Macro-level Criminological Perspectives and School Delinquency: A Meta-analysis

Abstract

School delinquency in public elementary, middle, and high schools has decreased in recent years, but is still a major issue that has negative mental health and academic implications for adolescents. Although research has focused on both individual-level and school-level explanations of school delinquency, it is not yet clear which macro-level criminological perspectives best explains it. Using 656 effect sizes nested within 75 studies and 30 unique datasets, this study addresses two questions using meta-analytic methods: Which macro-level criminological perspectives explain between-school differences in delinquency? Are effect sizes invariant across samples and research design? Results indicate that only concentrated disadvantage and social cohesion are significantly related to school delinquency. With the exception of concentrated disadvantage, effects are homogenous. This suggests that some school-level explanations are useful and future research should not exclude these factors. Practical implications suggest that improving social cohesion in schools may be more effective at preventing violence than target-hardening efforts.

Keywords

school violence bullying social disorganization situational crime prevention meta-analysis

Although school delinquency in elementary, middle, and high schools has decreased between 1992 and 2018 (Wang et al., 2020), this issue is of upmost importance. Data from the 2018 National Crime Victimization Survey show that the most common locations of adolescent victimization were connected with school, and 71% of public schools reported at least one violent incident in the 2017 to 2018 school year (Diliberti et al., 2019). School delinquency has many definitions, and researchers have conceptualized it to include a wide range of non-criminal misbehavior and criminal offending that take place on school campuses or in close proximity to school (Gottfredson, 2001; Smith-Adcock et al., 2013). These include both non-violent and violent behaviors. While most violent incidents in school involve physical bullying such as pushing and shoving, school delinquency also includes serious forms of violence including assault with a weapon and even murder. Twenty percent of students reported that they had been bullied at school in 2017, with higher estimates for females and LGBTQ+ students (Seldin & Yanez, 2019). Experiencing victimization at school may lead to long term implications, such as decreased academic achievement (Nakamoto & Schwartz, 2010) and increased internalizing problems, externalizing problems, and substance use (Ttofi et al., 2014). Those who commit acts of delinquency at school are at risk for dropping out of school (Peguero & Hong, 2019) and further involvement in crime (Hemphill et al., 2014).

Although rare, there were 66 school shootings in the 2018 to 2019 school year in the US (Wang et al., 2020). Widespread media coverage of these tragic events decreased perceptions of safety at school (Elsass et al., 2016), and policymakers were pressured to immediately address safety concerns. Funded by a $745 million federal grant, a situational crime prevention approach was quickly adopted in many schools, which focused on reducing problem behaviors by increasing surveillance and reducing opportunities for delinquency (Clarke, 1983). Methods included employing school resource officers (SROs), installing cameras and metal detectors, and using drug dogs to conduct locker searches (Jonson, 2017). However, public pressure for immediate attention to school safety led to situational crime prevention efforts to be enacted without empirical evidence to suggest that they were effective (Jonson, 2017). Additionally, they tend to focus on preventing rare forms of serious violence rather than more common forms of school delinquency (Gerlinger & Wo, 2016). While the climate of schools was also called into question, efforts to improve student-teacher relationships, authoritative school discipline, and connectedness were not enacted as readily, despite evidence that improving school climate decreased delinquency (Gerlinger & Wo, 2016).

Individual-level theories of crime have successfully explained a wide range of delinquent behaviors in school including violent crime, property crime, bullying, and general misbehavior (Chui & Chan, 2015; Popp & Peguero, 2012). However, school delinquency prevention efforts based on individual-level theories of crime may be difficult to enact, since the school may not be able to intervene with personal and family risk factors outside the school context (Rowan et al., 1983). School-level interventions are more malleable (Gottfredson et al., 2005), and might reach more students who have been placed at high-risk. Prior research suggests that using them in conjunction with individual-level interventions is more effective than individual-level interventions on their own (Tillyer et al., 2018). While a body of research has found that a range of school-level predictors can effectively reduce school delinquency, these concepts may be too broad. For example, the most effective school climate component in delinquency prevention has been found to be school connectedness, which is equivalent to a component of social disorganization theory’s concept of collective efficacy (Aldridge et al., 2018). Simply put, delinquency in schools might be better explained by theories designed to explain delinquency, rather than broader theories to might explain a variety of school outcomes not related to delinquency.

