Content uploaded by Montelle Felix
Author content
All content in this area was uploaded by Montelle Felix on Oct 31, 2023
Content may be subject to copyright.
Full Terms & Conditions of access and use can be found at
https://www.tandfonline.com/action/journalInformation?journalCode=udbh20
Deviant Behavior
ISSN: (Print) (Online) Journal homepage: https://www.tandfonline.com/loi/udbh20
The Impact of Social Learning and Social Bonds
on Juvenile Delinquency: An Empirical Study of
Secondary School Students in Saint Lucia
Montelle Marius Maradona Felix
To cite this article: Montelle Marius Maradona Felix (14 Oct 2023): The Impact of Social
Learning and Social Bonds on Juvenile Delinquency: An Empirical Study of Secondary School
Students in Saint Lucia, Deviant Behavior, DOI: 10.1080/01639625.2023.2270120
To link to this article: https://doi.org/10.1080/01639625.2023.2270120
Published online: 14 Oct 2023.
Submit your article to this journal
View related articles
View Crossmark data
The Impact of Social Learning and Social Bonds on Juvenile
Delinquency: An Empirical Study of Secondary School Students in
Saint Lucia
Montelle Marius Maradona Felix
Central Police University, Taoyuan City, Taiwan
ABSTRACT
There is a dearth of information regarding the phenomenon of juvenile
delinquency among adolescents in Saint Lucia. Using 268 secondary school
students as a sample, the researcher investigated the criminogenic risk
factors for juvenile delinquent behavior among adolescents from Saint
Lucia. The adolescent patterns of delinquent behavior (violent and nonvio-
lent delinquency) were analyzed in conjunction with conventional crimino-
logical theories (i.e. social learning and social bonds). The results showed that
there was markedly higher delinquency among male adolescents in compar-
ison to female adolescents. In addition, multiple regression and structural
equation modeling demonstrate that the social learning and social bond
variables analyzed in this study were found to be signicantly correlated with
violent and nonviolent delinquency. The ndings oer implications for
addressing the risk factors associated with criminogenic behavior in adoles-
cents and preventing them from engaging in violent or nonviolent delin-
quent behavior.
ARTICLE HISTORY
Received 18 May 2023
Accepted 8 October 2023
Introduction
Overview
The Caribbean region has some of the highest levels of violence in the world (Katz et al. 2023; United
Nations Office on Drugs and Crime [UNODC] 2022; World Health Organization [WHO] 2015).
Seven of the twenty nations with the highest homicide rates (Jamaica, St. Vincent and the Grenadines,
St. Kitts and Nevis, the Bahamas, Trinidad and Tobago, Dominica, and St. Lucia) are located in the
English-speaking Caribbean (UNODC 2021). Furthermore, youth make up a sizable proportion of the
population in many developing countries (between 14 and 25%), and juvenile delinquency adds
significantly to crime rates (Foss et al. 2013). In Jamaica, for example, adolescents aged 13–19 are
responsible for a quarter of major crimes, including armed robbery, assault, rape, and murder (Meeks-
Gardner et al. 2008).
Additionally, youth violence is one of the top priorities facing Caribbean policymakers and civil
society (Moestue, Moestue, and Muggah 2013), as youth violence has reached epidemic proportions in
the region (Agnich and Miyazaki 2013). An increase in violence can stymie the advancement of
effective and efficient democratic administration in Caribbean countries, economic investment in the
region, and general growth and security in civil society. According to the United Nations Development
Programme (UNDP) Citizen Security Survey (2012), which was administered to youth in seven
Caribbean countries (Antigua and Barbuda, Barbados, Guyana, Jamaica, Saint Lucia, Suriname, and
Trinidad and Tobago), Saint Lucia had the highest proportion of reported youth violence (2.7% of
CONTACT Montelle Marius Maradona Felix montelle.felix@gmail.com Department of Crime Prevention and Corrections,
Central Police University, Taoyuan City 33304, Taiwan
DEVIANT BEHAVIOR
https://doi.org/10.1080/01639625.2023.2270120
© 2023 Taylor & Francis Group, LLC
incidents of violence with weapons and 4.3 without). Moreover, according to the same survey, 21.7%
of the youth surveyed carried weapons at night, 16.2% carried weapons during the day, and 32.5% kept
weapons at home (UNDP 2012). This demonstrates how youth violence has limited youth options,
freedom, and opportunity while creating an environment conducive to increased violence. As a result
of this dread of victimization, some young people feel dismal about their chances of living long and
satisfying lives (Carter 2008).
Consequently, as a result of the rise in violence in the Caribbean, scholars have advanced work on
gang involvement (Katz and Fox 2010; Williams 2016), risk and protective factors (Katz, Maguire, and
Choate 2011; Laurent et al. 2011; Maguire and Fishbein 2016; Maguire, Wells, and Katz 2011) and
correlates of offending (Gentle-Genitty et al. 2017). However, despite the historic rise in violence
(UNDP 2012), limited research has been conducted on the prevalence and causes of adolescent
delinquency.
Aim for the study
This study was aimed at identifying the criminogenic risk factors for juvenile delinquent behavior in
Saint Lucian adolescents in a sample of year one and year two (formally referred to as “form one” and
“form two”) secondary school students. Grounded in the theoretical propositions of mainstream
criminological theories (i.e., social learning and social bonds), adolescents’ delinquent patterns (i.e.,
violent and nonviolent delinquency) are examined (Chan 2019). It is envisaged that this study would
prove beneficial to educational institutions and agencies charged with social welfare, crime prevention,
and interdiction; provide an excellent foundation to inform policy implementation, and eventually
redound to the benefit of the community and the country in efforts to avert and prevent crime and
delinquency, but most importantly in employing strategies that can bar young people from commit-
ting deviant acts or becoming hardened criminals over time. Drawing from extant literature from
Akers’ Social Learning Theory and Hirschi's Social Bond Theory, this study poses two central research
questions:
(1) Does increased social learning via negative association with delinquent peers significantly
influence adolescent delinquent behavior?
(2) Do weak social bonds to conventional society (e.g., school, family) significantly influence
adolescent delinquent behavior?
Over the past few years, violence has increased in several schools in Saint Lucia and other countries
within the region (Maharaj, Nunes, and Renwick 2009). According to the Caribbean Human
Development Report on Citizen Security, which reported on school violence in the Caribbean region,
there have been concerns that youth violence has become more vicious (UNDP 2012). According to
research, media, and anecdotal reports, school-based violence is more widespread in secondary
schools in the Caribbean; violent actions committed in this context include bullying, fights, vandalism,
sexual assault, and even homicide (Gentle-Genitty et al. 2017).
Although juvenile delinquency is a well-researched topic in developed countries, there is a paucity
of data (UNDP 2012) and “a serious deficiency of research on youth violence” (Deosaran 2007) in the
Caribbean. This makes Saint Lucia a key consideration for this research, but most importantly, to
focus on educational institutions, particularly the secondary school sector with a juvenile population.
This study on Saint Lucian youth emanates from the researcher’s desire to “expand the conceptual
framework of the discourse on youth delinquency” (Spina 2000) within the Saint Lucian context. This
study is focused on delinquency in schools because the school as a research site is a nexus of global,
regional, national, and local influence (Williams 2016). Furthermore, the school is a proximal context
in terms of its uniqueness to child growth and socialization (Baker 1998) because of the time youth
spend in that socializing space. Therefore, schools can play a major role in bolstering or interrupting
violence and delinquency (Seydlitz and Jenkins 1998).
2M. M. M. FELIX
The present literature
Social learning theory
The social learning theory established by Akers (1998) suggests that criminal behavior can be learned
just like any other behavior. The theory holds that the same learning process in a social context leads to
conforming behavior and deviant behavior. In other words, the difference depends on the direction of
behavioral influences (Akers and Sellers 2004). This theory explains criminal behavior by identifying
variables that induce or control criminal behavior and those that enable or hinder conformity (Akers
and Sellers 2004).
According to Akers’ theory of social learning (1998), four concepts are fundamental to developing
delinquent or conformist attitudes and behaviors: differential association, definitions, differential
reinforcement, and imitation. This study, however, only focuses on differential association and
definition. Each of these elements will be briefly explicated below:
(1) Differential association is associated with direct exposure to values and norms associated with
individuals who engage in specific behaviors (Durkin, Wolfe, and Clark 2005). In addition,
peer associations, such as family and friends, are crucial to the learning process (Akers 1998).
(2) Definitions are an individual’s attitudes and beliefs associated with a particular behavior (Akers
1998). Definitions distinguish between favorable and unfavorable acts, and desirable or unde-
sirable acts (Akers 1998). Moreover, Akers (1998) cites that the probability of criminal activity
increases for those who espouse favorable definitions of crime.
(3) Differential reinforcement exists when rewards and punishments are balanced according to the
behavior rewarded or punished (Akers and Sellers 2004:87).
(4) Imitation refers to performing an action after observing another person performing a similar
act (Akers 1998).
According to social learning researchers, delinquent behavior is related with problematic interactions
and connections with delinquent peers (Molleman, Ciranka, and van den Bos 2022; Seydlitz and
Jenkins 1998; Sullivan 2006). Peer interactions, particularly among males, have been identified as
a significant risk factor for violence and delinquency. When juvenile boys socialize with other
delinquent males, they are more likely to engage in aggressive behavior (Staden 2015). Furthermore,
unresolved emotions such as rage and increasing contact with delinquent peers might enhance the
likelihood that a traumatized youth will react violently (Howe and Parke 2001; Rodgers-Farmer 2000).
