activity Space and Risky Activities
Brantingham and Brantingham’s (1984, 2008) concept of activity spaces provides a
framework for distinguishing between residential neighborhoods as contexts for
development and criminogenic activity spaces as contexts for (criminal) actions. Their
discussion of activity spaces focuses on a person’s daily activities as they unfold across
space. Building on their approach, we view interactional settings or action contexts as
consisting of two components that influence the probability of criminal acts. The first
involves what individuals are doing, and the second involves the social and cultural
characteristics of the space where they are doing it. To be sure, activities and activity spaces
are related; however, we differentiate them because they are distinguishable and likely have
independent influences on criminogenic definitions. After all, one can attend a rowdy party
in an area high in social control and watch a movie or play charades in an area that is
SIMONS et al. Page 7
Criminology. Author manuscript; available in PMC 2015 November 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
dangerous and where deviant behavior is prevalent. Additionally, although activities and
spaces both influence the chances of crime through their effects on situational definitions
involving provocation, threat, or criminal opportunity, the two constructs do so in different
ways.
Risky activities increase the probability of criminogenic situational definitions to the extent
that they include, by their nature, a degree of disinhibition and spontaneity, and they involve
interaction within a boisterous crowd of strangers. Using these criteria, activities such as bar
hopping, frequenting strip clubs, hanging out in a pool hall, and getting drunk at a large
sports event are examples of risky activities. For instance, hanging out in a bar with a throng
of intoxicated, animated strangers increases the probability of events such as someone
cutting in line to order a drink, a socially insensitive remark, a purse being left on a bar
stool, or a patron flashing a large sum of money and then stepping into the alley for a smoke.
In contrast, activities such as going to a movie, eating out in a restaurant, watching
television at a friend’s house, or attending a party for close acquaintances are considered low
risk as they rarely involve events entailing provocations, threats, or criminal opportunities.
Although the nature of an activity directly influences the probability that it will lead to social
encounters favorable to crime, activity spaces contribute to criminogenic situational
definitions by dictating the norms and social controls that govern social encounters within an
area regardless of activity. Consistent with insights from Wikström et al. (2012) and drawing
on cultural (e.g., Anderson, 1999), structural/control (e.g., Sampson, 2012; Shaw and
McKay, 1969 [1942]), and routine activities theories (Cohen and Felson, 1979; Felson and
Cohen, 1980; Osgood et al., 2005), we view settings as criminogenic as a function of their
moral norms and the extent of formal and, more importantly, informal social control. The
moral norms of an area as they relate to crime are instantiated in the prevalence of crime and
deviance as well as in the existence of a street culture. Social control is indicated by the
willingness of individuals to intervene in conflicts when someone is breaking the law or
conventional norms (Sampson, Raudenbush, and Earls, 1997). Thus, areas are considered
criminogenic when the norms support deviant behavior and there is low social control. Such
settings increase the likelihood of situational definitions involving provocations, conflict,
and criminal opportunity, thereby making crime more likely.
Selection into Settings
An individual’s routine activities and activity space are not, of course, a connection of
random activity nodes. People do not indiscriminately end up at an opera house instead of at
a strip club. Instead, selection processes are operative; individuals select themselves into
certain settings as a result of their preferences.2 “Selection is a ‘kinds of people in kinds of
settings’ question,” and as several scholars have recently lamented, much prior work treated
selection as a bias to be controlled when examining contextual influences rather than as an
important causal force (Sampson, 2012; Wikström et al., 2012:37). Heeding these critiques,
2Of course, other factors are at play such as structural and cultural constraints. For example: Although my preferences may lead me to
elect to spend my leisure time at a beach-side mansion in Malibu, unfortunately, this option is not realistic given my monetary
resources (or lack thereof). Likewise, many teenagers may prefer to hang out in over-21 clubs, but only some can satisfy this
preference (with a fake ID or social connections).
SIMONS et al. Page 8
Criminology. Author manuscript; available in PMC 2015 November 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
we treat self-selection not as a bias but as an important explanatory factor. Individuals
develop personal characteristics and preferences that influence their participation in various
settings, which are differentially criminogenic. A full theoretical account of the environment
on individual actions, then, requires incorporating (1) the role of context as a site for
development and action and (2) the role that selection plays in linking the two. Wikström
and Sampson (2003: 127) argued, correctly in our view, that “what has been missing [from
criminological theory] is a concept that directly links community context to individual
behavior and actions.” We believe that the CKS provides such a linkage.
Individuals with high CKSs are attracted to risky activities and criminogenic activity spaces.
In such settings, they can engage in deviant behavior with friends unimpeded by
guardianship or conventional morality, which facilitate criminogenic definitions of the
situation. Thus, in addition to its direct effect on criminogenic situational definitions, the
CKS has an indirect effect on such definitions through selection into criminogenic settings.
Importantly, selection does not render the setting irrelevant. As we have noted, actions are in
response to situational stimuli. Individuals enter into situations with various goals (selection
and motivation) but revise those goals and act in light of situational factors (i.e.,
provocations, threats, and opportunities). Thus, although we expect that part of the effect of
risky activities and criminogenic activity spaces on situational definitions is a function of the
CKS, another part of this effect is a result of the features of the setting itself.
Moreover, we expect that the CKS and the setting interact in shaping situational definitions.
Individuals with high CKSs are more likely to attend to, encode, and respond to
criminogenic features of settings and therefore define those settings as compelling or
justifying crime than individuals with lower criminal propensity. Thus, given the same
situational stimuli—a shove in a bar— individuals with high CKS are more likely to respond
with crime—assaulting the pusher—than those with low CKS. Indeed, as Wikström et al.
(2012) have pointed out, individuals with low criminal propensity often do not even
perceive threats or criminal opportunities in the first place.
Summarizing our context for action arguments, we recognize that individuals act and react
in settings. Individuals’ CKSs influence their participation in settings that vary in
criminogenic features (namely, activities, moral norms, and social control); thus, the CKS
has an indirect effect on situational definitions conducive to crime (and, thus, crime itself)
through selection into criminogenic spaces and risky activities. Additionally, criminogenic
spaces and risky activities directly influence criminogenic situational definitions (and, thus,
crime). Finally, the CKS, criminogenic spaces, and risky activities interact such that those
with high CKSs are more likely to define features of criminogenic spaces and risky activities
as conducive to crime than those with a low CKS, and thus, they are more likely to engage
in crime.
CURRENT STUDY
The current study tests this elaborated version of SST, which is presented in figure 1. As
shown, our measures of social adversity focus on the three developmental contexts—quality
of parenting, community context, and racial discrimination—that were included in the
SIMONS et al. Page 9
Criminology. Author manuscript; available in PMC 2015 November 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
analyses reported by Simons and Burt (2011). We expect that these key contexts for
development—as well as sex/gender and past offending—influence individuals’ CKSs.
Contexts for development arguments are grounded in the idea that individuals adapt to their
environments; thus, we expect that persistent exposure to these environments
(operationalized by combining waves III, IV, and V) shapes the CKS measured at waves V
and VI. We hypothesize that the CKS, in turn, increases the likelihood of criminogenic
situational definitions directly, as well as indirectly through selection into risky activities
and criminogenic activity spaces (measured at wave VI). Risky activities and criminogenic
activity spaces, in turn, are expected to influence criminogenic situational definitions
positively. Additionally, we predict that the CKS interacts with criminogenic contexts and
amplifies their effects on criminogenic situational definitions. Finally, we hypothesize that
situational definitions are strongly associated with and fully mediate the effects of these
factors on crime.
The model is tested using waves III-VI of the FACHS, an ongoing investigation of the life-
course trajectories of several hundred African American youth and their families. The
FACHS is particularly well suited for testing the SST model. First, unlike most data sets, the
FACHS data have measures of both contexts for development and action in a longitudinal
design. Moreover, although many criminological theories emphasize the relevance of
situational states prior to the commission of crime, such as definitions, these states are
invariably unmeasured given data limitations. The latest wave of the FACHS data includes a
measure of situational definitions designed to test SST. Thus, the FACHS has many
strengths that make it apposite for testing the SST model. To be sure, the nature of the data,
especially the time intervals between waves, precludes our drawing causal connections for
the observed associations. Thus, the analyses that follow should be viewed as a preliminary
investigation concerned with establishing whether the basic pattern of associations is
consistent with our elaboration of SST.
METHOD
DATA
To test the proposed model, we used the latest four waves of data from the FACHS, an
ongoing investigation of the life-course trajectories of several hundred African American
youth and their families, all of whom were living in Iowa or Georgia at the initiation of the
study. The FACHS was designed to capture the diversity of African American families and
the variety of communities in which they live. Block groups (BGs) were used to identify
neighborhoods in Iowa and Georgia that varied on demographic characteristics, particularly
racial composition (percent African American) and economic level (percent families living
below the poverty line). These BGs (259 in total) were identified using 1990 U.S. Census
data. Families living within the chosen BGs were randomly selected and recruited by
telephone from rosters of all African American families who had a fifth grader (the target
child) in the public school system (Gibbons et al., 2004; Simons et al., 2002).
The first wave of data collection began in 1997-1998, and follow-up interviews with the
target children and their family members were conducted every 2-3 years thereafter. The
current study uses target child data from the third through sixth waves of data, collected in
SIMONS et al. Page 10
Criminology. Author manuscript; available in PMC 2015 November 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
2001-2002, 2004-2005, 2007-2008, and 2010-2011, respectively. These waves of data
capture information from mid-adolescence at wave III through early adulthood at wave VI.
Of the 889 targets interviewed at wave I, 699 (78.6 percent of the original sample)
participated more than a decade later at wave VI.
If targets were unable or unwilling to be interviewed at any given wave, they were not
removed from the study; rather, they were contacted for their participation at subsequent
waves. The analytic sample, then, consists of the 623 individuals (369 women and 254 men)
who provided complete data at wave VI and at least one earlier wave, most of whom (92.13
percent) provided complete data across all four waves used in this analysis. Little evidence
of selective attrition has been found over the course of the study (e.g., Simons et al., 2011).
Although when compared with earlier waves, a higher percentage of the wave VI
respondents were female and were slightly less delinquent, no significant differences were
found between participants and nonparticipants with regard to community measures, family
structure, or parenting practices at earlier waves.
PROCEDURES
To enhance rapport and cultural understanding, African American university students and
community members, all of whom received training in the administration of the self-report
instruments, served as field researchers to collect data from the families. At each wave, the
surveys were administered in the respondent’s home and took an average of 2 hours to
complete. In both waves III and IV, the instruments were presented on laptop computers.
The questions appeared in sequence on the screen, which both the researcher and the
participant could see. The researcher read each question aloud, and the participant entered an
anonymous response using a separate keypad. Because many of the instruments
administered at waves V and VI included questions regarding illegal behavior or potentially
embarrassing sexual activities, audio-enhanced, computer-assisted, self-administered
interviews were used to ensure further anonymity. Using this procedure, the respondent sat
in front of a computer and responded to questions that were presented both on the screen and
via earphones.