While individual-level predictors derived from criminological theory and broad categories of school-level variables have been found to explain delinquency, a much smaller volume of research exists that uses predictors derived from macro-level criminological theory. While the research that has been completed suggests that these explanations may be useful, it is not clear which predictors are the strongest, and estimates of the proportion of variance explained by macro-level factors vary. To shed some light on these issues, this study will use hierarchal meta-analytic methods to determine which macro-level criminological perspectives best explain delinquency in public elementary, middle, and high schools. Additionally, moderator analyses will be performed to determine whether effects are homogenous according to research design and sample characteristics.

Literature Review

While criminologists tend to focus on individual-level factors when explaining crime, there has been a renewed interest in using macro-level criminological perspectives to explain differences in crime rates between geographical areas such as neighborhoods, street segments, cities, and countries (Pratt & Cullen, 2005). Similarly, scholars studying delinquency in schools generally look at individual differences and close relationships, rather than school-level factors. While structural factors in neighborhoods and those in schools are not identical, scholars have successfully applied macro-level criminological perspectives to explain school delinquency (Plank et al., 2009). Research using multi-level models that include both individual and school-level factors have not consistently found similar variance accounted for at each level. Estimates range from 5.3% (Olsson et al., 2017) to 84% (Khoury-Kassabri et al., 2004) of variance attributed to school-level factors, depending on what variables are included.

Social disorganization theory posits that structural characteristics can explain differences in crime rates (Shaw & McKay, 1942), and has also been applied to schools. Schools that are more disorganized are more likely to have higher rates of delinquency due to a breakdown of informal social control. Prior research has found that schools with higher concentrations of poverty have higher rates of bullying victimization (Bradshaw et al., 2009) and violence (Adams & Mrug, 2019), however, this has been found to vary depending on school type (Bradshaw et al., 2009). Schools that are high in student mobility may make it difficult for students to form connections with others, which may inhibit the formation of social control (Wilson, 2004). However, higher mobility schools have also been found to have less bullying (Bradshaw et al., 2009). Schools that are ethnically heterogeneous are theorized to be more disorganized, as multiple social norms and cultures that co-exist may make bonding difficult, and students may be less likely to serve as a guardian to someone of a different race or culture than their own (Schreck et al., 2003). However, research is mixed, and relationships between heterogeneity and peer victimization has found to be dependent on immigrant status (Agirdag et al., 2011).

Along with the structural elements of social disorganization theory, collective efficacy has also been applied to schools. Cohesion and shared expectations of social norms among students invoke social control (Sampson et al., 1997), and students collectively intervene for the common good (i.e., enforcing school rules and preventing delinquency). Tests of collective efficacy in the school context have shown that social cohesion and normative beliefs have negative relationships with bullying (Olsson et al., 2017; Sapouna, 2010), and serious victimization (Zaykowski & Gunter, 2012). However, not all research has found direct effects of collective efficacy on outcomes. Some research has found no effect of social cohesion on misconduct when individual factors were controlled for (Demanet & Van Houtte, 2012; Tillyer et al., 2011), and some research has found mediating and moderating effects only. For example, the consensus of normative beliefs was found to mediate the relationship between student-teacher ratio and victimization (Gottfredson & DiPietro, 2011), and school efficacy was found to moderate the relationship between self-control and victimization (Tillyer et al., 2018).

Consistent with the broken windows perspective (Wilson & Kelling, 1982), signs of physical disorder in schools, such as graffiti, litter, and broken items send a message that “no one cares,” and decreases the likelihood that students will collectively influence social control, creating more opportunities for delinquency. Prior research has found that physical disorder in schools influenced students’ perceptions of delinquency in schools, which decreased social control and increased deviance (Johnson et al., 2017; Plank et al., 2009). However, inverse relationships have also been found, with schools with more vandalism having less instances of bullying. One explanation for this could be that older schools, which may be more likely to be vandalized, are more connected (Bradshaw et al., 2015).

Routine activity theory posits that a criminal event will happen when a motivated offender, a suitable target, and lack of guardianship converge in time and space (Cohen & Felson, 1979). Previous research on school delinquency has focused on the guardianship element and has measured this by the ratio of students to teachers. In schools with a high student-teacher ratio, it may be difficult to monitor the behavior of students, therefore increasing the opportunities that students have to commit acts of delinquency (Bradshaw et al., 2009). However, not all previous studies have found an effect of student-teacher ratio on bullying (Wei et al., 2010), suggesting that teachers may not be capable guardians.