Association with delinquent friends also predicts delinquency (Granic and Patterson 2006).
Researchers have found that adolescents who associate with delinquents have a five-fold higher risk
of committing a crime during adolescence than youths who do not associate with delinquents (Burt
and Klump 2013).
Burt and Klump (2013) argue that delinquent peer association cannot be entirely attributed to
peers’ socializing role; however, there is some evidence that the selection process moderately initiates
this association, such that delinquent youth tend to associate with peers who share similar traits
(Granic and Patterson 2006; Hill et al. 2008; Kendler et al. 2008). This technique of selecting peers
undoubtedly implies their shared interest in delinquent activity, particularly during adolescence, when
peer selection is essential in building groups (Hill et al. 2008; Kendler et al. 2008). Delinquent youth
are drawn to delinquent peer groups during adolescence because non-delinquent peer groups reject
them. This is mainly due to the youth’s disorderly behavior (Deater 2001; Hektner, August, and
Realmuta 2000).
Notably, delinquent peers have a more significant impact on the risk of juvenile delinquency among
school-age children than younger children (Gorman-Smith, Tolan, and Henry 2000; Kaufmann et al.
2007; Sullivan 2006). Furthermore, teenage peer interactions have a profound impact on their
decision-making processes and behavioral tendencies.
DEVIANT BEHAVIOR 3
Hirschi’s social bonds theory
Travis Hirschi argues in his classic book “Causes of Delinquency” (1969) that people are inherently
inclined to engage in delinquent behavior. Additionally, Hirschi (1969) suggests that one crucial
factor that hinders an individual from pursuing his or her criminal tendencies is how “bonded” he
or she feels to conventional society. Due to inadequate socialization, weak social bonds increase the
propensity to offend, while solid and high-quality social bonds inhibit this tendency (Peterson
et al. 2016).
Social bonding (also commonly known as social control) may be considered a lone concept in
the sense that all social bonding restrains delinquent behavior, but Hirschi (1969) identified four
interconnected yet independently measured components of social bonding: attachment to parents,
peers, and conventional institutions; commitment to conventional goals; involvement in conven-
tional pursuits, and belief in the moral legitimacy of the rules and norms of society. This study
focuses on two elements of the theory, attachment, and belief. Hirschi (1969) postulates that these
elements exist within three major institutions: family, peers, and school. Strong parental attach-
ment can hinder an individual’s behavior toward societal norms and customs. Attachment to pro-
social peers can restrict an individual’s tendency to engage in delinquent behavior. Lastly, indivi-
duals attached to school may refrain from engaging in delinquent behavior so that their educational
goals will not be jeopardized. Furthermore, the theory emphasizes that the more one is attached to
conventional society, the greater the likelihood of attachment to other aspects of society (Hirschi
1969).
Parent-child attachment is one of the most commonly used research elements of attachment
(Agnew 1991). It comprises the degree of parental supervision, the quality of parental interaction
and time spent with the children, and the parent’s knowledge of their children’s peers and activities
(Hirschi 1969). Furthermore, the transmission of pro-social values from parents to children and their
willingness to invest time in their children are inversely related to the tendency of their children to
commit delinquent acts (Wright, Cullen, and Miller 2001).
Dufur et al. (2019), Hoffmann and Dufur (2018), and Han and Grogan-Kaylor (2015) found
a negative association between attachment to parents and engagement in delinquent behavior
among youth in the United States of America. According to these studies, youth with higher levels
of parent-child attachment are more inclined to report fewer episodes of delinquent behavior. In
addition, other studies on social bonds and delinquency outside the United States found that strong
parent-child bonds are related with fewer incidents of delinquent behavior. Some noteworthy exam-
ples include Korea (Peterson et al. 2016), El Salvador (Springer et al. 2006), Finland (Salmi and
Kivivuori 2006), Germany (Boers et al. 2010), and Turkey (Yuksek and Solakoglu 2016).
Criminologists have extensively studied the correlation between social bonds theory and delin-
quency in schools (Gottfredson 2001; Özbay and Özcan 2006; Payne and Salotti 2007; Peterson et al.
2016). Furthermore, numerous scholarly publications have discovered that belief and school attach-
ment are associated with youth delinquency and delinquency in schools (Costello and Laub 2020;
James and Solomon 2021; Jenkins 1997; Payne 2008; Payne, Gottfredson, and Gottfredson 2003;
Welsh, Greene, and Jenkins 1999; Welsh, Stokes, and Greene 2000). It is commonly established that
violence in schools can lead to adverse outcomes such as academic failure and dropout, criminal
behavior during adulthood, mental health disorders, and drug use (Finkelhor 2008; Gottfredson 2001;
Lawrence 2007; Verdugo 1999). As a result, scholars have extensively employed social bond theory to
identify what influences youths to become delinquent in school (James and Solomon 2021; Payne
2008; Welsh, Greene, and Jenkins 1999).
In addition, Costello and Laub (2020) contend that weak social bonds between children and their
schools correlate with increased school misconduct. Arguably, Welsh, Greene, and Jenkins (1999)
postulate that students who have a solid attachment to school are more liable to conform to the rules
and regulations of the school. Accordingly, Jenkins (1997) asserts that students’ attachment to school
is crucial to the social bond theory for explaining school behavior.
4M. M. M. FELIX
Methodology
Research design
Considering the dearth of research on juvenile delinquency in schools in Saint Lucia, this research
design aimed to explore the criminogenic risk factors for juvenile delinquent behavior in Saint Lucian
adolescents in a sample of year one and year two secondary school students. This study used
a quantitative design which provided descriptive information from participants. The data from this
study derives from a random self-reported survey administered in February 2020 to form one and two
secondary school students from five different secondary schools in Saint Lucia. Considering the
limited time frame, cost constraints, and the variation in the student population among each school,
a total of 300 self-reported questionnaires were evenly distributed. Self-reports of delinquency are
more comprehensive than official reports because they include behaviors that are not reported or
otherwise unknown to authorities. Research has shown that young people are willing to provide
truthful information about minor and severe delinquent acts (Farrington et al. 1996). Furthermore,
self-report research generates a wealth of information about misbehaviors (such as aggression and
defiance) that, while not in themselves criminal, may serve as precursors to later delinquent behavior
(Loeber, Farrington, and Petechuk 2003).
Research procedure and setting
After receiving permission from the Ministry of Education to conduct research within the selected
schools, the researcher scheduled a meeting with the principals and various teachers at the respective
schools and appraised them of the details of the research. The researcher secured commitments with
the school principals to arrange a suitable research setting (e.g., classroom, time, etc.). Before
distributing the printed questionnaires, informed consent forms were given to the students as required
by law (below the age of consent {16 years}) to take to their parents/guardians to grant permission to
participate in the survey. After the consent forms were returned, the self-report questionnaires were
distributed to the students.
During this time, the purpose of the survey was explained, the right to refuse participation was
conveyed, and the level of confidentiality to be employed. While the respondents completed the
questionnaires, the researcher was available to answer any questions or address any concerns they may
have had. The participants were verbally informed that they were free to discontinue if they were
unable to complete the questionnaire. They were also encouraged to take their time and think carefully
while responding to the questions. They were further advised that the surveys would be collected in 45
minutes (one class period) after their initial distribution. After 45 minutes had elapsed and the school
bell rang, the questionnaires were collected. In the event that blank, or damaged questionnaires
appeared, they were disregarded.
Due to time constraints, as mentioned previously, the researcher could not select other students to
replace any respondent quotas that were not met. Through this tightened safeguard process, out of the
300 respondents, 268 students agreed to participate and complete the survey, representing an 89.3%
response rate of the targeted sample
Research measures
Dependent variable
The dependent variable, self-reported delinquency, was measured using an eleven (11) item scale. On
a count measure, Elliott, Huizinga, and Ageton (1985) developed these scales, which are widely
employed in delinquency research. This measure divided self-reported delinquency into two scales:
violent and nonviolent. Participants indicated how often they had engaged in the following behaviors
over the past year. Violent delinquency consisted of four items: 1) Hit someone with the idea of hurting
them, 2) Physically bullied someone, 3) Verbally bullied someone, and 4) Bullied someone via social
DEVIANT BEHAVIOR 5
media (Facebook, Instagram, WhatsApp, etc.). Nonviolent delinquency consisted of seven items: 1)
Skipped school (truancy), 2) Stolen something worth $500.00ECD or less, 3) Used marijuana, 4)
Smoked cigarettes, 5) Sold drugs (marijuana, cocaine), 6) Shared pornographic videos, and 7) Brought
a weapon (knife, firearm, etc.) to school. Response options ranged from 1 (Never) to 5 (more than 4
times).
Independent variables
The independent variables comprise two criminological theories to be employed in this study: Social
Learning Theory and Social Bonds Theory. Two concepts of Social Learning Theory were utilized:
differential association and definitions.