MEASURES3,4
Our general approach was to use multiple indicators of constructs when available. Given the
complexity of our model, however, we were not able to treat these multiple measures as
indicators of latent constructs. Rather, when multiple scales were available for a particular
construct, they were standardized and summed to form a composite measure of the variable.
The reliability of these composite constructs was assessed using Nunnally’s (1978) formula
for calculating the reliability of a linear combination of measures. As we will describe, these
coefficients were used in our structural equation models to correct for attenuation in
associations between constructs resulting from measurement error. Notably, structural
equation modeling (SEM) cannot be used to assess quality of measurement when composite
measures are used in place of latent constructs. Two assumptions are especially important
3All measures can be found in appendix A in the online supporting information.
4Additional supporting information can be found in the listing for this article in the Wiley Online Library at http://
onlinelibrary.wiley.com/doi/10.1111/crim.2014.52.issue-4/issuetoc.
SIMONS et al. Page 11
Criminology. Author manuscript; available in PMC 2015 November 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
when composite measures are used. First, the subscales used to assess a particular construct
need to load as a single factor, and second, they should show similar associations with other
variables in the model. We tested these assumptions prior to performing our SEM analyses,
and they were met for each of our composite measures.
Dependent Variable
Crime: Respondents’ engagement in crime was assessed at wave VI using self-reports on a
series of questions regarding how often during the preceding year they had engaged in 11
illegal acts, including physical assault, carrying a hidden weapon, pulling a knife or gun on
someone, shooting or stabbing someone, and breaking into a building or house. Responses
for each act were dummy coded (1 = yes, engaged in act, and 0 = no, did not engage in act)
and then summed at each wave, resulting in a count indicator of the number of criminal acts
in which the respondent participated in the previous year (α = .78). At wave VI, most (80
percent) respondents reported committing zero crimes. Among those who committed at least
one offense, approximately 48 percent reported engaging in two or more offenses,
representing significant variation in individual offending.
The control for prior delinquency is also a variety count of acts committed in the prior year
created by combining youth reports at waves III and IV. The items were gleaned from the
conduct disorder section of the Diagnostic Interview Schedule for Children, Version 4
(American Psychiatric Association, 1994). Respondents answered a series of questions
regarding how often during the preceding year the respondent engaged in 15 antisocial acts
such as shoplifting, physical assault, setting fires, vandalism, burglary, and robbery (for
more detail, see Simons and Burt, 2011). Although based on a different instrument than the
outcome measure, this measure was selected because it captures offending prior to the
measurement of the CKS in the model. (It is worth noting that the pattern of results is
identical whether an equivalent measure of offending is used from wave V and without a
control for prior offending.)
Adolescent Socialization—Our analyses are organized by our proposition that it is
persistent exposure to particular social contexts during late childhood and adolescence that
shapes criminogenic schemas (Simons and Burt, 2011). For these reasons, we average the
scores from waves III, IV, and V (15.5 to 21.5 years of age) to form measures of the
developmental contexts proposed to give rise to a criminogenic knowledge structure.
Supportive Parenting: We formed a composite measure of supportive parenting that
assessed the various components of effective parenting specified by family sociologists and
developmental psychologists (Simons, Simons, and Wallace, 2004). The instruments used in
creating the composite parenting measure across waves III through V were adapted from
scales developed for the Iowa Youth and Families Project (Conger and Elder, 1994) and
were the same as those used by Simons and Burt (2011). Responses for all instruments were
coded such that higher scores correspond to more supportive parenting. Target respondents
answered nine items at waves III and IV and six items at wave V concerning parental
warmth in the past year (e.g., “During the past 12 months, how often did your [Primary
Caregiver] let you know s/he really cares about you?”). Cronbach’s alpha for the warmth
SIMONS et al. Page 12
Criminology. Author manuscript; available in PMC 2015 November 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
scale was .90 at wave III, .91 at wave IV, and .89 at wave V. Target respondents answered
14 items at waves III and IV and 4 items at wave V concerning parental hostility in the past
year (e.g., “During the past 12 months, how often did your [Primary Caregiver] criticize you
or your ideas?”). Cronbach’s alpha for the hostility scale was .81 at wave III, .83 at wave IV,
and .65 at wave V.
In addition to parental warmth and hostility, both parents and target youth answered
questions about effective discipline. At wave III, target respondents also answered two
questions about their primary caregiver’s use of positive reinforcement (e.g., “When you do
something your [Primary caregiver] likes or approves of, how often does s/he let you know
s/he is pleased about it?”). Cronbach’s alpha for the positive reinforcement scale was .58. At
waves III and IV, both targets and primary caregivers also answered four questions about
their ability to solve problems (e.g., “How often do the same problems between you and
your [Primary Caregiver] come up again and again and never seem to get solved?”).
Cronbach’s alpha was .66 for targets and .56 for primary caregivers at wave III and .62 for
targets and .55 for primary caregivers at IV. Finally, at wave III, both targets and primary
caregivers answered five questions about inductive reasoning, or the extent to which
primary caregivers provide explanations for their decisions (e.g., “When you don’t
understand why your [Primary Caregiver] makes a rule for you to follow, how often does
s/he explain the reason?”). Cronbach’s alpha was .86 for targets and .77 for primary
caregivers. The inductive reasoning, problem solving, and positive reinforcement scales
were combined to form a composite measure of effective discipline.
After standardizing and averaging scales across waves to tap into the consistency of parental
support over time, we performed confirmatory factor analysis to establish that the three
parenting subscales (warmth, hostility, and effective discipline) loaded on a common
construct. Factor loadings were all above .5. Furthermore, the various subscales all showed
significant association with CKS. Hence, we standardized and summed the subscales to
form a composite indicator of supportive parenting. The reliability of this composite
measure was .81 and, like the other composite measures we will describe, was calculated
using Nunnally’s (1978) formula for the reliability of a linear combination of measures.
Community Context: For the sake of parsimony, we followed the example of Simons and
Burt (2011) and used a composite measure of community to assess community adversity.
Our composite measure was based on three subscales: community crime, criminal
victimization, and (lack of) collective efficacy. The measure of community crime was
assessed at waves III through V with a revised version of the community deviance scale
developed for the Project on Human Development in Chicago Neighborhoods (PHDCN;
Sampson, Raudenbush, and Earls, 1997). The measure is concerned with how often various
criminal acts occur within the target’s residential community. It includes behaviors such as
fighting with weapons, robbery, gang violence, and sexual assault. Responses ranged from 1
“never” to 3 “often,” and Cronbach’s alpha was .76 at wave III, .87 at wave IV, and .82 at
wave V. The measure of criminal victimization was based on targets’ responses to two items
at waves III through V. These items assessed the number of times that someone in “the
neighborhood surrounding where you lived for most of the past 12 months used violence,
such as in a mugging, fight, or sexual assault, against you or against any member of your
SIMONS et al. Page 13
Criminology. Author manuscript; available in PMC 2015 November 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
household?” and “against one of your friends?” Intercorrelations for these items ranged
from .62 to .83 across waves.
Finally, consistent with Sampson, Raudenbush, and Earls (1997), collective efficacy was
assessed with two subscales, one measuring social cohesion and one measuring informal
social control. Community social cohesion was assessed with a nine-item revised version of
the Social Cohesion and Trust Scale developed for the PHDCN (Sampson, Raudenbush, and
Earls, 1997) that was administered to the primary caregivers at wave III and to the targets at
waves IV and V. The items focus on the extent to which individuals in the area interact,
trust, and respect each other and share values (e.g., “People in your neighborhood do not
share the same values” and “People in this neighborhood can be trusted”). Cronbach’s alpha
for the social cohesion scale was .80 at wave III, .78 at wave IV, and .80 at wave V. The
social control scale, also answered by primary caregivers at wave III and targets at waves IV
and V, consists of six items (also adapted from the PHDCN; Sampson, Raudenbush, and
Earls, 1997) that assess the extent to which individuals in the neighborhood would take
action if various types of deviant behavior were evident. For example, items included the
following: “If some children were spray-painting graffiti on a local building, how likely is it
that your neighbors would do something about it?” and “The adults in the area would not
hesitate to call the authorities if a group of teens were fighting with each other.” Cronbach’s
alpha was .82 at wave III, .85 at wave IV, and .82 at wave V. Both the community cohesion
and social control indices were reversed coded, standardized, and averaged to form a
composite indicator of low collective efficacy.
After standardizing and averaging scales across waves to tap into the consistency of
community context over time, the three community subscales (crime, victimization, and low
collective efficacy) were then standardized and summed to form a composite indicator of
criminogenic community context. This variable was then logged to reduce positive skew.
Confirmatory factor analysis indicated that the three measures formed a one-dimensional
scale, and each of the subscales showed a significant association with CKS. Using
Nunnally’s (1978) formula for calculating the reliability of a linear combination of
measures, reliability for this composite measure was .76.
Racial Discrimination: At waves III through V, target respondents completed 13 items
from a revised version of the widely used and validated Schedule of Racist Events (SRE;
Landrine and Klonoff, 1996). The SRE was originally designed for African American
adults; the FACHS researchers revised the items to make them more appropriate for youth
from late childhood through emerging adulthood. The items assess the frequency (from 1 =
“never” to 4 = “frequently”) with which various discriminatory events were experienced
during the past year (e.g., “How often has someone said something insulting to you just
because of your race or ethnic background?” and “How often has someone suspected you of
doing something wrong just because of your race or ethnic background?”; see Burt et al.,
2012, for a list of items). Cronbach’s alpha was .91 at wave III, .91 at wave IV, and .90 at
wave V. The three scales were averaged across waves to create a measure of persistent
discrimination throughout late adolescence (α = .72).
SIMONS et al. Page 14
Criminology. Author manuscript; available in PMC 2015 November 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
Criminogenic Knowledge Structure—SST proposes that the CKS consists of three
interrelated schemas that come together to form a higher order knowledge structure. Hence,
three subscales (immediate gratification, low commitment to conventional norms, and a
hostile view of relationships) are used to assess the CKS. All subscales were assessed at
both waves V and VI. (For a list of all of the items in the CKS, see Burt and Simons, 2013.)
The first schema, immediate gratification, was assessed via 16 items that combine Kendall
and Williams’s (1982) inventory of self-constraint (e.g., “You would rather have a small gift
today than a large gift tomorrow”) and Eysenck and Eysenck’s (1977) scale of risk-taking
tendency (e.g., “Life with no danger would be dull for you”). The items tap into
respondents’ impulsivity and short sightedness, which are essential elements in Gottfredson
and Hirschi’s (1990) self-control theory. Cronbach’s alpha was .75 at both waves V and VI,
and the reliability across waves was .76. Disengagement from conventional norms was
assessed via 10 items that are similar to those used in Wikström et al.’s (2010) moral values
scale. The respondents were asked to indicate the degree to which they think it is wrong for
someone their age to engage in deviant acts, such as hitting someone in order to hurt them,
stealing or shoplifting, lying, and selling drugs. Cronbach’s alpha was .86 at wave V and .82
at wave VI. The reliability across waves was .82. Finally, the 18-item hostile view of
relationships subscale was designed to measure commitment to a hostile attribution bias
(Dodge, 2006) and consists of two dimensions: a cynical view of others’ intentions (e.g.,
“When people are friendly, they usually want something from you”) and a belief that
aggression is often necessary to avoid exploitation (e.g., “Being viewed as tough and
aggressive is important for gaining respect”). Cronbach’s alpha was .90 at wave V and .89 at
wave VI. The reliability across waves was .75.