Another form of guardianship may be provided by surveillance measures. Formal surveillance, including installing security cameras, employing school resource officers (SROs), and security guards shows mixed effectiveness. Some studies found that surveillance measures reduced delinquency in schools, as students might be deterred (Jennings et al., 2011; Theriot, 2009). However, others found that schools that employed SROs had higher incidents of delinquency (Fisher & Hennessy, 2016; Lesneskie & Block, 2017). This may be due to surveillance measures in schools contributing to a negative school environment and increasing feelings of fear at school, which erodes social control and increases delinquency (Fisher et al., 2018). Additionally, problem behaviors may be more likely to be detected in schools that use formal security, and these problems may be criminalized by SROs instead of being dealt with by school administration (Hirschfield, 2008).

Related to formal surveillance, Situational crime prevention (SCP) focuses on increasing risks and reducing opportunities for delinquency by manipulating the environment. Categories of SCP include access control (i.e., locking buildings, visitor sign in), controlling contraband (i.e., drug dogs, locker checks), and reducing anonymity (i.e., ID badges). Previous tests of SCP have found mixed effectiveness in preventing delinquency in schools. Some research has found that schools with more security measures have significantly less student delinquency (Gerlinger & Wo, 2016; Servoss, 2017). A larger volume of research exists that finds security measures have null effects in preventing violence (Sevigny & Zhang, 2018) and bullying (Blosnich & Bossarte, 2011), and may increase property crime (Tanner-Smith et al., 2018) and violence (Cuellar, 2018). Again, these target-hardening approaches may make the school feel institutionalized, promoting a sense of fear rather than safety. Because most research on security measures in schools is cross-sectional, it has been difficult for past research to disentangle its’ effect on school delinquency. It could be that schools with more delinquency are the ones where SCP methods are enacted (Fisher et al., 2018).

Current Focus

Schools have been quick to devote large amounts of resources toward preventing school delinquency, which is commendable. However, it is still not known exactly what the role of school-level variables are in explaining delinquency. Previous examinations of the effects of school factors on school delinquency are mixed, and it is not clear whether macro-level effects exist once individual-level explanations are controlled for. In order to get a better approximation of these mixed findings, this study will use hierarchal meta-analytic methods to estimate mean effect sizes of school level variables derived from macro-level criminological perspectives and their effects on school delinquency. These effects have been found to be heterogeneous, and vary depending school and student characteristics such as grade level (Bradshaw et al., 2009) and demographic composition of the school (Theriot, 2009). Based on gaps in previous research, several research questions were developed: (1) Can macro-level criminological perspectives explain school delinquency? (2) Which of these perspectives have the largest effect sizes? (3) Do these effect sizes vary according to research design and sample characteristics? Ultimately, better understanding how school factors influence delinquency will help school administrators enact policies to prevent it and ensure resources are used efficiently.

Data and Methods

Criteria for Inclusion

Several criteria were developed to search for studies suitable for inclusion. Both published and unpublished material, which includes theses and dissertations, were included. Studies from a variety of countries from a time range of 2001 to 2019 were included. Studies were limited to the English language. Both cross-sectional and longitudinal studies that used multivariate quantitative methods were included. Because the focus was on between-school differences, only studies that used the school (elementary, middle, high, or mixed) as the unit of analysis were included. The majority of studies used a hierarchal design and included both within-school and between-school differences. In these studies, only the level two effect sizes (between-school differences) were used. Some studies used non-hierarchal regression to look at between-school differences, and these studies were included as well. Studies had to have a predictor variable that was consistent with one of the following criminological perspectives: social disorganization theory, situational crime prevention, routine activity theory, and broken windows. Studies were included in they contained either a delinquency perpetration or victimization outcome.