Differential association consisted of seven items; participants were asked about the behavior of their
current friends (schoolmates) who have over the past year: 1) Skipped school, 2) Stolen something
worth ECD $100.00 or less, 3) Sold drugs (marijuana, cocaine), 4) physically bullied someone, 5)
bullied someone via social media (Facebook, Instagram, WhatsApp, etc.), 6) shared pornographic
videos, and 7) brought a weapon (firearm, knife, etc.) to school. Using a modified Likert-type scale,
responses ranged from 1 (none of them) to 4 (all of them). The differential association index combining
the seven items was calculated as the sum of scores divided by 7. Cronbach’s α for the index was
a modest 0.81, with an eigenvalue of 3.34 (47.72%).
Definitions consisted of seven items; participants were asked about their attitude toward deviant
behavior: 1) It’s ok to skip school if nothing important is going on in class, 2) It’s ok to steal, 3) It’s ok
to sell illegal drugs (marijuana, cocaine), 4) it’s ok to bring weapons (guns, knife, scissors, etc.) to
school, 5) it’s ok to physically bully students, 6) Its ok to bully someone via social media (Facebook,
Instagram, WhatsApp, etc.), and 7) It’s ok to share pornographic videos. Using a modified Likert-type
scale, responses ranged from 1 (strongly disagree) to 5 (strongly agree). The definitions index combin-
ing the seven items was calculated as the sum of scores divided by 7. Cronbach’s α for the index was
a modest 0.82, with an eigenvalue of 3.77 (53.86%).
Social bonds theory variables were broken down into four scales to include: beliefs, parent
attachment, teacher attachment, and school attachment.
Beliefs consisted of three items; participants were asked the following questions about their
beliefs: 1) If I do something wrong, people around me will blame me a lot, 2) I believe that I must
be punished if I do something bad, and 3) I believe that if I study hard, I will get good grades. Using
a modified Likert-type scale, responses ranged from 1 (strongly disagree) to 5 (strongly agree). The
belief index combining the three items was calculated as the sum of scores divided by 3. Cronbach’s α
for the index was a modest 0.304, with an eigenvalue of 1.27 (42.36%),
Parent attachment was subdivided into two scales to include female parent/guardian attachment
and male parent/guardian attachment and consisted of five items each; participants were asked the
following questions about their associations with parents/guardians: 1) my male/female parent/
guardian always treats me with love and affection, 2) I can talk freely to my male/female parent/
guardian, 3) I frequently talk about my thoughts and what I experience away from home with my
male/female parent/guardian, 4) my male/female parent/guardian ensure that I complete my assign-
ments on time, and 5) my male/female parent/guardian are concerned about my grades at school.
Using a modified Likert-type scale, responses ranged from 1 (strongly disagree) to 5 (strongly agree).
The female parent/guardian attachment and male/guardian attachment index combining the five
items was calculated as the sum of scores divided by 5 respectively. Cronbach’s α for the indexes was
a modest 0.750 and 0.789, with an eigenvalue of 2.53 (50.72%) and 2.721 (54.42%) respectively.
Teacher attachment consisted of three items; participants were asked about their associa-
tions with their teachers: 1) I have a close relationship with my teachers, 2) my teachers care
about my progress in class, and 3) my teachers ensure that I understand what is taught in
class. Using a modified Likert-type scale, responses ranged from 1 (strongly disagree) to 5
(strongly agree). The teacher attachment index combining the three items was calculated as the
6M. M. M. FELIX
sum of scores divided by 3. Cronbach’s α for the index was a modest 0.62, with an eigenvalue
of 1.76 (58.79%),
School attachment consisted of three items; participants were asked about what they thought
about school: 1) I like to go to school, 2) I think school is boring, and 3) homework is a waste of
my time. Using a modified Likert-type scale, responses ranged from 1 (strongly disagree) to 5
(strongly agree). The school attachment index combining the three items was calculated as the sum
of scores, divided by 3. Cronbach’s α for the index was a modest 0.83, with an eigenvalue of 2.22
(74.18%).
Control variables
The personal characteristics of gender, age, and socioeconomic status (SES) inclusive of parental
employment are included as control variables. Gender is expressed as a dichotomized variable (0=
male and 1= female. Age was measured in years. Parental employment was treated as a dichotomous
variable, where 1 represents parents employed, and 2 represents parents unemployed. Table 1 provides
further detail on how these variables were measured.
Statistical analysis
Using the path analysis program LISREL 8.80, the hypothesized relationship between the variables
was tested (Byrne 2001; El-Sheikh, Abonaze, and Gamil 2017). In addition, the researcher assessed
the pathways from social bond and social learning variables to violent and nonviolent
delinquency.
The hypothesized model was tested. As part of the evaluation, the researcher analyzed various
factors of model fit, such as the nonsignificant Chi-square (Kline 2011), root mean square error of
approximation (RMSEA) of 0.05 or less (Yu 2002), comparative fit index (CFI) of 0.95 or higher, and
adjusted goodness of fit index (AGFI) of 0.90 or higher (Chou et al. 2011). Finally, the paths among the
variables were examined. Direct and indirect effects were analyzed. Coefficients were presented in
both unstandardized and standardized forms. Only significant results were reported for indirect
effects.
Table 1. Variables and descriptive statistics (N = 268).
Variables Description Min Max M SD F %
Dependent Variable
Non-violent Delinquency
Seven-item index 7 28 9.51 4.23
Violent Delinquency: Four-item index 4 19 7.25 3.28
Independent Variable
Social Learning Theory
Definition
Seven-item index, α = .83 7 31 12.91 4.74
Differential Association Seven-item index, α = .81 7 25 11.67 3.65
Social Bond Theory
Belief
Female Parent Attachment
Male Parent Attachment
Teacher Attachment
School Attachment
Control Variable
SES
(Parent Employment)
Three-item index, α = .304
Five-item index, α = .750
Five-item index, α = .789
Three-item index, α = .62
Three-item index, α = .83
1 = Parents Employed
2 = Parents Unemployed
3
8
5
3
9
0
15
25
25
15
25
1
10.51
20.01
17.25
10.90
18.08
0.89
2.54
3.71
4.75
2.45
3.79
0.304
237
27
89.9
10.2
Gender 0 = Male
1 = Female
1 3 1.54 .521 125
138
47.0
51.9
Age 1 = 12 years
2 = 13 years
3 = 14 years
4 = 15 years
1 4 1.81 .751 102
119
42
4
38.2
44.6
15.7
1.5
DEVIANT BEHAVIOR 7
Results
Descriptive statistics
The descriptive statistics for this study are reported in Table 1 and represent a sample of 268
respondents. Overall, participants ranged from 12 to 15 years of age with the majority between
the ages of 12 and 13 years (M = 1.81, SD = 0.751). Gender represented 51.9% of females and
47% of males (M = 1.54, SD = 0.521). With regard to the socioeconomic status (SES) of the
respondents’ parents, more than 89% reported that at least one or both parents were
employed. In contrast, the remaining 10.2% reported that at least one or both of their parents
were unemployed (M = 0.89, SD = 0.304). Moreover, the mean for the dependent variables,
nonviolent delinquency (SD = 4.23) and violent delinquency (SD = 3.28) was 9.51 and 7.25,
respectively, suggesting that the majority of respondents reported engaging in nonviolent
delinquent behavior as compared to violent delinquent behavior.
Regarding social bond variables, the mean for belief (SD = 2.54) was 10.51, suggesting that
half of the respondents reported that it is right to abide by the rules and norms of society.
Not surprisingly, female parent attachment (M = 20.01, SD = 3.71), and male parent attach-
ment (M = 17.25, SD = 4.75) also accounted for about half of the respondents reporting
positive attachments to either parent. Moreover, respondents reported relatively high levels
of school attachment (M = 18.08, SD = 3.79) compared to relatively low levels of teacher
attachment (M = 10.90, SD = 2.45).
Multivariate analysis
Prior to conducting multivariate analysis, the data were cleaned and potential issues such as extreme
outliers, systematically missing values, and non-normality in distributions were all systematically
addressed. In addition, the researcher checked the variance inflation factor (VIF) which assessed the
multicollinearity of independent variables in models (Zuur, Ieno, and Elphick 2010), and the VIF for
all the variables were within the limit criterion of 4 (Kock and Lynn 2012), an observed value which is
within the acceptable range of scores reported in the social science research literature (Tabachnick and
Fidell 1996)
In the multivariate analysis, the variables derived from demographic characteristics models, social
learning, and social bond models were regressed on a scale of nonviolent, violent, and total delin-
quency. The results are presented in Table 2.
Table 2: Multiple Regression summary (N=268)
Nonviolent Delinquency Violent Delinquency Total delinquency
Variables B (β) B (β) B (β)
Age .510 (.088) .477 (.107) 1.109 (.136)**
Gender (female=1, male=0) -1.100 (-.132)** -.011 (-.002) -1.248 (-.106)*
SES (Parental Employment) -2.101 (.149)** 1.642 (.149)* -.733 (-.037)
Female Parent Attachment -.040 (-.032) .031 (-.032) -.048 (-.027)
Male Parent Attachment -.013 (-.014) -.040 (-.055) -.043 (-.033)
Teacher Attachment -.314 (-.176)*** -.105 (-.076) -.453 (-.181)***
School Attachment -.028 (-.024) -.214 (-.214)** -.227 (-.140)*
Belief -.260 (-.145)** -.029 (-.021) -.309 (-.122)*
Differential Association .290 (.311)*** .046 (.063) .333 (.253)***
Definitions .365 (.318)*** .168 (.189)** .521 (.323)***
(Constant) 11.955*** 9.053*** 21.049***
R
2
, Adjusted R
2
.545, .523 .205, .169 .535, .513
F 25.481*** 5.610*** 24.491***
*p≤0.05, **p≤0.01, ***p≤0.001
8M. M. M. FELIX
Nonviolent delinquency
Gender, parental employment, teacher attachment, belief, differential association, and defini-
tions are all statistically significant relative to nonviolent misconduct. These results depict that
males are more prone to nonviolent misconduct. Moreover, students whose parents are unem-
ployed, have low attachment to their teachers, hold little to no belief in societal norms,
intermingle with deviant peers and hold definitions favorable to deviant behavior are more
likely to engage in nonviolent misconduct. Among those significant variables, definitions is the
most robust predictor, followed by differential association, teacher attachment, and parental
employment. Overall, the explanatory variables were able to explain 52.3% of the variance in
this model.