Consonant with past studies (Burt and Simons, 2013; Simons and Barr, 2014; Simons and
Burt, 2011), confirmatory factor analyses indicated that the immediate gratification,
disengagement from conventional norms, and hostile view of relationships scales loaded on
a common factor with all loadings greater than .50. Also consistent with Simons and Burt
(2011), the three subscales showed comparable associations with other study variables
including the developmental contexts, risky activities, activity spaces, and crime. Finally,
consistent with SST’s assertion that the three schemas are mutually reinforcing and operate
in tandem, preliminary models using these three indicators as correlated traits rather than a
latent construct fit the data worse than those using the latent construct. Therefore, the scales
were standardized and summed to form a composite measure of criminogenic knowledge
structure. The resulting measure provides an indicator of criminogenic knowledge structure
during the transition to young adulthood (waves V and VI). The reliability of this composite
measure using Nunnally’s formula for a linear combination of measures was .88.
Criminogenic Settings
Risky Activities: At wave VI, we assessed the extent to which respondents spend their free
time in a range of potentially risky activities. Respondents were asked to think about how
they “spend [their] time on a typical weekend evening or night” and then to indicate how
often (1 = never and 5 = weekly) they engage in each of 19 activities (e.g., go bowling, go to
a movie, and watch TV or listen to music at a friend’s house). A focus group with young
SIMONS et al. Page 15
Criminology. Author manuscript; available in PMC 2015 November 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
African American adults was used to generate the list of activities. Eight of these activities,
including bar hopping, clubbing, hanging out at pool halls or strip clubs, and drinking or
getting high, were identified as risky activities that increase the probability of interactions
involving provocation, threat, or criminal opportunity. These items were summed to form a
measure of risky activities. Cronbach’s alpha for the eight-item index was .79. Because of
right skew, we use the logged version of this variable in all models.
Criminogenic Activity Space: After the questions regarding risky activities, respondents
were asked to indicate which activity they do most often and the area of town in which they
engage in said activity. This area indicates the respondents’ primary leisure activity space,
about which they were asked to answer a series of questions regarding its norms and social
controls. Because we conceptualize activity spaces as criminogenic to the extent that they
have low informal social control, a high incidence of criminal and deviant behavior, and a
collective commitment to the street culture, respondents were asked about each of these
components.
The first component, lack of informal social control, was assessed with six items similar to
those used in the community context measure (e.g., “Adults in the area would call the police
if they saw someone breaking the law.”). The responses ranged from 1 “very true” to 3 “not
at all true,” and Cronbach’s alpha for the index was .79. The second component, criminal
activity, also was assessed via six items, each asking about the frequency of criminal and
deviant behaviors that took place in the activity field and of which the respondent and his or
her friends were a part. These criminal and deviant behaviors included things like fighting
with a weapon, the selling of drugs, a sexual assault or rape, and a robbery or mugging.
Responses ranged from 1 “never” to 3 “often,” and Cronbach’s alpha for the index was .85.
Finally, commitment to a street culture was assessed via another six items that asked
respondents how strongly they felt people in the activity field would agree with statements
like the following: “People tend to respect a person who is tough and aggressive” and “It is
important to show other people that one cannot be intimidated.” The responses ranged from
1 “strongly disagree” to 4 “strongly agree,” and Cronbach’s alpha for the index was .94. In
addition to these three subscales, respondents were asked one question concerning how often
they hung out in “tough and dangerous” neighborhoods.
To form a composite measure of criminogenic activity spaces, the low informal social
control, crime, and street culture subscales, along with the “tough and dangerous” frequency
item, were standardized and then summed. Confirmatory factor analysis indicated that the
three subscales loaded on a common factor with all loadings greater than .50, and the
reliability of this composite measure calculated using Nunnally’s formula was .94.
Perception of the Situation
Criminogenic Situational Definitions: Situational definitions were assessed with 12 items
designed to test SST at wave VI. The respondents were asked to indicate how often during
the past year that they had encountered each of 12 different situations that have been
described in ethnographic research as fostering violent and antisocial behavior (e.g., Collins,
2008; Katz, 1988). Half of the items tap into perceived provocations and threats (e.g.,
SIMONS et al. Page 16
Criminology. Author manuscript; available in PMC 2015 November 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
“When you are out and about, how often do you encounter situations where you feel the
other people are not treating you with respect?”), and half relate to perceived opportunities
for getting over someone or an easy score (e.g., “When you are out and about, how often do
you encounter situations where you become aware that there is an opportunity to help
yourself at some sucker’s expense?”). The responses ranged from 1 “never” to 5 “this
happens all the time.” Given that <5 percent of respondents indicated a 4 “very often” or a 5
“this happens all the time,” responses were top coded to range from 1 “never” to 3
“frequently.” These items were summed to form the measure of criminogenic definitions.
Cronbach’s alpha was high at .90.
Control Variables: In all the models we present, the sex of the respondent is controlled.
This variable is coded as 1 = female and 0 = male. Furthermore, as indicated, we control for
prior offending at waves III and IV when predicting wave VI crime to assess the change in
offending in light of persistent exposure to environments.5
ANALYTIC STRATEGY
SEM was used to test our proposed model. Such an approach allows both for the estimation
of substantive parameters simultaneously in the context of a full-information model and
provides tests of significance for specific and general indirect effects. All analyses were
conducted using the statistical program Mplus, Version 7.0 (Muthén and Muthén, 2012).
Because our dependent variable, crime, is an overdispersed count measure, we used a
negative binomial equation model to account for this non-normally distributed outcome.
Furthermore, rather than use latent variables, which would unnecessarily complicate an
already complex and large model, we chose simply to treat the composite measures
described as observed and to specify, rather than to estimate, their measurement error
(Muthén and Muthén, 2012; see figure 2 for an example). With the exception of sex and
crime, we adjust all variables in the model for error in this way.
To assess overall model fit, we use criteria for the comparative fit index (CFI) and root mean
square error of approximation (RMSEA) proposed by Hu and Bentler (1999). A CFI greater
than .95 and an RMSEA smaller than .05 indicate good model fit.6 To compare the models
during the model reduction process as well as the paths constrained and unconstrained by
gender, we conduct chi-square difference tests using Satorra-Bentler scaled chi-square with
robust standard errors (Muthén and Muthén, 2012). Given the non-normality of our count
outcome variable, the Satorra-Bentler chi-square with robust standard errors divides the chi-
square by a scaling correction factor to approximate the chi-square under conditions of non-
normality.
5We also estimated the model controlling for prior offending at wave V using the same instrument used to assess wave VI offending
as well as without any control for prior offending. The pattern of results from these models is identical to that presented here, and the
former is presented in appendix A in the online supporting information.
6Given that the negative binomial estimator requires numerical integration, the indirect effects and model fit statistics cannot be
calculated. Hence, model fit indices and the calculations of indirect effects are based on a continuous model with a non-normality
robust estimator (MLM). Such a model allows for our indirect effects to approximate more closely the effects in the negative binomial
models. Indirect effects are calculated for the unconstrained model that includes paths from all adolescent predictors to all endogenous
variables.
SIMONS et al. Page 17
Criminology. Author manuscript; available in PMC 2015 November 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
RESULTS
DESCRIPTIVE INFORMATION
Table 1 presents the means, standard deviations, and ranges for all study variables for the
analytic sample. Also shown in this table are the zero-order correlations between variables.
The number of criminal acts committed by respondents ranges from 0 to 10, with a mean of .
46 at wave VI. At this later wave, most respondents committed zero violent crimes. Roughly
half (47.66 percent) of those who committed any crimes, however, engaged in two or more
different acts, representing significant individual variation in offending. As expected, the
wave III and IV deviance and wave VI crime measures are significantly correlated at .19 (p
< .001).
Other zero-order correlations among study variables are largely as expected. All variables
are significantly correlated with the dependent variable, crime. Exclusive of prior crime,
these correlations range from −.10 to .34. Furthermore, all of the adolescent social-
environmental variables are significantly related to the adult criminogenic knowledge
structure in the expected directions. Perhaps most importantly, however, the criminogenic
knowledge structure is significantly and positively related to both measures of criminogenic
setting (risky activities: r = .44, p < .001; activity field: r = .48, p < .001) and to
criminogenic definitions of the situation (r = .53, p < .001). Last, as expected, both risky
activities (r = .42, p < .001) and activity field (r = .45, p < .001) are significantly associated
with criminogenic definitions.
SEM RESULTS
Given that the model to be tested is an expansion of past work (Simons and Burt, 2011), we
began our analyses with the full structural model. We then proceeded to improve model fit
by dropping nonsignificant paths (t < 1.5) and by adding paths that were not part of the
hypothesized model but were indicated in the modification indices to be significant. Given
that chi-square difference tests based on log-likelihood values and scaling correction factors
indicated that the model in which effects were free to vary by sex did not fit the data better
than the constrained model (X2 = 22.26, d.f. = 18, p > .05), figure 3 displays the results of
the best fitting model for the full sample. With few exceptions, this final model maps onto
the hypothesized model well. Although model fit indices are not available for the negative
binomial model, the fit indices for the continuous model with the non-normality robust
maximum likelihood estimator indicate that the model fits the data well (CFI = .996;
RMSEA = .045).7,8
Given the complexity of this model, we progress through a discussion of the results in four
stages. First, we focus on the left side of the model to explore the effects of persistent
7We also compared this model with the fully saturated model using the Satorra–Bentler scaled chi-square with robust standard errors.
The nonsignificant chi-square test (chi-square = 6.819, d.f. = 3, p > .05) indicates that the reduced model presented here fits no worse
than the fully saturated model in which all paths are estimated.
8The models presented here contain some overlap with regard to developmental periods (that is, adolescence is measured as waves III
through V, emerging adulthood as wave V, and early adulthood as wave VI). It should be noted that the pattern of findings in models
with no overlapping waves is similar to that presented here. We opted to present results for the overlapping waves (1) given the
increased sample size it afforded and (2) to capture the process of developmental change over time.