Search Methods

The search for relevant materials took place between May 2020 and July 2020. The following electronic databases were searched: Criminal Justice Abstracts, Education Full Text, ERIC, Family Studies Abstracts, NCJRS, OpenDissertations, PsychArticles, PsychINFO, ProQuest, SocINDEX, and Sociological Abstracts. These databases were searched using the Boolean search terms “school” AND “bullying OR victimization OR crime OR violence OR theft OR delinquency OR aggression OR problem behavior OR shooting” AND “social disorganization OR collective efficacy OR school efficacy OR cohesion OR residential mobility OR poverty OR ethnic heterogeneity OR SRO OR school resource officer OR security OR disorder OR surveillance OR camera OR target hardening OR situational crime prevention OR metal detector.” These keywords were restricted to the abstract. Material was limited to journal articles, theses, dissertations, conference papers, and technical reports. The next step involved examining previous meta-analyses and review articles on school-level variables and delinquency (Azeredo et al., 2015; Fisher & Hennessy, 2016; Gonzalez et al., 2016; Johnson, 2009; Jonson, 2017; Saarento et al., 2015; Steffgen et al., 2013). In total, 2,554 items were identified. Out of these, 2,308 were eliminated once the title and/or abstract was read because it was clear that the study did not fit the search criteria. Two hundred forty-six full articles were accessed. Out of these, 68 were eliminated because they only examined between-individual differences, rather than between-school differences. Fifty-six were eliminated because they did not have a predictor variable that was associated with the criminological perspectives discussed. Thirty-three were eliminated because they had insufficient methods to be included in the meta-analysis, including having only bivariate statistics, not having the information to compute the standard error, or unstandardized coefficients that did not have the required information to transform into standardized coefficients. Fourteen studies were eliminated because the dependent variable did not measure some form of delinquency in an elementary, middle, or high school (i.e., fear of crime, perceptions of problems in school). In total, 75 studies were retained for coding from 30 unique datasets. Included studies are presented in Figure 1 and Table 1.

Figure 1.

PRISMA flowchart of included studies.

Table 1.

Included Studies.

Authors	n Schools	School type	Location	Published
Agirdag et al. (2011)	68	Elementary	Belgium	Yes
Agnich and Miyazaki (2013)	4,885	Middle	Global	Yes
Augustine (2003) ⁵	60	Mixed	USA	No
Bates (2008) ⁴	992	Mixed	USA	No
Bauer (2008) ¹	2,724	Mixed	USA	No
Bradshaw et al. (2009) ²	95	Elem, Middle	USA	Yes
Bradshaw et al. (2015) ²	52	High	USA	Yes
Boswell (2016)	26	High	USA	No
Brezina et al. (2001)	87	High	USA	Yes
Broll and Lafferty1 (2018)	735	Mixed	USA	Yes
Byrd et al.3 (2015)	580	High	USA	Yes
Chen (2008)	712	High	USA	Yes
Chen and Jacobson (2013)	16	Middle	USA	Yes
Coffey (2017)	2,560	Mixed	USA	No
Cornell and Huang (2016)	323	High	USA	Yes
Cuellar (2018)	936	High	USA	Yes
Demanet and Van Houtte (2011) ⁸	85	High	Belgium	Yes
Demanet and Van Houtte (2012) ⁸	85	High	Belgium	Yes
Demanet and Van Houtte (2019)	111	High	Europe	Yes
Devlin and Gottfredson (2018) ¹	480	Mixed	USA	Yes
Dhungana (2012)	681	Mixed	USA	No
DiPietro et al. (2015)	30	Mixed	USA	Yes
Fink et al. (2018)	648	Elementary	England	Yes
Fisher et al. (2018) ³	531	High	USA	Yes
Fissel et al. (2019) ⁵	137	Middle	USA	Yes
Gibbs (2002)	13	Elementary	USA	No
Gottfredson and DiPietro (2011) ⁶	253	High	USA	Yes
Gregory et al. (2010) ⁷	290	High	USA	Yes
Hoffmann and Dufur (2008) ⁴	883	High	USA	Yes
Jennings et al. (2011) ¹	932	High	USA	Yes
Johnson et al. (2019) ⁵	107	Mixed	USA	Yes
Jones (2003) ⁵	67	Mixed	USA	No
Klein and Cornell (2010)7	290	High	USA	Yes
Koo et al. (2012) ³	580	High	USA	Yes
Kurlychek (2004) ⁴	745	Mixed	USA	No
Latimore et al. (2018) ³	580	High	USA	Yes
Larochette et al. (2010)	116	Mixed	Canada	Yes
Lee et al. (2014) ²	58	High	USA	Yes
Lehman (2014) ³	575	High	USA	Yes
Lesneskie and Block (2017) ¹	1,942	Mixed	USA	Yes
Lleras (2008) ⁴	659	High	USA	Yes
Martinez et al. (2016)	13	Mixed	USA	Yes
Maskaly et al. (2011) ¹	1,853	Mixed	USA	Yes
Maume et al. (2010) ¹	1,936	Mixed	USA	Yes
Na and Gottfredson (2013) ¹	470	Mixed	USA	Yes
Nickerson and Martens (2008) ¹	440	Mixed	USA	Yes
Olmstead (2005) ¹	2,270	Mixed	USA	No
O’Neill and McGloin (2007) ¹	1,678	Mixed	USA	Yes
Parker et al. (2017) ²	58	High	USA	Yes
Payne (2008) ⁶	253	High	USA	Yes
Payne (2009) ⁶	253	High	USA	Yes
Payne (2012) ⁶	253	High	USA	Yes
Peguero (2011)3	580	High	USA	Yes
Peguero and Jiang (2014) ³	580	High	USA	Yes
Peguero and Jiang (2016) ³	580	High	USA	Yes
Pigott et al. (2018) ¹	950	Mixed	USA	Yes
Popp and Peguero (2011) ³	580	High	USA	Yes
Rezapour et al. (2019)	42	Mixed	Iran	Yes
Richard et al. (2012)	478	Middle	France	Yes
Sevigny and Zhang (2018) ¹	996	High	USA	Yes
Stickle (2009) ⁹	130	Mixed	USA	No
Stretesky and Hogan (2005)	371	High	USA	Yes
Swartz et al. (2017) ⁵	103	Mixed	USA	Yes
Tanner-Smith et al. (2018) ¹	10,340	Mixed	USA	Yes
Theriot (2009)	28	Mixed	USA	Yes
Tillyer et al. (2010) ⁵	107	Mixed	USA	Yes
Tillyer et al. (2011) ⁵	58	Mixed	USA	Yes
Tillyer et al. (2018) ⁵	65	Mixed	USA	Yes
Vogel and Barton (2013) ⁹	75	High	USA	Yes
Wang (2013)	1,524	Mixed	USA	No
Way (2003) ⁴	877	High	USA	No
Wei et al. (2010)	12	Middle	Taiwan	Yes
Wilcox et al. (2006) ⁵	65	Middle	USA	Yes
Wilcox and Clayton (2001)	21	Mixed	USA	Yes
Zhao (2018)	302	High	China	Yes