Violent delinquency
parental employment, school attachment, and definitions are statistically significant. These results
suggest that students whose parents are unemployed, have a low attachment to their school, and hold
definitions favorable to deviant behavior are more likely to engage in violent misconduct. Among
those significant variables, school attachment is the most robust predictor, followed by definitions and
parental employment. In general, the explanatory variables were able to explain 16.9% of the variance
in this model.
Total delinquency
Age, teacher attachment, differential association, and definitions are statistically significant. These
results signify that as age increases, so does delinquency. Moreover, students with low attachment
to their teachers, who intermingle with deviant peers and hold definitions favorable to deviant
behavior are more likely to engage in delinquent misconduct. Among those significant variables,
definitions is the most robust predictor, followed by differential association, teacher attachment,
and age. Generally, the explanatory variables were able to explain 51.3% of the variance in this
model.
Path analysis (violent delinquency)
The hypothesized model in Figure 1 includes a representation of the relationship among the social
bond variables (family attachment, school attachment, belief), social learning variables (differential
association and definitions), and a control variable (gender) on violent delinquency.
The pooled model shows a particularly good model fit to the data (Chi-Square {χ
2
} = 1.12, p = 0.771;
(RMSEA) = 0.000; CFI = 1.000; AGFI = 0.990). The unstandardized and standardized results are pre-
sented in Table 3.
Direct eects
Table 3 shows parameter estimates of the violent delinquency path analysis. Gender, social bonds, and
social learning variables had multiple influences on students’ violent delinquent behavior. Gender was
significantly linked to school attachment (β = 1.116, p < .05) and social learning (β =-3.237, p < .001).
Respondents reported low attachments to their school; male respondents reported higher levels of
social learning than their female counterparts. Family attachment exerted a significant direct effect on
school attachment (β = .249, p < .001) and a significant direct effect on belief (β = .069, p < .01). School
attachment was found to have a significant effect on belief (β = .073, p < .05), social learning (β = -.633,
p < .001) and violent delinquency (β = -.182, p < .001). Social learning was significantly associated with
violent delinquency (β = .075, p < .05).
As displayed in Table 3, four of the thirteen variables examined were significant. Family attachment
not only had a significant direct effect on school attachment and belief but also had a significant indirect
effect on belief through school attachment (β = .018, p < .05), social learning through school attachment
DEVIANT BEHAVIOR 9
(β = -.0157, p < .001) and violent delinquency through school attachment (β = -.045, p < .01). School
attachment had a significant direct effect on violent delinquency, and a significant indirect effect on
violent delinquency through social learning (β = -.047, p < .05).
Nonviolent delinquency
The hypothesized model in Figure 2 includes a representation of the relationship among the
social bond variables (family attachment, school attachment, belief), social learning variables
(differential association and definitions), and a control variable (gender) on nonviolent
delinquency. The pooled model shows a particularly good model fit to the data (Chi-Square
{χ
2
} = 6.018, p = 0.111; RMSEA = 0.0608; CFI = 0.991; AGFI = 0.948). The unstandardized and
standardized results are presented in Table 4.
Direct effects
Table 4 shows parameter estimates of the nonviolent path analysis. Age, social bonds and social
learning variables had multiple influences on students’ nonviolent delinquent behavior. Age was
significantly linked to nonviolent delinquency (β = .478*, p < .05). Family attachment exerted
a significant direct effect on school attachment (β = .244, p < .001), and belief (β = .069, p < .01).
School attachment was found to have a significant effect on belief (β = .073, p < .05), social learning (β
= .684, p < .001) and nonviolent delinquency (β = .101, p < .05). Belief had a significant direct effect on
nonviolent delinquency (β = .146*, p < .05). Social learning was found to have a direct effect on
nonviolent delinquency (β = .374, p < .001).
As displayed in Table 4, four of the ten variables examined were significant. Family attachment not
only had a significant direct effect on belief, but also had a significant indirect effect on belief through
school attachment (β = .017, p < .05), social learning through school attachment (β = -.166, p < .001), and
Figure 1. Path analysis showing the relationship among the variables gender, family attachment, school attachment, belief, social
learning, and violent delinquency. Standardized coefficients are in parentheses. *p < .05; **p < .01; ***p < .001.
10 M. M. M. FELIX
nonviolent delinquency through school attachment (β = .162, p < .001). In addition, school attachment
had a significant direct effect on nonviolent delinquency via social learning (β = .255, p < .001).
Discussion
Key ndings
This study investigated the criminogenic risk factors of social learning and social bonds in relation to
the involvement in violent and nonviolent delinquency in a sample of secondary school students in
Saint Lucia. Further discussion of several of the significant findings is warranted. However, the
findings indicate that increased social learning with delinquent peers and weak social bonds can
influence adolescent delinquent behavior.
The first research question was whether increased social learning (i.e., differential association and
definitions) significantly influences adolescent delinquent behavior. The social learning variables, as
evidenced in the multiple regression analysis, were found to be consistently significant across non-
violent delinquency and total delinquency among adolescents. Differential association, however, was
not significant across violent delinquency. However, when the social learning variables were combined
and analyzed in the path analysis model, they were found to have a direct effect on violent delin-
quency. Previous literature indicates that the age-crime curve of delinquency for most adolescents
typically increases after the age of 13, surges to a peak at age 16, and then decline steeply at first to the
mid-20s and, thereafter, more steadily (Farrington 1986, 2003; Loeber and Farrington 2014; McVie
2005; Moffitt, Lynam, and Silva 1994). Furthermore, the descriptive statistics in Table 4 illustrate that
Table 3. Violent delinquency direct and indirect from path analysis (N = 268).
Parameter Estimate Unstd. β Std. β
Direct effects
Gender→School Attachment 1.116* 0.115
Gender→Social Learning −3.237*** −0.247
Family Attachment→School Attachment 0.249*** 0.346
Family Attachment→Belief 0.069** 0.190
Family Attachment→Social Learning 0.057 0.058
Family Attachment→Violent Delinquency −0.013 −0.028
School Attachment→Belief 0.073* 0.146
School Attachment→Social Learning −0.633*** −0.468
School Attachment→Violent Delinquency −0.182*** −0.282
Belief→Social Learning −0.139 −0.052
Belief→Violent Delinquency −0.063 −0.049
Social Learning→Violent Delinquency 0.075* 0.158
Indirect effects
Gender→School Attachment→Belief 0.081 0.016
Gender→School Attachment→Social Learning −0.706 −0.053
Gender→School Attachment→Violent Delinquency −0.203 −0.032
Gender→Social Learning→Violent Delinquency −0.242 −0.039
Family Attachment→School Attachment→Belief 0.018* 0.05
Family Attachment→School Attachment→Social Learning −0.157*** −0.161
Family Attachment→School Attachment→Violent Delinquency −0.045** −0.097
Family Attachment→Belief→Violent Delinquency −0.004 −0.009
Family Attachment→Belief→Social Learning −0.009 −0.009
School Attachment→Belief→Social Learning −0.01 −0.007
School Attachment→Belief→Violent Delinquency 0.004 −0.007
School Attachment→Social Learning→Violent Delinquency −0.047* −0.073
Belief→Social Learning→Violent Delinquency −0.01 −0.008
Note: Chi Square (χ
2
) = 1.12, p = 0.771; Normal Theory Weighted Least Squares Chi-Square = 1.129, p =
0.770; Root Mean Square of Approximation (RMSEA) = 0.000; Comparative Fit Index (CFI) = 1.000;
Adjusted Goodness of Fit Index (AGFI) = 0.990.
+ p < .10; *p < .05; **p < .01; ***p < .001.
DEVIANT BEHAVIOR 11
Figure 2. path analysis showing the relationship among the variables age, family attachment, school attachment, belief, social
learning, and violent delinquency. Standardized coefficients are in parentheses. * p < .05; ** p < .01; *** p < .001.
Table 4. Nonviolent delinquency direct and indirect from path analysis (N = 268).