SIMONS et al. Page 18
Criminology. Author manuscript; available in PMC 2015 November 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
exposure to harsh, unpredictable environments on the development of a CKS. Doing so aids
in our understanding of the developmental process whereby individuals acquire schemas
conducive to crime. Second, we move toward the right side of the model to explore the
extent to which criminogenic situational definitions mediate the impact of the CKS on
crime. This mediating effect was explicitly hypothesized by Simons and Burt (2011) and is
an essential, yet untested, element of SST. Third, keeping to the right side of the model, we
examine the extent to which both criminogenic situational definitions and crime are a
function of the characteristics of the setting. Hence, we discuss paths linking the CKS to
risky activities and criminogenic activity spaces and those linking these two variables to
both criminogenic definitions and crime. Finally, we examine whether and how the CKS
interacts with characteristics of the setting to enhance criminogenic definitions. We close out
our presentation of results with a discussion of the mediation results.
With respect to the development of the CKS, figure 3 shows that all of the adolescent social-
environmental variables are significantly associated with the criminogenic knowledge
structure as predicted. Whereas both criminogenic community context and racial
discrimination are positively associated with this knowledge structure (γ = .235 and .103,
respectively), supportive parenting is negative in its association (γ = −.269). In addition to
these socialization variables, both sex (γ = −.223) and prior delinquency (γ = .112) are
significantly related to the CKS. Importantly, with the exception of prior delinquency, none
of the adolescent social-environmental variables maintains a direct effect on crime in young
adulthood. Rather, as shown in tables 1 and 2 and as will be discussed in greater detail later,
their effects are indirect, largely through the criminogenic knowledge structure.9 Such
findings are consonant with those of Simons and Burt (2011) and consistent with the SST
model that past experiences influence future offending through their effects on cognitive
schemas about the value of delaying gratification, the wisdom of following conventional
rules, and the trustworthiness and intentions of others.
SST predicts that criminogenic situational definitions account for the link between the CKS
and offending. That is, the development of a CKS is expected to enhance perceptions of
provocation, threat, and opportunity, thereby increasing the likelihood of crime. Consistent
with this idea, figure 3 reveals that the robust link between the CKS and crime can be
explained largely by criminogenic definitions of the situation. As shown on the right side of
the model presented in figure 3, the CKS is positively associated with criminogenic
definitions (β = .346, p < .001), which, in turn, is positively associated with crime (β = .475,
p < .01). As shown in table 2, this indirect effect is highly significant and renders the direct
effect from the criminogenic knowledge structure to crime nonsignificant.
SST proposes that criminogenic situational definitions are a function of both an individual’s
CKS as well as the features of the setting and that individuals with a high CKS select
themselves into criminogenic activity spaces and risky activities. Consonant with these
predictions, figure 3 reveals that the CKS is significantly and positively associated with both
9MPlus has two options, the delta and bootstrapping methods, for calculating the standard errors for indirect effects. The magnitude
and significance levels of effects were found across methods (bootstrap with 1,000 replications). Hence, we present significance levels
based on the default delta method.
SIMONS et al. Page 19
Criminology. Author manuscript; available in PMC 2015 November 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
involvement in risky activities (
β
= .507, p < .001) and criminogenic activity spaces (β = .
401, p < .001), and both of these variables go on to influence criminogenic situational
definitions. More specifically, both risky activities (β = .193, p < .001) and criminogenic
activity (β = .198, p < .001) significantly and positively predict criminogenic situational
definitions. SST further predicts that situational definitions fully mediate the effects of both
the CKS and criminogenic contexts for action on offending. Although the effect of activity
spaces on crime is wholly indirect through criminogenic definitions (see figure 3), the
measure of risky activities continues to have a direct association with crime. In fact,
independent of criminogenic definitions, a 1 standard deviation increase in the logged risky
activities scale predicts more than a twofold increase in the expected count of violent crimes.
We return to this unexpected direct path in the discussion.
Finally, SST predicts that individuals with a high CKS are more likely to attend to
criminogenic features of the situation. This prediction implies an interaction effect,
specifically that the CKS amplifies the effects of criminogenic settings on criminogenic
situational definitions. Consistent with this expectation, and as shown by the dashed lines in
figure 3, the CKS augments the associations between risky activities and situational
definitions and between criminogenic activity spaces and situational definitions. These
moderating effects are illustrated in figures 4 and 5, respectively, and indicate that the
association between risky activities and criminogenic definitions (β = .097, p < .05) as well
as that between activity field and criminogenic definitions (β = .069, p < .05) are stronger for
those with more criminogenic knowledge structures.
With the exception of a direct effect of risky activities on crime, the findings to this point are
largely as predicted by SST. However, three other findings shown in figure 3 were
unexpected. First, racial discrimination had a direct positive effect on criminogenic
situational definitions unmediated by the CKS (γ = .147, p < .001). Hence, independent of
criminal propensity, the experience of discrimination seems to enhance the degree to which
young African Americans define situations as provocative and opportunistic. Second,
although generally supportive of a self-selection effect with regard to characteristics of the
setting, our model suggests that, independent of the CKS, the community context in which
one lives seems to constrain one’s choice of leisure activity spaces, as the direct path from
community context to activity spaces is substantial and significant (γ = .233, p < .001).
Finally, unsurprising albeit not predicted, the direct effect from sex or gender to risky
activities is significant and negative. This finding indicates that, independent of the CKS,
females engaged in these risky activities less frequently than did males (γ = −.162, p < .001).
Aside from the few unexpected findings, the results presented in figure 4 provide much
support for the SST model. Moreover, it should be noted that the model explains a
significant portion of variance in all of our endogenous variables. The proportion of variance
explained ranged from 19 percent for our outcome measure (although this statistic is based
on the continuous model rather than on the more appropriate negative binomial one) to 45
percent for criminogenic definitions of the situation. Furthermore, tables 2 and 3 reveal that
most total and specific indirect effects in the model were statistically significant. For
instance, as shown in table 2, all of the effect of CKS on crime was indirect. In looking at
the specific indirect effects from table 3, one can see that criminogenic definitions (CDS)
SIMONS et al. Page 20
Criminology. Author manuscript; available in PMC 2015 November 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
mediated approximately 25 percent of this effect (indirect effect through CDS = .052/total
effect of .205 = .254), whereas risky routine activities mediated nearly 60 percent (indirect
effect through activities = .121/total effect of .205 = .590). Thus, these findings support SST
contentions about both the important factors and the mechanisms through which these
individual and contextual factors influence the development of criminal propensity and
actual offending.
DISCUSSION
Criminological theories tend to focus on either the role of factors related to the development
of criminal propensity or the situational factors conductive to criminal events, but they
rarely incorporate both (Wikström and Sampson, 2003; but see Wikström et al., 2012).
Furthermore, criminological theories tend to emphasize either identifying salient
criminogenic factors or the processes that link such factors to criminal behavior. As a result,
despite having a rich body of theories, indeed what some would consider a surfeit of
theories, criminology finds itself in a theoretical morass. We argue that in moving forward,
criminology needs more general unifying theories that identify key criminogenic factors and
link these to criminal propensity and events in a relatively parsimonious manner. We
embrace a holistic approach that gives priority to the mechanisms underlying social
influences on both criminal propensity and offending. SST, as presented by Simons and Burt
(2011), is intended to be such a theory. It is grounded in the learning paradigm but improves
on existing learning theories in many ways, especially by being more precise regarding the
key sites of learning and the messages learned and by linking learning to criminal propensity
and events in a life-course model.
This study represented a theoretical elaboration of SST and a test of the model and its
extensions. In particular, two extensions were examined. First, with the addition of
theoretical measures of criminogenic definitions of the situation to the FACHS, we tested
the key idea that criminogenic situational definitions mediate the link between individual
propensities (the CKS) and offending. Furthermore, the SST model was broadened to
include the role of contexts for action in addition to the previous incorporation of contexts
for development. These extensions, tested with waves III through VI of the FACHS data,
will be discussed in this article. This discussion will be followed by a consideration of the
limitations of the current study, the implications of these findings, and directions for future
research.
Consistent with prior tests of SST, the results provide strong support for the theoretical
model. The social-environmental factors we examined, which are theorized to vary in the
key dimensions of supportiveness and predictability versus hostility and dangerousness,
were all strongly associated with the development of the CKS. Specifically, persistent
exposure to supportive parenting was negatively linked to the CKS, whereas racial
discrimination and criminogenic community contexts produced an increase in the CKS.
Additionally, and consistent with SST, being female and prior delinquency were associated
with a lower and higher CKS, respectively.
SIMONS et al. Page 21
Criminology. Author manuscript; available in PMC 2015 November 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
The findings also provide preliminary support for our extensions of the SST model. First,
consistent with a core proposition of the SST model that crime results when individuals
come to define situations as requiring, compelling, or excusing offending, the findings
indicated that much of the effect of both the CKS (and, hence, contexts for development)
and contexts for action on offending is through criminogenic situational definitions. For
example, criminogenic definitions mediated 25.4 percent, 13.5 percent, and 100 percent of
the effect of CKS, risky activities, and activity spaces, respectively, on crime.
The results also are consonant with SST’s context for action arguments, as they show that
both criminal propensities and settings influence involvement in crime; moreover,
propensity and settings interact such that individuals with a high CKS are more likely to
attend to and respond to potentially criminogenic situational cues with offending. Contrary
to the SST model, however, risky activities continued to have a direct effect on offending
after controlling for criminogenic situational definitions. Although this finding was not
expected, it is understandable. The SST model proposes that situational definitions mediate
all of the effects of social and individual factors on offending. We argue that even at a rapid
or reflexive level or when acting out of habit, individuals encode and respond to situational
cues when acting and reacting. However, capturing all of the potential situational definitions
that might result in criminal behavior is an impracticable task in rather large surveys. Thus,
we would argue that this finding is a result of the inherent limitations in measuring
situational definitions. Even so, the measure of situational definitions had a robust effect in
the model, with 25.4 percent of the CKS and 100.0 percent of criminogenic activity spaces
on offending being mediated through such definitions.
Three other significant pathways were not consonant with the SST model presented in figure
1. First, not all of the effects of the criminogenic contexts for development were through the
CKS. Consistent with Simons and Burt (2011), the effect of racial discrimination was not
fully mediated by the CKS, as it had a direct effect on criminogenic situational definitions.
Such a finding is consistent with recent theorizing that has suggested that racial
discrimination increases offending through factors that are unique to the worldview of
African Americans, such as through schemas about the injustices of the criminal justice
system (e.g., Unnever and Gabbidon, 2011). This theory implies that racially specific factors
are operative that are not captured in the SST model.
In addition to its effects through the CKS, criminogenic community context had a direct
positive effect on respondents’ involvement in criminogenic activity spaces. Although not
hypothesized, we believe this effect is likely a result of structural constraints, which
manifest in two different ways. First, individuals who reside in highly disadvantaged,
dangerous, and often isolated communities likely have less mobility. Moreover, such
communities are themselves generally surrounded by similarly situated neighborhoods given
patterns of concentrated disadvantages and spatial interdependence (e.g., Sampson, 2012).
This finding is thus consistent with Sampson’s (2012) argument that ecologically
concentrated neighborhood disadvantage affects individual offending “through the interplay
of structure and purposeful choice” (2012: 64) and that “social choices are governed by
spatial proximity” (2012: 239).