School Survey on Crime and Safety.

Maryland Safe and Supportive Schools.

Education Longitudinal Study 2002.

National Education Longitudinal Study 1988.

Rural Substance Abuse and Violence Project.

National Study of Delinquency Prevention in Schools.

Virginia High Schools.

Flemish Educational Assessment.

ADD Health.

Predictor Variables

The predictor variables explain between-school differences of delinquency and are separated into several domains that represent macro-level criminological perspectives.

Social disorganization theory

While originally used to explain crime at the neighborhood level, elements of social disorganization have been applied to schools as well. The measure of social cohesion comes from the collective efficacy in neighborhoods literature (Sampson et al., 1997). It is measured as a continuous variable and captures whether students have common goals and values, a sense of belonging to their school, and help one another, with higher scores indicating more social cohesion. Ethnic heterogeneity includes effect sizes that use several similar diversity indices that calculate the proportion of different ethnicities represented in the student population. It is expressed as a continuous variable with higher values indicating greater ethnic heterogeneity in the school. Studies that used the percentage of minority students as a measure of ethnic heterogeneity were not included, as this is not a true measure of heterogeneity. Student mobility was measured as the proportion of students that transfer in and out of the school during the school year, with higher values indicating more student mobility. Concentrated disadvantage was measured using the percentage of students that were eligible for a free or reduced-price lunch.

Situational crime prevention

Safety features was measured as the number of features used in a school that were consistent with situational crime prevention. Several categories of situational crime prevention were included, such as controlling access (visitor sign in, locked doors), controlling contraband (drug dogs, locker checks, metal detectors, clear bookbags), and reducing anonymity (ID badges). SRO measured whether the school employed a school resource officer, and formal surveillance measured whether the school used visible security guards and/or cameras. Both of these variables were measured dichotomously.

Broken windows

Consistent with the broken windows perspective, physical disorder was included. This variable was measured using Likert scale responses that indicated the amount of physical disorder in the school. Items included broken windows, vandalism, graffiti, broken desks, broken lights, etc., as well as signs of ownership (murals). Signs of ownership were reverse-coded. Higher scores indicated a higher level of physical disorder in the school.

Routine activity theory

While larger schools generally have more delinquency, the ratio of students to teachers has been found to be a more accurate measure of opportunity to offend than school size (Khoury-Kassabri et al., 2004). In schools with higher student/teacher ratios, it may be more difficult to monitor students, decreasing guardianship and creating opportunity for delinquency. Student-teacher ratio was included as a measure of the guardianship element of routine activity theory.