Parameter Estimate Unstd. β Std. β
Direct effects
Age→Social Learning −0.667 −0.073
Age→Nonviolent Delinquency 0.478* 0.086
Family Attachment→School Attachment 0.244*** 0.340
Family Attachment→ Belief 0.069** 0.190
Family Attachment→ Social Learning 0.079 0.081
Family Attachment→Nonviolent Delinquency −0.035 −0.059
School Attachment→ Belief 0.073* 0.145
School Attachment→ Social Learning −0.684*** −0.504
School Attachment→Nonviolent Delinquency −0.101* −0.123
Belief→Social Learning −0.161 −0.060
Belief→Nonviolent Delinquency −0.146* −0.089
Social Learning→Nonviolent Delinquency 0.374*** 0.616
Indirect effects
Age→Social Learning→Nonviolent Delinquency 0.020 −0.044
Family Attachment→School Attachment→Belief 0.017* 0.049
Family Attachment→School Attachment→Social Learning −0.166*** −0.171
Family Attachment→School Attachment→Nonviolent Delinquency −0.162*** −0.041
Family Attachment→Belief→Social Learning −0.011 −0.011
Family Attachment→Belief→Nonviolent Delinquency −0.010 −0.016
School Attachment→Belief→Social Learning −0.011 −0.008
School Attachment→Belief→Nonviolent Delinquency −0.010 −0.012
School Attachment→Social Learning→Nonviolent Delinquency −0.255*** −0.31
Belief→Social Learning→Nonviolent Delinquency −0.060 −0.036
Note: Chi Square (χ2) = 6.018, p = 0.111; Normal Theory Weighted Least Squares Chi-Square = 5.951, p = 0.114;
Root Mean Square of Approximation (RMSEA) = 0.0608; Comparative Fit Index (CFI) = 0.991; Adjusted
Goodness of Fit Index (AGFI) = 0.948.
+ p < .10 ;*p < .05; **p < .01; ***p < .001.
12 M. M. M. FELIX
the sample accounts for 38.2% of the sampled students were 12 years old, and 44.6% were 13 years old.
Given that most students were between 12 and 13 years old, this phenomenon may suggest that the
differential association was non-significant across violent delinquency may result from the students
not having reached the peak of their adolescent stage. Nonetheless, the findings are coherent with
Akers’ social learning theory of delinquency (Akers 1985, 1998; Akers and Sellers 2004) that involve-
ment in delinquent peer groups is a strong predictor of delinquent behavior (Thornberry 1996),
especially in minor offending (Moffitt 1993; Moffitt and Caspi 2001).
According to Gifford-Smith et al. (2005), Molleman, Ciranka, and van den Bos (2022), and
Simons et al. (1994), deviant peer relationships do not play a major role in an adolescent’s
violent behavior. Although much more research is needed, Loeber, Farrington, and Petechuk
(2003) contend that accelerated paths to delinquency and more serious crime may be attrib-
uted to various factors, including the antisocial tendencies of youth who frequently show early
disruptive behaviors, association with deviant peers, and peer rejection. Consequently, as
children grow and mature, they become assimilated within the school environment and the
larger community, the range of risk factors for delinquency gradually increases. A critical
question that can be asked is whether “birds of a feather flock together?” or does “bad
company corrupt?” According to some hypotheses, delinquent peers can cause youth who
have not shown previous signs of delinquent behavior to become delinquent; they may also
increase delinquency levels in youth who are already delinquent (Loeber, Farrington, and
Petechuk (2003).
The second research question was whether weak social bonds to conventional society (e.g., school,
family) significantly influence adolescent delinquent behavior. The social bonding variables demon-
strated in this study were crucial factors associated with violent and nonviolent delinquency.
Essentially, adolescents with weak bonds to conventional society are more prone to engage in violent
and nonviolent delinquent behaviors.
The results showed that students who engaged in nonviolent and total delinquent acts reported
having low attachment to their teachers. Moreover, students who committed violent and total
delinquent acts reported low attachment to their school. The path analysis model illustrated that
although school attachment (a combination of teacher and school attachment) directly impacted both
violent and nonviolent delinquency, school attachment had a greater negative impact on violent
delinquency. Nonetheless, these results are consistent with previous studies (e.g., Cunningham
2007; Popp and Peguero 2012), which exemplify that delinquent behavior is correlated to low
academic commitment and/or school disengagement. Conversely, Sprott, Jenkins, and Doob (2005)
highlighted the importance of schools in protecting at-risk youth from violent and nonviolent
delinquency by alleviating risk factors such as early aggression, cumulative risks, and peer deviance.
In addition, a low belief in an internalized moral code was also found to impact delinquency,
particularly in nonviolent delinquency. Adolescents who did not have an internalized conventional
moral code and the belief that it is right to abide by the rules and norms of society were found to
engage in nonviolent acts of delinquent behavior. Surprisingly, male and female parental attachment
remained nonsignificant across the multiple regression and path analysis models. Consistent with past
social bond studies, adolescents are less likely to engage in delinquent behavior if they have a healthy
and secure relationship with their parents (Chan and Wong 2019). A recent meta-analysis piloted by
Hoeve et al. (2012) surveyed over 70 studies, predominantly Western samples, suggesting that
delinquency is significantly associated with poor parent-child bonds regardless of gender. Evidently,
a strong parent-child bond is a protective factor that prevents adolescents from participating in
delinquent behaviors (Chan, 2015; Chan and Chui 2015).
Although socio-demographic variables relating to delinquency were not the focus of the study,
some findings were worth mentioning. First, male adolescents reported significantly more nonviolent
delinquency than their female counterparts. Moreover, parental socio-economic status (SES) or
parental employment was significant across nonviolent and violent delinquency. Research has
shown that the delinquent behavior of youths from low-SES families is more prevalent than that of
DEVIANT BEHAVIOR 13
youths from high-SES families (Bjerk 2007; Jarjoura, Triplett, and Brinker 2002). Conversely, only
focusing on parental employment can provide limited views about the link between SES and delin-
quency (Gutierrez and Shoemaker 2008; Ring and Svensson 2007).
Implications of ndings
Findings from this study could have implications for practices aimed at preventing juvenile delin-
quency. Moreover, delinquency was an important factor influencing adolescents’ engagement in both
violent and nonviolent delinquent behavior. Adolescents become more oriented toward their peers
and spend more time away from their parents as they mature (Sullivan 2014). This study should
particularly emphasize the negative influences of peers on adolescents because the variables associated
with social learning were a powerful predictor of deviant behavior among Saint Lucian adolescents.
Peer educators and counselors may provide adolescents with an effective method to prevent them
from drifting into deviant peer groups (Davis, Tang, and Ko 2000).
Additionally, a healthy parent-child relationship is undoubtedly protective of adolescent delin-
quency. Effective parenting, including parental guardianship and efficient parent-child interaction, is
essential in promoting healthy psychosocial development in adolescents. Psychosocial well-being (e.g.,
self-efficacy, empathy, increased self-esteem, and self-control) can reduce adolescents’ proclivity to
espouse delinquent behavior (Chan and Chui 2015). Hoeve et al. (2012) observed that parental control
and attachment (i.e., parental supervision, setting rules, and discipline) are likely to affect the
adolescent’s behavior more than the parent-child relationship alone. School administrators, counse-
lors, and social workers can identify problematic adolescents (e.g., those with school discipline
problems, and members of gangs) to study their family dynamics. If they are found to have weak
parental relationships, the involvement of social service providers is critical in bridging a gap between
parents and adolescents and enabling more effective parent-child bonding and healthy family func-
tioning. Nurturing a commitment to educational values is especially essential in preventing at-risk
youth from violent and nonviolent offending.
Implementing these recommendations can only be successful if all parties (e.g., parents, caregivers,
school administrators, teachers, and social workers) work together. Individual efforts (school inter-
ventions without parental involvement or vice versa) may not be as successful in preventing adoles-
cents from participating in delinquent behaviors.
Research limitations and conclusion
The results of this study should be interpreted with caution, as it is not methodologically
flawless in the purview of numerous limitations. First, the results were limited due to the self-
reported data used in this study. Social desirability and memory recall biases may lead to
underreporting of delinquent behaviors. Researchers should consider incorporating a response
bias measure into future studies. A second limitation of this study is that the association
between the adolescents’ criminogenic factors and self-reported delinquent behaviors was not
examined due to its cross-sectional design. Moreover, understanding the offending phenomena
associated with this population requires a longitudinal approach in future research. Thirdly,
the researcher was unable to increase the sample size and the number of schools to be
sampled due to time limitations and financial constraints. For example, the researcher did
not factor in the timeframe needed to receive permission from the Ministry of Education; or
the time it would take to arrange a meeting with the school principals and teachers at the
selected schools. Subsequent research should be adequately planned, and prior investigations
must be conducted to understand the necessary prerequisites needed to undertake research of
this caliber within Secondary Schools. Fourthly, an inherent limitation of this study relates to
the non-independence of observations, which are nested within the hierarchical structure of
schools. The clustering of observations within schools introduces the potential for downward
14 M. M. M. FELIX
bias in standard error estimation, subsequently affecting the interpretation of statistical sig-
nificance associated with estimated coefficients. This clustering issue has been acknowledged
within the study. Future research can address this limitation by adopting a more collaborative
approach involving multiple researchers and/or by conducting studies with larger, diverse
samples. This approach can help mitigate the clustering issue by providing a broader repre-
sentationof observations across different schools or clusters. Additionally, future studies can
employ advanced statistical techniques specifically designed for clustered data from the outset,
allowing for more accurate standard error estimation and robust statistical analyses.
Furthermore, researchers can explore alternative study designs that minimize or account for
clustering effects, such as conducting studies at the individual level rather than within clusters.
These proactive measures can contribute to more reliable and generalizable research outcomes
in future investigations.