SIMONS et al. Page 22
Criminology. Author manuscript; available in PMC 2015 November 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
Furthermore, being female was not fully mediated by the CKS as it had a direct effect on
participation in risky contexts. Although speculative, we believe that this effect is likely at
least partly caused by the higher degree of monitoring of females (e.g., Hagan, Simpson, and
Gillis, 1979) as well as their greater risk aversion (e.g., Byrnes, Miller, and Schafer, 1999).
Finally, only part of the effect of prior offending was mediated through the CKS. This
finding deserves further research attention and may be a result of structural effects of
offending that are not captured in the current model or related to peer affiliations, which was
not included for the reasons we will discuss next. Overall, however, the findings generally
replicated prior studies of SST.
The study is, of course, not without limitations, and several, in particular, deserve mention.
First, with regard to our measures, all of our constructs, with one exception, relied on
respondents’ self-reports. The exception was the parenting scales, which employed both
primary caregiver and child reports. Individuals’ experiences and perceptions are central to
our model, thus necessitating the use of self-reports for several of our constructs. However,
to the extent possible, future research should incorporate more objective reports of
community and situational conditions. Furthermore, our assessments of situational
definitions, routine activities, activity spaces, and crime were all taken in the same wave of
data collection because of the multiyear intervals between waves. Thus, causal priorities
cannot be established in the current study. Ideally, we would have had multiple waves of
data separated by shorter lengths of time so that causal priorities could have been more
clearly established.
Another limitation that needs to be mentioned is the homogeneity of our sample. All of the
respondents were African American and resided in Iowa and Georgia at the first wave.
Although this raises issues regarding the generalizability of findings, research is needed on
the causes of offending among African Americans given that past research has established
that they suffer from higher rates of crime than other ethnic groups. Use of an African
American sample also had the benefit of allowing us to incorporate racial discrimination into
the model: a factor that recent research has indicated is an important predictor of crime
among African Americans (e.g., Burt, Simons, and Gibbons, 2012; Simons et al., 2006;
Unnever et al., 2009). Although we cannot think of any reason why our results would be
specific to African Americans, our findings clearly need to be replicated with more diverse
samples.
Finally, the SST model we tested did not include deviant or criminal peers. Despite our
belief that peers do influence both propensity and context, we decided not to include peers
given their potential reciprocal relationship with all of the factors in the model, as well as the
fact that our measure of peers is a perceptual measure, which has been shown to be biased
(e.g., Young et al., 2011). Future research needs to examine the interplay of peers with all of
the facets of the SST model in a way that recognizes both peer effects and individual
selection into peer groups.
Despite these limitations, in addition to providing support for the SST model, this study
contributes to criminological knowledge more broadly by supporting an integrative, holistic
approach that combines explanations of propensity and action into a unified developmental
SIMONS et al. Page 23
Criminology. Author manuscript; available in PMC 2015 November 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
model. We believe that criminology needs an updated learning theory that better integrates
extant findings into a life-course model and that SST can fill this gap in theoretical
knowledge. Much more work remains to be done. In addition to testing the theory in
different and especially more diverse samples, the model can be extended in various ways.
For example, the model might incorporate biosocial findings about differential susceptibility
to environmental factors, thus, elucidating individual differences in the effects of social
contexts. In particular, this line of work suggests that some individuals will be more
responsive to environmental conditions, whether supportive or hostile, and thus, we will find
evidence of more change in response to such social factors (Belsky and Pluess, 2009;
Simons, Beach, and Barr, 2012). The SST model also might be refined to incorporate the
relevance of “sensitive periods for change” in response to social conditions, as well as a
consideration of factors that might be more salient at one developmental stage than another
(e.g., Burt, Sweeten, and Simons, 2014; Ellis et al., 2012).
The SST model might be elaborated on to include the role of highly traumatic or memorable
events that may have a much greater influence on the individual than routine daily situations
have. To be sure, such potential experiences are already incorporated into the model in the
form of community criminal victimization, but other traumatic or memorable positive events
may affect the individual and his or her knowledge structures in powerful ways. Future
theorizing and research might consider the effects of such events. Finally, the model might
be expanded to include consideration of transitions and potential turning points and their
effects on offending. As we have noted, one study has already shown that a key adult role
transition, involvement in a satisfying romantic relationship, reduces offending by
decreasing the CKS (Simons and Barr, 2014). Future work might incorporate and test the
effects of other salient life transitions such as work, incarceration, or having a child.
In sum, we believe that SST provides a needed step in the direction of moving
criminological explanations in the direction of more comprehensive, integrated, and
developmental theories that recognize both cumulative continuity as well as the capacity for
change. Clearly, more theoretical work needs to be done in explaining and understanding
crime, and we present SST in this spirit: “It is better to forge ahead and fail than to ignore
the hard questions” (Sampson, 2012: 23). We hope this work stimulates more scholars to
work toward this end whether by challenging or improving on our efforts.
Supplementary Material
Refer to Web version on PubMed Central for supplementary material.
Acknowledgments
This research was supported by the National Institute of Mental Health (MH48165 and MH62669) and the Center
for Disease Control (029136-02). Additional funding for this project was provided by the National Institute on Drug
Abuse (DA021898 and 1P30DA027827) and the National Institute on Alcohol Abuse and Alcoholism
(2R01AA012768 and 3R01AA012768-09S1).
SIMONS et al. Page 24
Criminology. Author manuscript; available in PMC 2015 November 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
Biographies
Ronald L. Simons is a distinguished research professor in the Department of Sociology and
the Center for Contextual Genetics at the University of Georgia. His current research
investigates the manner in which social factors become biologically embedded and influence
development and health across the life course.
Callie H. Burt is an assistant professor in the School of Criminology and Criminal Justice at
Arizona State University. Her research takes a biopsychosocial approach to understand the
pathways through which social factors influence criminal offending and development over
the life course.
Ashley B. Barr is an assistant professor in the Sociology Department at The University at
Buffalo, SUNY. Her research focuses on the development of romantic relationships and
their influence on education, health, and deviance across the life course.
Man-Kit Lei is a research scientist at the Center for Family Research and the Institute for
Behavioral Research at the University of Georgia. His current research focuses on the ways
in which neighborhood factors and genotypes combine to influence well-being across the
life course.
Eric Stewart is a professor of criminology at Florida State University. His research focuses
on contextual and individual dimensions of offending, victimization, and criminal justice
outcomes.
REFERENCES
Agnew, Robert A. Pressured into Crime: An Overview of Strain Theory. Roxbury; Los Angeles, CA:
2006.
Akers, Ronald L. Deviant Behavior: A Social Learning Approach. Wadsworth; Belmont, CA: 1985.
Alexander, Richard D. The Biology of Moral Systems. Aldine; New Brunswick, NJ: 1987.
American Psychiatric Association. Diagnostic Interview Schedule for Children, Version 4 (DISC-IV).
American Psychiatric Association; Washington, DC: 1994.
Anderson, Elijah. Code of the Street: Decency, Violence, and the Moral Life of the Inner City.
University of Chicago Press; Chicago, IL: 1999.
Baldwin, Mark W. Relational schemas and the processing of social information. Psychological
Bulletin. 1992; 112:461–84.
Bernasco, Wim; Stijn, Ruiter; Bruinsma, Gerben J. N.; Pauwels, Lieven J. R.; Weerman, Frank.
Situational causes of offending: A fixed-effects analysis of space-time budget data. Criminology.
2013; 51:895–926.
Belsky, Jay; Pluess, M. Beyond diathesis-stress: Differential susceptibility to environmental
influences. Psychological Bulletin. 2009; 135:885–908. [PubMed: 19883141]
Belsky, Jay; Scholomer, Gabriel L.; Ellis, Bruce J. Beyond cumulative risk: Distinguishing harshness
and unpredictability as determinants of parenting and early life history strategy. Developmental
Psychology. 2012; 48:662–73. [PubMed: 21744948]
Black, Donald. The Social Structure of Right and Wrong. Academic; New York: 1998.
Bourdieu, Pierre. The Logic of Practice. Stanford University Press; Stanford, CA: 1990.
Brantingham, Paul; Brantingham, Patricia. Patterns in Crime. Macmillan; New York: 1984.
SIMONS et al. Page 25
Criminology. Author manuscript; available in PMC 2015 November 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
Brantingham, Paul; Brantingham, Patricia. Crime pattern theory. In: Wortley, Richard; Mazerolle,
Lorraine Green, editors. Environmental Criminology and Crime Analysis. Routledge; New York:
2008.
Burt, Callie H.; Simons, Ronald L. Interpersonal racial discrimination, ethnic-racial socialization, and
offending: Risk and resilience among African American females. Justice Quarterly. 2013 E-pub
ahead of print.
Burt, Callie H.; Simons, Ronald L.; Gibbons, Frederick X. Racial discrimination, ethnic-racial
socialization, and crime: A micro-sociological model of risk and resilience. American Sociological
Review. 2012; 77:648–77. [PubMed: 24058204]
Burt, Callie H.; Sweeten, Gary; Simons, Ronald L. Self-control through emerging adulthood:
Instability, multidimensionality, and criminological significance. Criminology. 2014; 52:450–87.
Byrnes, James P.; Miller, David C.; Schafer, William D. Gender differences in risk taking: A meta-
analysis. Psychological Bulletin. 1999; 125:367–83.
Collins, Randall. Violence: A Micro-Sociological Theory. Princeton University Press; Princeton, NJ:
2008.
Cohen, Lawrence E.; Felson, Marcus. Social change and crime rate trends: A routine activity
approach. American Sociological Review. 1979; 44:588–608.
Conger, Rand D.; Elder, Glen H. Families in Troubled Times: Adaptation to Change in Rural America.
Aldine; New York: 1994.
Crick, Nicki R.; Dodge, Kenneth A. A review and reformulation of social-information processing
mechanisms in children’s social adjustment. Psychological Bulletin. 1994; 115:74–101.
De Waal, Frans. Primates and Philosophers: How Morality Evolved. Princeton University Press;
Princeton, NJ: 2006.
Dodge, Kenneth A. Translational science in action: Hostile attributional style and the development of
aggressive behavior problems. Development and Psychopathology. 2006; 18:791–814. [PubMed:
17152401]
Dodge, Kenneth A.; Pettit, Gregory S. A biopsychosocial model of the development of chronic
conduct problems in adolescence. Developmental Psychology. 2003; 39:349–71. [PubMed:
12661890]
Ellis, Bruce J.; Figueredo, Aurelio José; Brumbach, Barbara H.; Schlomer, Gabriel L. Fundamental
dimensions of environmental risk: The impact of harsh versus unpredictable environments on the
evolution and development of life history strategies. Human Nature. 2009; 20:204–68. [PubMed:
25526958]
Ellis, Bruce J.; Del Giudice, Marco; Dishion, Thomas J.; Figueredo, Aurelio José; Gray, Peter;
Griskevicius, Vladas; Hawley, Patricia H.; Jacobs, W. Jake; James, Jenée; Volk, Anthony A.;
Wilson, David Sloan. The evolutionary basis of risky adolescent behavior: Implications for
science, policy, and practice. Developmental Psychology. 2012; 48:598–623. [PubMed:
22122473]
Eysenck, Sybil B. G.; Eysenck, Hans Jürgen. The place of impulsiveness and venturesomeness in a
dimensional system of personality description. British Journal of Social and Clinical Psychology.