Moderating Variables

While the term school delinquency encompasses a wide range of misbehavior, it is likely that different forms of delinquency are not all best explained by the same predictor variables. In order to control for this, effect sizes were coded by outcome type: general delinquency included a mixed measure of criminal delinquency (i.e., studies that did not separate violence and theft) and misconduct that may not be criminal (substance use at school, incorrigibility). Violence included measures of violent acts and physical bullying. Theft included property crimes. Bullying included any bullying that was not physical in nature, and mixed measures of bullying. Each of the four outcomes types was coded for whether it was perpetration or victimization, therefore eight outcomes were included. While the vast majority of studies specified that the outcome had to take place on school property, several studies did not specify that outcomes occurred only at school. This was included as a moderator. Studies that explicitly stated that the outcome took place outside of school were eliminated from analysis. Some studies used self-reports from the students in the school, and some studies used reports by administrators. This was included as a moderator. All samples used in the studies were school samples, and studies were coded for whether they took place in elementary, middle, high, or a mix of more than one type of school. Additionally, studies were coded for whether they took place in North America (USA and Canada) or outside of North America. Studies from Belgium, Iran, England, and France, among others, were included, but there were not enough international studies to code each country separately. Studies were coded for whether they used a cross-sectional or longitudinal design, as well as whether they used multi-level models or not. Multi-level models control for the influences of between-individual differences as well as between-school differences, while the studies that only use the school as the unit of analysis do not. This was controlled for to determine if including the between-individual differences affects the mean effect size.

Multivariate models can vary widely depending on what controls are included. Several of the most salient predictors of school delinquency were included as controls. Each effect size was coded for whether or not it included each of several different controls: School size, since larger schools have been found to have more delinquency than small schools (Nickerson & Martens, 2008); Percent minority, as schools with higher proportions of minority students have been found to have more delinquency (Payne et al., 2003); Poverty in the neighborhood where the school is located, as delinquency has been found to be higher in impoverished neighborhoods (Bradshaw et al., 2009); Urban, suburban, or rural location, as urban schools have been found to have more delinquency (Payne, 2009); Neighborhood crime rate, as schools in criminogenic neighborhoods tend to have more delinquency (Jennings et al., 2011); and percent male students, as males are generally more involved in delinquency at school than girls (Servoss, 2017). Additionally, effect sizes were coded for whether or not they had a predictor from a competing perspective in the same model. For example, if an effect size for SRO also contained an effect size for student/teacher ratio, it was coded as “yes.”

Plan of Analysis

This meta-analysis used Fisher’s z scores as a measure of mean effect size. Fisher’s z scores are a common metric used in meta-analyses, as they provide a correction for the assumption that samples are normally distributed and sampling variances are known (Borenstein et al., 2011). Multivariate effect sizes, including odds ratios and standardized regression coefficients were converted into r correlation coefficients, and then Fisher’s z scores using established formulas (Peterson & Brown, 2005; Pratt et al., 2014). Studies that contained unstandardized coefficients were standardized using the formula provided by Rosenthal (1994). Standard errors were required to compute mean effect sizes, and if not provided, were calculated using p values and confidence intervals.

An important assumption of meta-analyses is that effect sizes are independent of each other. However, studies that use the same dataset and studies that contain multiple models violate this assumption. A solution to this is to aggregate common effect sizes so that one effect size per study is used. However, this decreases statistical power and masks important moderating differences. To address these issues, a hierarchal meta-analysis design was used in this study. This method assesses the variance at three levels: sampling variance of individual effect sizes (level 1), variance within datasets (level 2), and variance between datasets (level 3). Datasets, rather than studies, were used to assess variance since a number of studies in the meta-analysis used the same data. The School Survey on Crime and Safety (SSOCS) surveys a different sample of schools every 2 years. However, overlap of schools from one survey year to the next is around 50%, so about half of schools are surveyed in multiple years. Additionally, many of the studies used multiple years in their analyses (Fisher et al., 2019). Because of this, the decision was made to count every year of the SSOSC study as one dataset to ensure that some schools are not counted more than once. Since there are a large number of studies using the SSOSC data that are eligible for inclusion, this is favorable to eliminating SSOSC studies from the analysis completely.

Analyses were performed using the metafor package in RStudio (Assink & Wibbelink, 2016; Viechtbauer, 2010). First, mean effect sizes for each predictor were estimated and heterogeneity was assessed. Next, for each predictor variable that showed heterogeneity, variance was assessed at each level using likelihood ratio tests. Next, if sufficient variance at level 2 or 3 was found, moderating analyses were performed. Finally, publication bias was assessed by estimating trim-and-fill funnel plots for each predictor variable.