Finally, the author recognizes the significance of addressing a notable limitation within the confines of
this study. This limitation pertains to the non-inclusion of specific criminogenic variables, most notably
low self-control, within the study’s statistical models. While the primary research focus is centered on
examining the impact of social learning and social bonds on juvenile delinquent behavior among Saint
Lucian adolescents, it is imperative to acknowledge the inherent constraints associated with this omission.
This limitation, although inherent, does provide an avenue for future research to further enrich the depth
and breadth of this investigation. In forthcoming studies in this domain, the author envisages
a comprehensive approach that encompasses a broader spectrum of criminogenic determinants, including
low self-control, within the analytical models of this study. Such an endeavor will undoubtedly enhance the
analytical robustness of this research, facilitating a more nuanced understanding of the intricate interplay
among multifaceted variables contributing to juvenile delinquency among Saint Lucian adolescents.
In summary, while this study bears the limitation of not incorporating specific criminogenic
variables, the author views this as an opportunity to pave the way for future research endeavors that
are poised to contribute significantly to the scholarly discourse in the realm of criminology. This
proactive stance aligns seamlessly with the author’s unwavering commitment to advancing empirical
foundations within the field.
The results of this study showed that adolescent deviant behavior is significantly correlated with
social learning variables. Furthermore, adolescents’ participation in delinquency is also significantly
influenced by weak social bonds. Notwithstanding the noted limitations of this study, the findings
have nonetheless presented a significant step forward in our understanding of the criminogenic risk
factors of social learning and social bond variables associated with the involvement in violent and
nonviolent offending among a sample of school-aged adolescents in Saint Lucia.
Acknowledgements
I am deeply thankful to the numerous individuals and institutions who have provided unwavering support, exceptional
knowledge, and ongoing encouragement in bringing this project to fruition.
At the vanguard of this journey, I want to express my gratitude and admiration to my academic advisor, Dr. Lai Yung
Lien. His steady mentorship, perceptive observations, and enduring counsel have been vital, lighting each aspect of this
research with his vast expertise and dedication. I also want to thank Professor Dr. Chen Yu Shu; her meticulous attention
to detail and dedicated efforts significantly increased the analytical depth of this study, enriching both the results and
subsequent conversations.
My heartfelt gratitude goes to Saint Lucia's Ministry of Education. Their authorization to conduct the survey within
the schools helped this research and demonstrated their dedication to scholarly inquiry and the knowledge of juvenile
delinquency in Saint Lucia. I'd be remiss if I didn't mention the importance of the principals and teachers from the
participating schools. Their collaboration and willingness to help with this research were critical to its success.
To my cherished family, who have been a constant source of encouragement and faith in my abilities, your unfailing
support has been my anchor, primarily through difficult times. I owe you more gratitude than words can express. Finally,
regardless of the scale of their contribution to this research, I send my heartfelt gratitude to each individual. Every
comment, advice, and recommendation helped enrich this study.
DEVIANT BEHAVIOR 15
Disclosure statement
The author hereby declares no known competing financial interests or personal relationships that could have appeared
to influence the work reported in this paper.
Funding
The data utilized in this study were self-funded and conducted by the researcher.
Notes on contributor
Montelle Marius Maradona Felix is a distinguished Ph.D. candidate at the Central Police University in Taiwan, R.O.C.
He holds a Master of Science Degree in Management from the University of Maryland, with a focus on Criminal Justice
Management. Beyond academia, Mr. Felix boasts over 15 years of dedicated service as a police officer with the Royal
Saint Lucia Police Force. His scholarly pursuits primarily revolve around comparative correctional systems, inmate
misconduct, juvenile delinquency, and intricate aspects of policing. His latest research contributions can be found in the
Journal of Criminology.
ORCID
Montelle Marius Maradona Felix http://orcid.org/0000-0001-5112-786X
References
Agnew, R. 1991. “A Longitudinal Test of Social Control Theory and Delinquency.” Journal of Research in Crime and
Delinquency 28(2):126–56. doi:10.1177/0022427891028002002.
Agnich, L. E. and Y. Miyazaki. 2013. “A Cross-National Analysis of principals’ Reports of School Violence.”
International Criminal Justice Review 23(4):378–400. doi:10.1177/1057567713515273.
Akers, R. L. 1985. Deviant Behavior: A Social Learning Approach (3rd ed.). Belmont, CA: Wadsworth Pub.
Akers, R. L. 1998. Social Learning and Social Structure: A General Theory of Crime and Deviance. Boston, MA:
Northeastern University Press.
Akers, R. L. and C. S. Sellers. 2004. Criminological Theories: Introduction, Evaluation, and Application. 4
th
ed. Los
Angeles, CA: Roxbury.
Baker, J. A. 1998. “Are We Missing the Forest for the Trees? Considering the Social Context of School Violence.” Journal
of School Psychology 36(1):29–44. doi:10.1016/s0022-4405(97)00048-4.
Bjerk, D. 2007. “Measuring the Relationship Between Youth Criminal Participation and Household Economic
Resources.” Journal of Quantitative Criminology 23(1):23–39. doi:10.1007/s10940-006-9017-8.
Boers, K., J. Reinecke, D. Seddig, and L. Mariotti. 2010. “Explaining the Development of Adolescent Violent
Delinquency.” European Journal of Criminology 7(6):499–520. doi:https://doi.org/10.1177/1477370810376572.
Burt, S. A. and K. L. Klump. 2013. “Delinquent Peer Affiliation as an Etiological Moderator of Childhood Delinquency.”
Psychological Medicine 43(6):1269–78. doi:10.1017/s0033291712000013.
Byrne, B. M. 2001. “Structural Equation Modeling with AMOS, EQS, and LISREL: Comparative Approaches to Testing
for the Factorial Validity of a Measuring Instrument.” International Journal of Testing 1(1):55–86. doi:10.1207/
S15327574IJT0101_4.
Carter, R. 2008. Caribbean Youth: An Integrated Literature Review. Commission on Youth Development. Georgetown,
Guyana: CARICOM Secretariat.
Chan, H. C. 2019. “Violent Offending, Nonviolent Offending, and General Delinquency: Exploring the Criminogenic
Risk Factors of Hong Kong Male and Female Adolescents.” International Journal of Offender Therapy and
Comparative Criminology 65(9):975–98. doi:10.1177/0306624X19881917.
Chan, H. C. O. and W. H. Chui. 2015. “The Influence of Low Self-Control on Violent and Nonviolent Delinquencies:
A Study of Male Adolescents from Two Chinese Societies.” Journal of Forensic Psychiatry and Psychology 28
(5):599–619. doi:10.1080/14789949.2015.1012534.
Chan, H. C. O. and D. S. W. Wong. 2019. “Traditional School Bullying and Cyberbullying Perpetration: Examining the
Psychosocial Characteristics of Hong Kong Male and Female Adolescents.” Youth & Society 51(1):3–29. doi:10.1177/
0044118X16658053.
Chou, C. Y., Y. J. Su, H. M. Wu, and S. H. Chen. 2011. “Child Physical Abuse and the Related PTSD in Taiwan: The Role
of Chinese Cultural Background and victims’ Subjective Reactions.” Child Abuse & Neglect 35(1):58–68. doi:10.1016/
j.chiabu.2010.08.005.
16 M. M. M. FELIX
Costello, B. J. and J. H. Laub. January 132020. “Social Control Theory: The Legacy of Travis Hirschi’s Causes of
Delinquency.” Annual Review of Criminology 3(1):21–41. doi:10.1146/annurev-criminol-011419-041527.
Cunningham, N. J. 2007. “Level of Bonding to School and Perception of the School Environment by Bullies, Victims, and
Bully Victims.” The Journal of Early Adolescence 27(4):457–78. doi:10.1177/0272431607302940.
Davis, C., C. Tang, and J. Ko. 2000. “The Impact of Peer, Family, and School on Delinquency.” International Social Work
47(4):489–502. doi:10.1177/0020872804046255.
Deater, D. K. 2001. “Annotation: Recent Research Examining the Role of Peer Relationships in the Development of
Psychopathology.” Journal of Child Psychology and Psychiatry 42(5):565–79. doi:10.1111/1469-7610.00753.
Deosaran, R. 2007. “School Violence and Delinquency: The Dynamics of Race, Gender, Class, Age, and Parenting in the
Caribbean.” in Pp. 89–132 in Crime, Delinquency, and Justice: A Caribbean Reader, edited by R. Deosaran. Jamaica:
Ian Randle Publishers
Dufur, M. J., J. D. Thorpe, H. S. Barton, J. P. Hoffmann, and T. L. Parcel. 2019. “Can Social Capital Protect Adolescents
from Delinquent Behavior, Antisocial Attitudes, and Mental Health Problems?” Archives of Psychology 3(6):1–22.
doi:10.31296/aop.v3i6.119.
Durkin, K. F., T. W. Wolfe, and G. A. Clark. 2005. “College Students and Binge Drinking: An Evaluation of Social
Learning Theory.” Sociological Spectrum 25(3):255–72. doi:10.1080/027321790518681.
Elliott, D. S., D. Huizinga, and S. Ageton. 1985. Explaining Delinquency and Drug Use. Beverly Hills, CA: Sage.
El-Sheikh, A. A., M. R. Abonaze, and N. Gamil. 2017. “A Review of Software Packages for Structural Equation Modeling:
A Comparative Study.” Applied Mathematics and Physics 5(3):85–94. doi:10.12691/amp-5-3-2.