1977; 16:57–68. [PubMed: 843784]
Felson, Marcus; Cohen, Lawrence E. Human ecology and crime: A routine activities approach. Human
Ecology. 1980; 8:389–406.
Gibbons, Frederick X.; Gerrard, Meg; Cleveland, Michael J.; Wills, Thomas A.; Brody, Gene.
Perceived discrimination and substance use in African American parents and their children: A
panel study. Journal of Personality and Social Psychology. 2004; 86:517–29. [PubMed: 15053703]
Gottfredson, Michael R.; Hirschi, Travis. A General Theory of Crime. Stanford University Press;
Stanford, CA: 1990.
Hagan, John; Simpson, John H.; Gillis, AR. The sexual stratification of social control: A gender-based
perspective on crime and delinquency. The British Journal of Sociology. 1979; 30:25–38.
[PubMed: 540220]
Hauser, Marc D. Moral Minds: The Nature of Right and Wrong. Harper; New York: 2006.
Hirschi, Travis. Causes of Delinquency. University of California Press; Berkeley: 1969.
SIMONS et al. Page 26
Criminology. Author manuscript; available in PMC 2015 November 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
Hu, Li-tze; Bentler, Peter M. Cutoff criteria for fit indexes in covariance structure analysis:
Conventional criteria versus new alternatives. Structural Equation Modeling-A Multidisciplinary
Journal. 1999; 6:1–55.
Katz, Jack. Seductions of Crime: Moral and Sensual Attractions in Doing Evil. Basic Books; New
York: 1988.
Kendall, Philip C.; Williams, Carrie L. Assessing the cognitive and behavioral components of
children’s self-management. In: Karoly, Paul; Kanfer, Frederick H., editors. Self-Management and
Behavior Change. Pergamon Press; New York: 1982.
Lamont, Michele; Small, Mario Lois. How culture matters: Enriching our understanding of poverty.
In: Lin, Ann Chih; Harris, David R., editors. The Colors of Poverty. Russell Sage Foundation;
New York: 2008.
Landrine, Hope; Klonoff, Elizabeth A. The schedule of racist events: A measure of racial
discrimination and a study of its negative physical and mental health consequences. Journal of
Black Psychology. 1996; 22:144–68.
Loeber, Rolf; Farrington, David P. Young children who commit crime: Epidemiology, developmental
origins, risk factors, early interventions, and policy implications. Development and
Psychopathology. 2000; 12:737–62. [PubMed: 11202042]
Mickelson, Kristen D.; Kessler, Ronald C.; Shaver, Phillip R. Adult attachment in a nationally
representative sample. Journal of Personality and Social Psychology. 1997; 73:1092–1106.
[PubMed: 9364763]
Muthén, Linda K.; Muthén, Bengt O. Mplus User’s Guide, Seventh Edition. Muthén & Muthén; Los
Angeles, CA: 2012.
Nunnally, Jum C. Psychometric Theory. McGraw-Hill; New York: 1978.
Osgood, D. Wayne; Anderson, Amy L.; Shaffer, Jennifer N. Unstructured leisure in the after-school
hours. In: Mahoney, Joseph L.; Larson, Reed W.; Eccles, Jacquelynne S., editors. Organized
Activities as Contexts of Development. Lawrence Erlbaum; Mahwah, NJ: 2005.
Wayne, Osgood, D.; Wilson, Janet K.; O’Malley, Patrick M.; Bachman, Jerald G.; Johnston, Lloyd D.
Routine activities and individual deviant behavior. American Sociological Review. 1996; 61:635–
55.
Sallaz, Jeffrey J.; Zavisca, Jane. Bourdieu in American sociology, 1980-2004. Annual Review of
Sociology. 2007; 33:21–41.
Sampson, Robert J. Great American City: Chicago and the Enduring Neighborhood Effect. University
of Chicago Press; Chicago, IL: 2012.
Sampson, Robert J.; Laub, John H. Crime in the Making: Pathways and Turning Points throughout
Life. Harvard University Press; Cambridge, MA: 1993.
Sampson, Robert J.; Raudenbush, Stephen W.; Earls, Felton. Neighborhoods and violent crime: A
multilevel study of collective efficacy. Science. 1997; 277:918–24. [PubMed: 9252316]
Sampson, Robert J.; Wilson, William Julius. Toward a theory of race, crime, and urban inequality. In:
Hagan, John; Peterson, Ruth, editors. Crime and Inequality. Stanford University Press; Stanford,
CA: 1995.
Shaw, Clifford R.; McKay, Henry. Juvenile Delinquency and Urban Areas. University of Chicago
Press; Chicago, IL: 1969. [1942]
Shermer, Michael. The Science of Good and Evil. Henry Holt and Company; New York: 2004.
Simons, Ronald L.; Barr, Ashley B. Shifting perspectives: Cognitive changes mediate the impact of
romantic relationships on desistance from crime. Justice Quarterly. 2014; 31:793–821. [PubMed:
25328280]
Simons, Ronald L.; Beach, Steven R. H.; Barr, Ashley B. Differential susceptibility to context: A
promising model of the interplay of genes and the social environment. Advances in Group
Processes. 2012; 29:139–63.
Simons, Ronald L.; Burt, Callie H. Learning to be bad: Adverse social conditions, social schemas, and
crime. Criminology. 2011; 49:553–98. [PubMed: 21760641]
Simons, Ronald L.; Lei, Man-Kit; Beach, Steven R. H.; Brody, Gene H.; Philibert, Robert A.;
Gibbons, Frederick X. Social environment, genes, and aggression: Evidence supporting the
differential susceptibility perspective. American Sociological Review. 2011; 76:883–912.
SIMONS et al. Page 27
Criminology. Author manuscript; available in PMC 2015 November 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
Simons, Ronald L.; Murry, Velma; McLoyd, Vonnie; Lin, Kuei-Hsiu; Cutrona, Carolyn; Conger, Rand
D. Discrimination, crime, ethnic identity, and parenting as correlates of depressive symptoms
among African American children: A multilevel analysis. Development and Psychopathology.
2002; 14:371–93. [PubMed: 12030697]
Simons, Ronald L.; Simons, Leslie Gordon; Burt, Callie H.; Drummund, Holli; Stewart, Eric; Brody,
Gene H.; Gibbons, Frederick X.; Cutrona, Carolyn. Supportive parenting moderates the effect of
discrimination upon anger, hostile view of relationships, and violence among African American
boys. Journal of Health and Social Behavior. 2006; 47:373–89. [PubMed: 17240926]
Simons, Ronald L.; Simons, Leslie Gordon; Wallace, Lora Ebert. Families, Delinquency, and Crime.
Oxford; New York: 2004.
Smith, David Livingstone. The Most Dangerous Animal: Human Nature and the Origins of War. St.
Martin’s Press; New York: 2007.
Snow, David; Benford, Robert D.; Mueller, Carol McChung. Master frames and cycles of protest. In:
Monis, Aldon D., editor. Frontiers in Social Movement Theory. Yale University Press; New
Haven, CT: 1992.
Tittle, Charles R. Control Balance: Toward a General Theory of Deviance. Westview Press; Boulder,
CO: 1995.
Unnever, James D.; Cullen, Francis T.; Mathers, Scott A.; McClure, Timothy E.; Allison, Marisa C.
Racial discrimination and Hirschi’s criminological classic: A chapter in the sociology of
knowledge. Justice Quarterly. 2009; 26:377–409.
Unnever, James D.; Gabbidon, Shaun L. A Theory of African American Offending: Race, Racism, and
Crime. Routledge; New York: 2011.
Weerman, Frank M.; Bernasco, Wim; Bruinsma, Gerben J. N.; Pauwels, Lieven J. R. When is
spending time with peers related to delinquency? The importance of where, what, and with whom.
Crime & Delinquency. 2013 E-pub ahead of print.
Wikström, Per-Olof H. Individuals, settings, and acts of crime: Towards a developmental ecological
action theory of crime involvement and its changes. In: Wikström, Per-Olaf H.; Sampson, Robert
J., editors. The Explanation of Crime: Context, Mechanisms and Development. Cambridge
University Press; Cambridge: 2006.
Wikström, Per-Olof H.; Ceccato, Vania; Hardie, Beth; Treiber, Kyle. Activity fields and the dynamics
of crime: Advancing knowledge about the role of the environment in crime causation. Journal of
Quantitative Criminology. 2010; 26:55–87.
Wikström, Per-Olof H.; Oberwittler, Dietrich; Treiber, Kyle; Hardie, Beth. Breaking Rules: The Social
and Situational Dynamics of Young People’s Urban Crime. Oxford University Press; Oxford,
U.K.: 2012.
Wikström, Per-Olof H.; Sampson, Robert J.; Lahey, Benjamin B.; Moffitt, Terrie E.; Caspi, Avshalom.
Causes of Conduct Disorder and Serious Juvenile Delinquency. Guilford Press; New York: 2003.
Social mechanisms of community influences on crime and pathways in criminality.
Wikström, Per-Olof H.; Treiber, Kyle. The role of self-control in crime causation: Beyond Gottfredson
and Hirschi’s general theory of crime. European Journal of Criminology. 2007; 4:237–64.
Wilson, James Q.; Herrnstein, Richard J. Crime and Human Nature. Simon & Schuster; New York:
1985.
Young, Jacob T. N.; Barnes, JC.; Meldrum, Ryan C.; Weerman, Frank M. Assessing and explaining
misperceptions of peer delinquency. Criminology. 2011; 49:599–630.
SIMONS et al. Page 28
Criminology. Author manuscript; available in PMC 2015 November 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
Figure 1. Elaboration of Social Schematic Theory from Early Adolescence to Young Adulthood
SIMONS et al. Page 29
Criminology. Author manuscript; available in PMC 2015 November 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
Figure 2. Adjusting Observed Variables for Error
a Error = (1 − reliability) × variance.
SIMONS et al. Page 30
Criminology. Author manuscript; available in PMC 2015 November 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
Figure 3. Reduced Negative Binomial Model Predicting All Crime: All Mediating Paths
NOTES: Standardized coefficients presented. Exponentiated unstandardized coefficients
(incident rate-ratios) in parentheses. Female and prior delinquency controlled on all
endogenous variables; only significant paths shown. All exogenous variables correlated. p
= .078; CFI = .996; RMSEA = .045.bp < .10; *p < .05; **p < .01; ***p < .001.
a R2 for this count outcome taken from a continuous, noninteractive model using Satorra–
Bentler scaled chi-square and robust standard errors.
b Fit stats taken from a continuous, noninteractive model using Satorra–Bentler scaled chi-
square and robust standard errors.