Results

Descriptive Statistics

In total, 656 effect sizes were included in the analysis. The descriptive statistics of the studies are presented in Table 2. Because most datasets included multiple studies, many of the descriptive statistics exceed 100%.

Table 2.

Descriptive Statistics.

Moderator variable	% Datasets
Measurement of DV
General delinquency	50.0
General victimization	13.3
Violent delinquency	40.0
Violent victimization	16.7
Theft delinquency	10.0
Theft victimization	3.3
Bullying perpetration	43.3
Bullying victimization	23.3
Location of DV
At school	66.7
Mixed	33.3
DV reported by
Administration/faculty	20.0
Students	83.3
School type
Elementary	13.3
Middle	20.0
High	43.3
Mixed	30.0
Study design
Cross-sectional	80.0
Longitudinal	20.0
Location
North America	70.0
Non-North America	30.0
Analysis type
HLM	83.3
Non-HLM	16.7
Publication status
Published	86.7
Unpublished	26.7
Control for school size
Yes	70.0
No	40.0
Control for % minority
Yes	73.3
No	43.3
Control for school location
Yes	30.0
No	80.0
Control for % male
Yes	66.7
No	53.3
Control for low SES
Yes	76.7
No	30.0
Control for competing theories
Yes	36.7
No	70.0

Note. Because studies in a dataset may have different moderators, some descriptive statistics add up to more than 100%.

Main Effects

Mean effect sizes and the percentage of variance at each level is presented in Table 3. Only two predictor variables had a significant relationship with school delinquency. Higher student/teacher ratios (z = 0.034, p < .001) and increased concentrated disadvantage (z = 0.044, p = .027) were associated with increased delinquency.

Table 3.

Mean Effect Sizes and Moderators.

Predictor	#Dataset	#E.S.	Mean z	95% C.I.	%Var. Level 1	Level 2 Var.	% VarLevel 2	Level 3 Var.	% Var. Level 3	F (df1, df2)
SRO	3	30	0.021	−0.064, 0.106	100	0.000	0	0.000	0	—
Safety features	3	71	0.005	−0.026, 0.035	100	0.000	0	0.000	0	—
Surveillance	3	64	0.013	−0.032, 0.059	100	0.000	0	0.000	0	—
Physical disorder	4	53	0.051	−0.009, 0.111	100	0.000	0	0.000	0	—
Student/teacher ratio	11	81	0.034***	0.015, 0.054	100	0.000	0	0.000	0	—
Social cohesion	6	81	−0.025	−0.056, 0.007	100	0.000	0	0.000	0	—
Ethnic heterogeneity	7	30	0.052	−0.027, 0.130	100	0.000	0	0.000	0	—
Student mobility	5	31	0.027	−0.017, 0.072	86.43	0.000	0	0.001	13.57	—
Concentrated disadvantage	25	215	0.045*	0.010, 0.080	48.62	0.000	2.23	0.004***	49.15	—
Moderators-concentrated disadvantage
Outcome	—	—	—	—	—	—	—	—	—	F (7, 207) = 3.629***
School type	—	—	—	—	—	—	—	—	—	F (3, 211) = 0.144
Reported by	—	—	—	—	—	—	—	—	—	F (1, 213) = 0.652
DV measure	—	—	—	—	—	—	—	—	—	F (1, 213) = 0.002
Study design	—	—	—	—	—	—	—	—	—	F (1, 213) = 0.362
Size	—	—	—	—	—	—	—	—	—	F (1, 213) = 0.139
Percent minority	—	—	—	—	—	—	—	—	—	F (1, 213) = 0.015
School location	—	—	—	—	—	—	—	—	—	F (1, 213) = 0.800
Neighborhood crime	—	—	—	—	—	—	—	—	—	F (1, 213) = 3.157
Percent male	—	—	—	—	—	—	—	—	—	F (1, 213) = 5.877*
Competing theories	—	—	—	—	—	—	—	—	—	F (1, 213) = 0.143
HLM model	—	—	—	—	—	—	—	—	—	F (1, 213) = 0.446
Country	—	—	—	—	—	—	—	—	—	F (1, 213) = 2.627

Note. #E.S. = number of effect sizes; z = Fisher’s z; F = f-statistic; df = degrees of freedom.

p < .05. **p < .01. ***p < .001.

Moderating Effects

The concentrated disadvantage predictor variable showed significant variance at level 3 (between studies) so moderating analyses were completed. Two moderating variables showed significant variation. First, relationships between concentrated disadvantage and school delinquency depended on the outcome used. This relationship was only significant for studies that used an outcome of general delinquency (F[7, 207] = 3.629, p < .001). Additionally, the relationship between concentrated disadvantage and school problem behaviors was only significant when studies controlled for the percentage of male students (F[1, 213] = 5.877, p = .016). Moderating analyses for concentrated disadvantage are presented in Table 3.