Farrington, D. P. 1986. “Age and Crime.” in Crime and Justice: An Annual Review of Research, edited by M. Tonry and
N. Morris (Vol. 7, pp. 189–250). Chicago: University of Chicago Press.
Farrington, D. P. 2003. “Developmental and Life-Course Criminology: Key Theoretical and Empirical Issues - the 2002
Sutherland Award Address.” Criminology 41(2):221–25. doi:10.1111/j.1745-9125.2003.tb00987.x.
Farrington, D. P., R. Loeber, M. Stouthamer-Loeber, W. B. Van Kammen, and L. Schmidt. 1996. “Self-Reported
Delinquency and a Combien Delinquency Seriousness Scale Based on Boys, Mothers, and Teachers: Concurrent
and Predictive Validity for African-Americans and Caucasians.” Criminology 34(4):493–517. doi:10.1111/j.1745-
9125.1996.tb01217.x.
Finkelhor, D. 2008. Childhood Victimization: Violence, Crime, and Abuse in the Lives of Young People. New York, NY:
Oxford University Press.
Foss, A., B. Laurent, D. de Garcia, H. Charles, M. Puucilowski, N. Hazel, R. Sahai, and S. McKenzie. 2013. Eastern and
Southern Caribbean Youth Assessment (ESCYA). Arlington, VA: United States Agency for International Development
(USAID).
Gentle-Genitty, C., J. Kim, E. H. Yi, D. Slater, B. Reynolds, and N. Bragg. 2017. “Comprehensive Assessment of Youth
Violence in Five Caribbean Countries: Gender and Age Differences.” Journal of Human Behavior in the Social
Environment 27(7):745–59. doi:10.1080/10911359.2016.1273811.
Gifford-Smith, M., K. A. Dodge, T. J. Dishion, and J. McCord. 2005. “Peer Influence in Children and Adolescents:
Crossing the Bridge from Developmental to Intervention Science.” Journal of Abnormal Child Psychology 33
(3):255–65. doi:10.1007/s10802-005-3563-7.
Gorman-Smith, D., H. T. Tolan, and D. B. Henry. 2000. “A Developmental-Ecological Model of the Relation of Family
Functioning to Patterns of Delinquency.” Journal of Quantitative Criminology 16(2):169–98. doi:10.1023/
A:1007564505850.
Gottfredson, D. C. 2001. Schools and Delinquency. Cambridge, UK: Cambridge University Press.
Granic, I. and G. R. Patterson. 2006. “Toward a Comprehensive Model of Antisocial Development: A Dynamic Systems
Approach.” Psychological Review 113(1):101–31. doi:10.1037/0033-295x.113.1.101.
Gutierrez, F. and D. J. Shoemaker. 2008. “Self-Reported Delinquency of High School Students in Metro Manila: Gender
and Social Class.” Youth and Society 40(1):55–85. doi:10.1177/0044118x07309986.
Han, Y. and A. Grogan-Kaylor. 2015. “Social Capital and the Onset of Health-Risk Behaviors Among Korean Youths.”
Social Work Research 39:199–211. doi:10.1093/swr/svv027.
Hektner, J. M., G. J. August, and G. M. Realmuta. 2000. “Patterns and Temporal Changes in Peer Affiliation Among
Aggressive and Non-Aggressive Children Participating in a Summer School Program.” Journal of Clinical Child
Psychology 29(4):603–14. doi:10.1207/s15374424jccp2904_12.
Hill, J. E., R. E. Emery, K. P. Harden, J. Mendle, and E. Turkheimer. 2008. “Alcohol Use in Adolescent Twins and
Affiliation with Substance Using Peers.” Journal of Abnormal Child Psychology 36(1):81–94. doi:10.1007/s10802-007-
9161-0.
Hirschi, T. 1969. Causes of Delinquency. Berkeley, CA: University of California Press.
Hoeve, M., G. J. J. M. Stams, C. E. van der Put, J. S. Dubas, P. H. van der Laan, and J. R. M. Gerris. 2012. “A
Meta-Analysis of Attachment to Parents and Delinquency.” Journal of Abnormal Child Psychology 40(5):771–85.
doi:10.1007/s10802-011-9608-1.
Hoffmann, J. P. and M. J. Dufur. 2018. “Family Social Capital, Family Social Bonds, and Juvenile Delinquency.”
American Behavioral Scientist 62(11):1525–44. doi:10.1177/0002764218787020.
DEVIANT BEHAVIOR 17
Howe, T. R. and R. D. Parke. 2001. “Friendship Quality and Socio-Metric Status: Between Group Differences and Links
to Loneliness in Severely Abused and Non-Abused Children.” Child Abuse and Neglect 25(5):585–607. doi:10.1016/
s0145-2134(01)00230-7.
James, T. A. and S. J. Solomon. September 9, 2021. “Exploring How Interactions Between Race and Social Bonds
Influence Substance Use: Findings from a Predominately Black Juvenile Sample.” Deviant Behavior 43(10):1225–44.
doi:10.1080/01639625.2021.1976605.
Jarjoura, G. R., R. A. Triplett, and G. P. Brinker. 2002. “Growing Up Poor: Examining the Link Between Persistent
Childhood Poverty and Delinquency.” Journal of Quantitative Criminology 18(2):159–87. doi:10.1023/
A:1015206715838.
Jenkins, P. H. 1997. “School Delinquency and the School Social Bond.” Journal of Research in Crime and Delinquency 34
(3):337–67. doi:10.1177/0022427897034003003.
Katz, Charles M., Hyunjung Cheon, Kayla Freemon, and Lidia E. Nuño. 2023. “Delinquency, Drug Use, and Gang
Membership in the English-Speaking Caribbean.” Children and Youth Services Review, Elsevier 144:106758. doi:10.
1016/j.childyouth.2022.106758.
Katz, C. M. and A. M. Fox. 2010. “Risk and Protective Factors Associated with Gang-Involved Youth in Trinidad and
Tobago.” Revista Panamericana de Salud Pública 27(3):187–202. doi:10.1590/s1020-49892010000300006.
Katz, C. M., E. R. Maguire, and D. Choate. 2011. “A Cross-National Comparison of Gangs in the United States and
Trinidad and Tobago.” International Criminal Justice Review 21(3):243–62. doi:10.1177/1057567711417179.
Kaufmann, D. R., P. A. Wyman, E. Forbes-Jones, and J. Barry. 2007. “Prosocial Involvement and Antisocial Peer
Affiliations as Predictors of Behaviour Problems in Urban Adolescents: Main Effects and Moderating Effects.”
Journal of Community Psychology 35(4):417–34. doi:10.1002/jcop.20156.
Kendler, K. S., K. C. Jacobson, J. M. Myers, and L. J. Eaves. 2008. “A Genetically Informative Developmental Study of the
Relationship Between Conduct Disorder and Peer Deviance in Males.” Psychological Medicine 38(7):1001–11. doi:10.
1017/s0033291707001821.
Kline, R. B. 2011. “Convergence of Structural Equation Modeling and Multilevel Modeling.”
Kock, N. and G. Lynn. 2012. “Lateral Collinearity and Misleading Results in Variance-Based SEM: An Illustration and
Recommendations.” Journal of the Association for Information Systems 13(7):546–80. doi:https://doi.org/10.17705/
1jais.00302.
Laurent, B., A. Barnes, T. Chambers, and S. Gray. 2011. “Caribbean Basin Security Initiative Juvenile Justice
Assessment.” USAID Barbados https://democracyinternational.com/media/Caribbean%20Basin%20Security%
20Initiative%20Juvenile%20Justice%20Assessment.pdf .
Lawrence, R. 2007. School Crime and Juvenile Justice. New York, NY: Oxford University Press.
Loeber, R. and D. P. Farrington. 2014. “Age-Crime Curve.” in Encyclopedia of Criminology and Criminal Justice, edited
by G. Bruinsma and D. Weisburd. New York, NY: Springer. doi:10.1007/978-1-4614-5690-2_474.
Loeber, R., D. P. Farrington, and D. Petechuk. 2003. Child Delinquency: Early Intervention and Prevention. Washington,
DC: U.S. Dept. of Justice, Office of Justice Programs, Office of Juvenile Justice and Delinquency Prevention. https://
www.ojp.gov/pdffiles1/ojjdp/186162.pdf .
Maguire, E. R. and D. H. Fishbein. 2016. “The Influence of Family Characteristics on Problem Behaviors in a Sample of
High‐Risk Caribbean Adolescents.” Family Relations 65(1):120–33. doi:10.1111/fare.12179.
Maguire, E. R., W. Wells, and C. M. Katz. 2011. “Measuring Community Risk and Protective Factors for Adolescent
Problem Behaviors: Evidence from a Developing Nation.” Journal of Research in Crime and Delinquency 48
(4):594–620. doi:10.1177/0022427810395148.
Maharaj, R. G., P. Nunes, and S. Renwick. 2009. “Health Risk Behaviours Among Adolescents in the English-Speaking
Caribbean: A Review.” Child and Adolescent Psychiatry and Mental Health 3 (1): doi: 10.1186/1753-2000-3-10.
McVie, S. 2005. “Patterns of Deviance Underlying the Age-Crime Curve: The Long-Term Evidence.” Criminology is the
Interdisciplinary Study of Crime and Deviant Behaviour 7:1–15.