SIMONS et al. Page 31
Criminology. Author manuscript; available in PMC 2015 November 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
Figure 4. Effect of Risky Activities on Criminogenic Definitions at High and Low Levels of CKS
SIMONS et al. Page 32
Criminology. Author manuscript; available in PMC 2015 November 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
Figure 5. Effect of Criminogenic Activity Spaces on Criminogenic Definitions at High and Low
Levels of CKS
SIMONS et al. Page 33
Criminology. Author manuscript; available in PMC 2015 November 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
SIMONS et al. Page 34
Table 1
Correlations and Descriptive Statistics for All Study Variables (N = 623)
Variable 1 2 3 4 5 6 7 8 9 10
1 Crime (wave VI) 1.000
2 Prior deviance
(waves III and IV) .188*** 1.000
3 Female −.096*.008 1.000
4 Supportive parenting
(waves III–V) −.131** −.207*** −.0731.000
5 Community context
(waves III–V) .171*** .250*** −.009 −.279*** 1.000
6 Discrimination
(waves III–V) .166*** .124** −.066−.134*** .211*** 1.000
7 Activity fields (wave VI) .295*** .213*** −159*** −.146*** .309*** .110** 1.000
8 Risky activities
(wave VI) .285*** .097*−.241*** −.094*.091*.110** .390*** 1.000
9 Criminogenic
definitions of the
situation (wave VI)
.327*** .218*** −.186*** −.189*** .243*** .233*** .453*** .421*** 1.000
10 Criminogenic
knowledge structure
(waves V and VI)
.339*** .245*** −.200*** −.333*** .290*** .209*** .480*** .442*** .528*** 1.000
Mean .462 .234 .592 .015 1.384 −.015 −.033 2.643 17.455 −.079
Standard deviation 1.248 .555 .492 2.419 .419 .812 2.742 .337 5.063 2.225
Minimum .000 .000 .000 −9.766 .000 −1.232 −3.961 2.079 12.000 −4.893
Maximum 10.000 4.000 1.000 5.375 2.814 3.556 12.121 3.584 36.000 7.199
Reliability .809 .557 .720 .939 .790 .900 .876
p < .10;
*p < .05;
**p < .01;
***p < .001 (two-tailed tests).
Criminology. Author manuscript; available in PMC 2015 November 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
SIMONS et al. Page 35
Table 2
Total and Indirect Effects (N = 623)
Predictors Outcome
Crime CDS Risky Activities Activity Fields
Indirect
Effect Total
Effect Indirect
Effect Total
Effect Indirect
Effect Total
Effect Indirect
Effect Total
Effect
Female −.089 *** −.090 −.159 *** −.189 −.126 *** −.273 −.100 −.162 ***
Supportive parenting −.035 −.045 −.107 *** −.107 −.152 *** −.152 −.121 − 121 ***
Community context .046 .168 .158 *** .231 .132 *** .132 .105 .417 ***
Discrimination .046 *.109 .047 .171 .058 .085 .046 .046
Prior deviance .025 .099 .064 *.064 .060 *.067 .048 .086
CKS .205 *** .205 .195 *** .521
Risky activities .034 *.251
Activity fields .028 .028
NOTES: Standardized indirect effects reported. Indirect effects calculated using the continuous, noninteractive SEM with a non-normality robust MLM estimator. Indirect effects calculated with
unconstrained model.
ABBREVIATIONS: CDS = criminogenic definitions; CKS = criminogenic knowledge structure.
p < .10;
*p < .05;
***p < .001 (two-tailed tests).
Criminology. Author manuscript; available in PMC 2015 November 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
SIMONS et al. Page 36
Table 3
Specific Indirect Effects (N = 623)
Predictor Mediating Path Outcome
Crime CDS
CKS --> Activities --> .121*** .117***
--> Activities --> CDS --> .019*
--> CDS --> .052*
--> Fields --> .077***
--> Fields --> CDS --> .012
Community context --> CKS --> .077**
--> CKS --> Activities --> .029*.028*
--> CKS --> Activities --> CDS --> .004
--> CKS --> CDS --> .012
--> CKS --> Fields --> .018*
--> CKS --> Fields --> CDS --> .003
--> Fields --> .054**
--> Fields --> CDS --> .009
Discrimination --> CDS --> .020
--> CKS --> .034
--> CKS --> Activities --> .013.012
--> CKS --> Activities --> CDS --> .002
--> CKS --> CDS --> .005
--> CKS --> Fields --> .008
--> CKS --> Fields --> CDS --> .001
Female --> Activities --> −.032*−.031*
--> Activities --> CDS --> −.005
--> CKS --> −.073***
--> CKS --> Activities --> .027** −.026**
--> CKS --> Activities --> CDS --> −.004
--> CKS --> CDS --> −.012*
--> CKS --> Fields --> −.017**
--> CKS --> Fields --> CDS --> −.003
Supportive parenting --> CKS --> −.089***
--> CKS --> Activities --> −.033** −.032**
--> CKS --> Activities --> CDS --> −.005
--> CKS --> CDS --> −.014*
--> CKS --> Fields --> −.021**
Criminology. Author manuscript; available in PMC 2015 November 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
SIMONS et al. Page 37
Predictor Mediating Path Outcome
Crime CDS
--> CKS --> Fields --> CDS --> −.003
NOTES: Standardized indirect effects reported. Indirect effects calculated using the continuous, noninteractive SEM with a non-normality robust
MLM estimator. Indirect effects calculated with unconstrained model.
ABBREVIATIONS: CDS = criminogenic definitions; CKS = criminogenic knowledge structure.
p ≤ .10,
*p ≤ .05,
**p ≤ .01,
***p ≤ .001 (two-tailed tests).
Criminology. Author manuscript; available in PMC 2015 November 01.
... Consequently, these three schemas coalesce into a criminogenic knowledge structure (CKS) that fosters situational definitions that make crime appear legitimate and compelling. Several studies have provided support for the tenants of SST (Simons & Burt, 2011;Simons et al., 2014). In the present study, we examine the extent to which repeated exposure to adverse parenting styles increases commitment to CKS, with CKS serving to mediate the impact of these parenting styles on adult crime. ...
... Data for this research come from the Family and Community Health Study (FACHS), a multisite investigation designed to examine family, neighborhood, and peer processes that impact youth development and behavior over time (Simons et al., , 2014. FACHS focuses on a nonprobability sample of 889 African American fifth graders (54 percent girls) and their primary caregivers (PCs) recruited in 1997 from neighborhoods in Iowa and Georgia in which African Americans made up 10 percent or more of the population. ...
... At wave 5 and 6 assessments, participants were asked to indicate how many times in the last year they had engaged in each of 11 criminal acts, such as breaking and entering, larceny, property damage, fighting, and attacking someone with a weapon (Simons et al., 2014). To account for skewness, each item was coded as follows: 0 = never engaged in the activity, 1 = engaged 1-2 times, 2 = engaged 3-5 times, and 3 = engaged 6 or more times. ...
Article
Although several criminological theories suggest that variations in parenting increase the probability of adult crime, most studies limit focus to the association between parenting and adolescent delinquency. Thus, research exploring the association between parenting and adult crime is rare. The present study used path analyses and prospective, longitudinal data from a sample of 318 African American men to examine the effects of eight parenting styles on adult crime. Furthermore, we investigated the extent to which significant parenting effects are mediated by criminogenic schemas, negative emotions, peer affiliations, adult transitions, and involvement with the criminal justice system. Consonant with the study hypotheses, the results indicated that parenting styles with high demandingness, regardless of whether it co‐occurred with responsiveness or corporal punishment, reduced the risk of adult crime. On the other hand, parenting styles low on demandingness but high on responsiveness or corporal punishment were associated with a robust increase in risk for adult crime. These parenting effects were mediated, in large measure, by criminogenic schemas and affiliation with adult deviant peers. The findings held after taking into account the effect of adolescent experiences and traits such as delinquency, deviant peer affiliations, community violence, discrimination, negative emotionality, and poor self‐control.
... The age of samples ranged from 12 to 20 years, with most of the samples representing middle adolescence (13-16 years old). In addition to GPS and GEMA tools already described, some studies used perceived measures of the primary leisure activity space (Simons, Burt, Barr, Lei, & Stewart, 2014) or predetermined activity space locations (Jackson, Browning, Krivo, Kwan, & Washington, 2016) to create networks based on clusters of activity locations (Browning, Soller, & Jackson, 2015). The majority of the studies assessed direct associations between activity space and externalizing, whereas a smaller set examined moderation. ...
... Conversely, Browning et al. (2015) found that eco-network reinforcement, similar to weak social ties or loose social cohesion with shared activities, was negatively associated with delinquency. Although these two studies found direct associations between measures of activity space and problem behaviors, a third found that the effect of activity space risk on offending was mediated by individual's interpretation of the environment as legitimizing or compelling criminal behavior (Simons et al., 2014). The authors suggest that certain individuals select into criminogenic activity spaces, or, to use our dimensions, select into activity spaces with higher social toxicity. ...
Article
Over the last decade, two lines of inquiry have emerged from earlier investigations of adolescent neighborhood effects. First, researchers began incorporating space-time geography to study adolescent development within activity spaces or routine activity locations and settings. Second, cultural-developmental researchers implicated neighborhood settings in cultural development, to capture neighborhood effects on competencies and processes that are salient or normative for minoritized youth. We review the decade’s studies on adolescent externalizing, internalizing, academic achievement, health, and cultural development within neighborhoods and activity spaces. We offer recommendations supporting decompartmentalization of cultural-developmental and activity space scholarship to advance the science of adolescent development in context.
... Crime is increasingly explained by reference to the interplay between person and environment (Barnum and Solomon, 2019;Beier, 2016;Berg et al., 2012;Ernst and Lenkewitz, 2020;Simons et al., 2014;Zimmerman, 2010). In this vein of research, Situational Action Theory (SAT; Wikström, 2006Wikström, , 2014Wikström et al., 2012) provides a particularly comprehensive and detailed framework that explicitly integrates person and environmental approaches, and puts their interplay at the centre of the explanation of crime. ...
Article
Full-text available
Wikström's Situational Action Theory (SAT) explains rule-breaking by reference to the cognitive perception-choice process, which indicates how a person's propensity to break rules interacts with the setting's criminogeneity. SAT's situational model claims that the interaction between personal morality and the moral norms of the setting, the so-called moral filter, is critical in the explanation of rule-breaking, and that the influence of self-control is subordinate to this process. Self-control becomes relevant when individuals whose personal morality discourages rule-breaking are exposed to settings in which the moral norms encourage rule-breaking, that is, if the moral filter is conflicted. Whereas most previous studies have equated the moral filter with personal morality, we consider the moral norms of the setting as well. This allows for a more rigorous test of the moral filter, and thus the conditionality of self-control. Here, we investigate student cheating, using data from two waves of a large-scale German school panel study, and we conceptualise the setting's moral norms by reference to the descriptive norm: other students’ cheating behaviour. This ensures the spatio-linkage between the setting's criminogeneity and rule-breaking, which is necessary for investigating SAT. Additionally, our estimation strategy – person and school fixed-effect models – controls for alternative explanations by the selection of people into settings with different levels of criminogeneity. Moreover, it controls for heterogeneity across persons and schools. The findings are in line with SAT's predictions. In cases of a correspondence between personal morality and the moral norms of a setting, students with rule-abiding morality are least likely to cheat, whereas students with a rule-breaking morality are the most likely to cheat. Also, in line with SAT, self-control only matters for students with rule-abiding morality when they are exposed to moral norms that encourage rule-breaking.
... 40 hours of community service is equivalent to 1 month of imprisonment, but the term of imprisonment should not exceed one year (Government Offices of Sweden, 1965). This approach seems quite correct and is approved by some foreign lawyers (Simons et al., 2018). On the one hand, the term of punishment prescribed in the sentence related to isolation from society has a psychological effect on the convicted person, encouraging him not to bring the situation to the point of replacing mandatory work with a more serious type of punishment. ...
... Moreover, if adolescents are asked to report offending in surveys, some acts of retaliation are likely to be included in their self-reports. Delinquent behavior and attitudes supporting norm-breaking put adolescents on a path towards favoring retaliation, since delinquency contributes to a drift from conventional society (Simons et al., 2014;Sykes & Matza, 1957) and renders cooperation with the police (whose prime task is the enforcement of conventional norms) problematic. Personal experiences of delinquent behavior as well as influence by delinquent peers may reinforce attitudes that favor alternative modes of conflict resolution. ...
Article
Strong and viable modern states have limited the use of private force to narrowly-defined situations of self-defense. Yet, evidence from crime surveys shows that a significant proportion of violent and property crimes is not reported to police. Instead of calling the police, people either take no action or employ a variety of mechanisms, including retaliation, to settle disputes. Drawing on data from a survey of 2,921 young people in two German cities, we investigate the propensity of adolescents to resort to self-justice retaliation. The results show widespread propensity to engage in retaliatory actions, particularly among adolescent males of low socio-economic status. Further, attitudes to the police, unsupervised routine activities, and delinquency-related variables were the most influential correlates of propensity to engage in retaliatory actions.
... One notable example of a theory that attempts to provide a model by which to integrate psychological and sociological explanations of crime is the Social Schematic Theory of crime (SST; Simons & Burt, 2011;Simons, Burt, Barr, Lei, & Stewart, 2014). SST is conceptually less specific than SAT. ...
Chapter
This chapter describes the problems and solutions that justify the need for appropriate analysis of the convergence of people in environments to explain action. Thus, it justifies the research approach and methods that are described and explained in this volume.
... One notable example of a theory that attempts to provide a model by which to integrate psychological and sociological explanations of crime is the Social Schematic Theory of crime (SST; Simons & Burt, 2011;Simons, Burt, Barr, Lei, & Stewart, 2014). SST is conceptually less specific than SAT. ...
Chapter
Research should be driven by theory and served by method. Studying person-environment interaction as defined by Situational Action Theory (SAT) requires an interactive approach. Operationalisation of concepts, data collection methods, and analytical techniques must each be consistent with an interactive worldview and the specific implications of the situational model of SAT. Situational action refers to the behavioural outcome of the interaction of an individual (and their features and state) and an environment (and its characteristics and conditions). In order to study situational interaction, data must therefore capture the spatio-temporally linked convergence of a particular person in a particular environment at the level of the situation and the resultant action. This requires specialist data collection methods, for example, real-world Space-Time Budgets and randomised experimental hypothetical scenarios. The selection and application of analytical techniques must primarily be guided by interactive theoretical principles in order to retain the convergent properties of the situational data. Appropriate methods include comparative presentations of rates, risk ratios, and machine learning techniques that do not divide or nest component parts of the data. In practice, the various features and states of individuals and environments captured in situational data are rarely all measured at the situational level. Therefore, to complement the interactive situational methods, analytical methods may be applied that can take account of methodological obstacles, such as various regression-based techniques that include some account of measurement error. Suggestions of fruitful areas for improvement, development, and replication are included within these recommendations about the data collection and analytical methods that are appropriate for the study of situational interaction.
... One notable example of a theory that attempts to provide a model by which to integrate psychological and sociological explanations of crime is the Social Schematic Theory of crime (SST; Simons & Burt, 2011;Simons, Burt, Barr, Lei, & Stewart, 2014). SST is conceptually less specific than SAT. ...
Chapter
Overall, this SpringerBrief monograph argues that studying the convergence of individuals in environments (situational interaction) is the most appropriate way to study human action, including crime. Despite that being the main argument of the volume, this chapter argues for the continued relevance of evidence of statistical interaction for the study of person-environment (situational) interaction in human behaviour. This is because there are very few studies in any field that can collect the situation-level data about real-world interactions and behavioural outcome that is required for fully appropriate study of situational interaction. In contrast, individual-level exposure data about individuals and environments is much cheaper to collect, but it only allows assessment of statistical interaction (dependence), most commonly using regression methods. This chapter collates and evaluates a range of techniques employed by studies that test the situational model of SAT by assessing person-environment dependency in (most often) individual-level data. In so doing, the chapter highlights the challenges of demonstrating and interpreting dependency, particularly in outcome data that commonly exhibits a problematic distribution (such as crime counts). The chapter concludes that important future directions in the study of statistical interaction are (i) improvements to, and more accurate application of, existing methods, (ii) development and application of new appropriate methods, and (iii) replication in various forms using multiple methods. These future directions are all essential to maintaining and enhancing the utility and relevance of statistical interaction as a complementary element of evidence about the causes of action.
... One notable example of a theory that attempts to provide a model by which to integrate psychological and sociological explanations of crime is the Social Schematic Theory of crime (SST; Simons & Burt, 2011;Simons, Burt, Barr, Lei, & Stewart, 2014). SST is conceptually less specific than SAT. ...
Book
In response to misconceptions and sub-optimal assessment of situational interaction in the criminological literature, this volume is a comprehensive resource for researchers of person-environment interaction in human behavioural outcomes, with a focus on acts of crime. It provides a bridge between strong complex theory about causal situational interaction in crime and the appropriate methods for empirically testing proposed situational mechanisms. It is underwritten by the principle that research should be driven by theory and served by method. This volume clarifies the key concepts of interaction and situation within the framework of Situational Action Theory (SAT). It details the implications of these conceptual issues for an appropriate integrative analytical approach to data collection and analysis that places situational interaction at the heart of research into the causes of behaviour (such as acts of crime). Using existing examples of attempts to analyse person-environment interaction, the volume distinguishes and showcases different methods and evaluates their appropriateness for the study of situational interaction in behaviour. Appropriate for researchers in criminology and the behavioural sciences more generally, Studying Situational Interaction is essential for those studying the individual and environmental causes of human actions such as crime.
... One notable example of a theory that attempts to provide a model by which to integrate psychological and sociological explanations of crime is the Social Schematic Theory of crime (SST; Simons & Burt, 2011;Simons, Burt, Barr, Lei, & Stewart, 2014). SST is conceptually less specific than SAT. ...
Chapter
This chapter describes how adequate study of person-environment interaction requires an integrative model of action that explains precisely how the criminogenic features of individuals and environments interact to result in crime. It presents the situational model of Situational Action Theory (SAT) as the only suitable model of acts of crime. The SAT framework has implications for an integrative analytical approach to data collection and analysis that places situational interaction at the heart of research into the causes of behaviour. In outlining these implications, this chapter clarifies the inter-related concepts of interaction and situation; distinguishes dependence from convergence in order to avoid the ecological fallacy; and defines setting, environment, and situation as distinct concepts to improve empirical testing and interpretation of theory and findings. In order to provide guidance to those wishing to empirically test the situational hypotheses of SAT, the chapter concludes by stipulating the implications of SAT for the family of methodological procedures that are appropriate for the study of situational interaction in acts of crime. Primarily, this involves the collection of event-level situational data that captures the convergence of people in environments.
Book
Why do certain people commit acts of crime? Why does crime happen in certain places? Presenting an ambitious new study designed to test a pioneering new theory of the causes of crime, Breaking Rules: The Social and Situational Dynamics of Young People's Urban Crime demonstrates that these questions can only go so far in explaining why crime happens - and, therefore, in preventing it. Based on the work of the Peterborough Adolescent and Young Adult Development Study (PADS+), Breaking Rules presents an analysis of the urban structure of Peterborough and its relation to young people's social life. Contemporary sciences state that behaviour is the outcome of an interaction between people and the environments to which they are exposed, and it is precisely that interaction and its relation to young people's crime involvement that PADS+ explores. Driven by a ground-breaking theory of crime, Situational Action Theory, which aims to explain why people break rules, it implements innovative methods of measuring social environments and people's exposure to them, involving a cohort of 700 young people growing up in the UK city of Peterborough. It focuses on the important adolescent time window, ages 12 to 17, during which young people's crime involvement is at its peak, using unique space-time budget data to explore young people's time use, movement patterns, and the spatio-temporal characteristics of their crime involvement. Presenting the first study of this kind, both in breadth and detail, with significant implications for policy and prevention, Breaking Rules should not only be of great interest to academic readers, but also to policy-makers and practitioners, interested in issues of urban environments, crime within urban environments, and the role of social environments in crime causation.
Book
Integration of disciplines, theories and research orientations has assumed a central role in criminological discourse yet it remains difficult to identify any concrete discoveries or significant breakthroughs for which integration has been responsible. Concentrating on three key concepts: Context, mechanisms, and development, this volume aims to advance integrated scientific knowledge on crime causation by bringing together different scholarly approaches. Through an analysis of the roles of behavioural contexts and individual differences in crime causation, The Explanation of Crime seeks to provide a unified and focused approach to the integration of knowledge. Chapter topics range from individual genetics to family environments and from ecological behaviour settings to the macro-level context of communities and social systems. This is a comprehensive treatment of the problem of crime causation that will appeal to graduate students and researchers in criminology and be of great interest to policy-makers and practitioners in crime policy and prevention.
Article
A little more than a century ago, the famous social scientist W.E.B. Du Bois asserted that a true understanding of African American offending must be grounded in the –real conditions— of what it means to be black living in a racial stratified society. Today and according to official statistics, African American men - about six percent of the population of the United States - account for nearly sixty percent of the robbery arrests in the United States. To the authors of this book, this and many other glaring racial disparities in offending centered on African Americans is clearly related to their unique history and to their past and present racial subordination. Inexplicably, however, no criminological theory exists that fully articulates the nuances of the African American experience and how they relate to their offending. In readable fashion for undergraduate students, the general public, and criminologists alike, this book for the first time presents the foundations for the development of an African American theory of offending.