Meeks-Gardner, J., H. L. Lee, P. Chevannes, C. Coore, J. Thomas, and H. B. Henningham. (2008). Violence Against
Children in the Caribbean: A Desk Review. https://www.researchgate.net/publication/300078841_Violence_against_
Children_in_the_Caribbean_A_Desk_Review .
Moestue, H., L. Moestue, and R. Muggah. 2013. “Youth Violence Prevention in Latin America and the Caribbean:
A Scoping Review of the Evidence.” Norwegian Peacebuilding Resource Centre Report. https://igarape.org.br/wp-
content/uploads/en/2013/08/Muggah-Moestue-Igarape_NOREF_Youth-violence-prevention-in-LA-and-
Caribbean_Aug-2013_FINAL.pdf .
Moffitt, T. E. 1993. “Adolescence-Limited and Life-Course-Persistent Antisocial Behavior: A Developmental
Taxonomy.” Psychological Review 100(4):674–701. doi:10.1037/0033-295x.100.4.674.
Moffitt, T. E. and A. Caspi. 2001. “Childhood Predictors Differentiate Life-Course Persistent and Adolescence-Limited
Antisocial Pathways Among Males and Females.” Development and Psychopathology 13(2):355–75. doi:10.1017/
s0954579401002097.
Moffitt, T. E., D. R. Lynam, and P. A. Silva. 1994. “Neuropsychological Tests Predicting Persistent Male Delinquency.”
Criminology 32(2):277–300. doi:10.1111/j.1745-9125.1994.tb01155.x.
18 M. M. M. FELIX
Molleman, L., S. Ciranka, and W. van den Bos. 2022. “Social Influence in Adolescence as a Double-Edged Sword.”
Proceedings of the Royal Society B: Biological Sciences 289(1977):20220045. doi:https://doi.org/10.1098/rspb.2022.0045.
Özbay, Ö. and Y. Z. Özcan. 2006. “A Test of Hirschi’s Social Bonding Theory: Juvenile Delinquency in the High Schools
of Ankara, Turkey.” International Journal of Offender Therapy and Comparative Criminology 50(6):711–26. doi:10.
1177/0306624X05283525.
Payne, A. 2008. “A Multilevel Model of the Relationships Among Communal School Disorder, Student Bonding, and
Delinquency.” Journal of Research in Crime and Delinquency 45(4):429–55. doi:10.1177/0022427808322621.
Payne, A., D. Gottfredson, and G. Gottfredson. 2003. “Schools as Communities: The Relationships Among Communal
School Organization, Student Bonding, and School Disorder.” Criminology 41(3):749–77. doi:10.1111/j.1745-9125.
2003.tb01003.x.
Payne, A. and S. Salotti. 2007. “A Comparative Analysis of Social Learning and Social Control Theories in the Prediction
of College Crime.” Deviant Behavior 28(6):553–73. doi:10.1080/01639620701357933.
Peterson, B., L. Daiwon, A. Henninger, and M. Cubellis. 2016. “Social Bonds, Juvenile Delinquency, and Korean
Adolescents.” Crime & Delinquency 62(10):1337–63. doi:10.1177/0011128714542505.
Popp, A. M. and A. A. Peguero. 2012. “Social Bonds and the Role of School-Based Victimization.” Journal of
Interpersonal Violence 27(17):3366–88. doi:10.1177/0886260512445386.
Ring, J. and R. Svensson. 2007. “Social Class and Criminality Among Young People: A Study Considering the Effects of
School Achievement as a Mediating Factor on the Basis of Swedish Register and Self-Report Data.” Journal of
Scandinavian Studies in Criminology and Crime Prevention 8(2):210–33. doi:10.1080/14043850701610451.
Rodgers-Farmer, A. Y. 2000. “Parental Monitoring and Peer Group Association: Their Influence on Adolescent
Substance Abuse.” Journal of Social Service Research 27(2):1–18. doi:10.1300/j079v27n02_01.
Salmi, V. and J. Kivivuori. 2006. “The Association Between Social Capital and Juvenile Crime: The Role of Individual and
Structural Factors.” European Journal of Criminology 3(2):123–48. doi:10.1177/1477370806061967.
Seydlitz, R. and P. Jenkins. 1998. “The Influence of Families, Friends, Schools, and Community on Delinquent
Behavior.” Pp. 53–97 in Advances in Adolescent Development: An Annual Book Series. Delinquent Violent Youth:
Theory And Interventions, edited by, T. P. Gullotta, G. R. Adams, and R. Montemayor. Sage Publications, Inc. doi:10.
4135/9781483328256.n3.
Simons, R. L., C. Wu, D. Conger, and F. O. Lorenz. 1994. “Two Routes to Delinquency: Differences Between Early and
Late Starters in the Impact of Parenting and Deviant Peers.” Criminology 32(2):247–76. doi:10.1111/j.1745-9125.1994.
tb01154.x.
Spina, S. U. 2000. “Preface.” Pp. i–xx in Smoke and Mirrors: The Hidden Context of Violence in Schools and Society, edited
by S.U. Spina. Lanham, MD: Rowman & Littlefield
Springer, A., G. Parcel, E. Baumler, and M. Ross. 2006. “Supportive Social Relationships and Adolescent Health Risk
Behavior Among Secondary School Students in El Salvador.” Social Science & Medicine 62(7):1628–40. doi:10.1016/j.
socscimed.2005.08.018.
Sprott, J. B., J. M. Jenkins, and A. N. Doob. 2005. “The Importance of School: Protecting At-Risk Youth from Early
Offending.” Youth Violence and Juvenile Justice 3(1):59–77. doi:10.1177/1541204004270943.
Staden, N. V. 2015. Psychosocial Factors Contributing to Juvenile Delinquency in the ZFM-Region. Northern Cape, South Africa.
Sullivan, C. J. 2006. “Early Adolescent Delinquency: Assessing the Role of Childhood Problems, Family Environment,
and Peer Pressure.” Youth Violence and Juvenile Justice 4(4):291–313. doi:10.1177/1541204006292656.
Sullivan, C. J. 2014. “Individual, Social, and Neighborhood Influences on the Launch of Adolescent Antisocial Behavior.”
Youth Violence and Juvenile Justice 12(2):103–20. doi:10.1177/1541204013483779.
Tabachnick, B. G, and L. S. Fidell. 1996. Using Multivariate Statistics. 3rd ed. New York: Harper Collins.
Thornberry, T. 1996. “Empirical Support for Interactional Theory: A Review of the Literature.” Pp. 198–235 in
Delinquency and Crime: Current Theories, edited by J. D. Hawkins. Cambridge, UK: Cambridge University Press
United Nations Development Programme. (2012). Caribbean Human Development Report 2012: Human Development
and Shift to Better Citizen Security. New York, NY: Author. https://infosegura.org/sites/default/files/2023-02/
Caribbean_HDR2012.pdf .
UNODC (2022). “Strategic Vision for Latin America and the Caribbean 2022-2025”. https://www.unodc.org/res/
strategy/STRATEGIC_VISION_LATIN_AMERICA_AND_THE_CARIBBEAN_2022_2025_ENE17_EDsigned.pdf .
UN Office on Drugs and Crime. (2021). International Homicides [Data Set]. https://data.worldbank.org/indicator/VC.
IHR.PSRC.P5?most_recent_value_desc=true .
Verdugo, R. 1999. “Safe Schools: Theory, Data, and Practices.” Education and Urban Society 31(3):267–76. doi:10.1177/
0013124599031003001.
Welsh, W. N., J. R. Greene, and P. H. Jenkins. 1999. “School Disorder: The Influence of Individual, Institutional, and
Community Factors.” Criminology 37(1):73–116. doi:10.1111/j.1745-9125.1999.tb00480.x.
Welsh, W. N., R. Stokes, and J. R. Greene. 2000. “A Macro-Level Model of School Disorder.” Journal of Research in Crime
and Delinquency 37(3):243–83. doi:10.1177/0022427800037003001.
WHO. (2015). Homicide Rate [Data Set]. https://apps.who.int/violence-info/homicide .
DEVIANT BEHAVIOR 19
Williams, H. M. 2016. “Lingering Colonialities as Blockades to Peace Education: School Violence in Trinidad.” Pp.
141–56 in Peace Education: International Perspectives, edited by M. Bajaj and M. Hantzopoulos. London: Bloomsbury
Academic
Wright, J. P., F. T. Cullen, and J. T. Miller. 2001. “Family Social Capital and Delinquent Involvement.” Journal of
Criminal Justice 29(1):1–9. doi:10.1016/s0047-2352(00)00071-4.
Yu, C. Y. (2002). Evaluating cutoff criteria of model fit indices for latent variable models with binary and continuous
outcomes. Unpublished doctoral dissertation. Los Angeles, CA: University of California.
Yuksek, D. A. and O. Solakoglu. 2016. “The Relative Influence of Parental Attachment, Peer Attachment, School
Attachment, and School Alienation on Delinquency Among High School Students in Turkey.” Deviant Behavior
37(7):723–47. doi:10.1080/01639625.2015.1062683.
Zuur, A. F., E. N. Ieno, and C. S. Elphick. 2010. “A Protocol for Data Exploration to Avoid Common Statistical
Problems.” Methods in Ecology & Evolution / British Ecological Society 1:3–14. doi:10.1111/j.2041-210X.2009.00001.x.
20 M. M. M. FELIX