Publication Bias

A concern when conducting meta-analyses is the “file cabinet” effect, where studies that have null findings are less likely to be submitted for publication. This may artificially inflate the mean effect sizes in a meta-analysis (Rosenthal, 1979). To test for this bias, trim and fill funnel plots, which impute datasets that may be missing from the analysis, were estimated after completing the main analyses. The funnel plots for concentrated disadvantage, physical disorder, formal surveillance, situational crime prevention, SRO, and ethnic heterogeneity revealed imputed datasets on the left side, indicating that studies using these predictor variables may be positively biased. Social cohesion, student teacher ratio, and residential mobility did not show evidence of publication bias. However, because there are a small number of included datasets, these funnel plots should be interpreted with caution (Nakagawa & Santos, 2012). Because of space restrictions, funnel plots are available from the author upon request.

Discussion

This study used hierarchal meta-analytic methods to examine the effects of macro-level criminological perspectives on school delinquency. While each of the relationships between the predictor variables and outcomes were in the expected direction, only concentrated disadvantage and student/teacher ratio were significant. When the heterogeneity of effects was examined, only concentrated disadvantage showed heterogeneity, which suggested that moderator analyses be conducted. For moderating analyses, only studies that used an outcome of general delinquency and those that controlled for percentage of male students were significant, indicating that the effect of concentrated disadvantage differed depending on these two moderators. Implications for theory, future research, and practice, as well as limitations, are discussed.

Implications for Theory

The limited number of macro-level criminological perspectives that predicted school delinquency suggests that individual-level explanations may be more influential. This is consistent with other research that has found that once individual factors are controlled for, school-level factors become insignificant in explaining delinquency (Azeredo et al., 2015; Servoss, 2017). Criminological perspectives that look at individual differences, such as social bonds, low self-control, and general strain have had success in explaining delinquency in schools (Moon & Alarid, 2015; Patchin & Hinduja, 2011; Popp & Peguero, 2012). However, in this meta-analysis, only one element of social disorganization theory (concentrated disadvantage) and one element of routine activity theory (lack of capable guardianship) significantly predicted between-school differences. Scholars have had recent success integrating individual-level and macro-level criminological theories when explaining school delinquency. Some examples of this include routine activity theory with social bonds (Cecen-Celik & Keith, 2019), and situational crime prevention with social bonds (Fisher et al., 2019). Findings from this meta-analysis suggest that individual-level criminological perspectives should be integrated with macro-level perspectives in future research on school delinquency.

All in all, when comparing target hardening efforts and formal surveillance to teachers as capable guardians, teachers appear to be more effective. Mean effect sizes for situational crime prevention strategies, surveillance, and SROs did not reduce delinquency in schools. In fact, the effect sizes for these predictors had a positive relationship with delinquency, although they did not reach significance. This is not surprising, as employing these techniques in schools has been found to create a climate that is more like a prison than a school, effectively decreasing school connectedness and feelings of safety (Theriot, 2009). While there is some concern that this lack of effectiveness is simply due to the fact that delinquency is more likely to be detected in schools that use SCP and formal surveillance (Devlin & Gottfredson, 2018), this was not likely the case in this meta-analysis. Studies were coded for whether administrators or students reported the dependent variable, and if increased delinquency was an artifact of increased detection and reporting, the studies that used administrative data would have had larger effect sizes than those that used student self-reports. However, this moderating effect was not present.

Implications for Research

While previous research has found that relationships between school-level factors and delinquency differed according to demographic composition and research design (Bradshaw et al., 2009), in this meta-analysis, there was limited heterogeneity of effect sizes both within and between datasets. While this could be attributed, again, to individual-level differences being more important in explaining school delinquency, recent meta-analyses that examined risk factors for juvenile offending and victimization at the individual-level found limited heterogeneity as well (Moule & Fox, 2021; Myers et al., 2020). However, two moderators were significant for the concentrated disadvantage predictor variable. Concentrated disadvantage was significantly associated with general delinquency only, and not victimization, bullying, or violence. Additionally, the relationship was significant only when the percentage of males in the sample was controlled for. While this moderator examined if gender composition is controlled for, rather than different effects of concentrated disadvantage on gendered violence in school, previous research has found that girls were more likely to be victimized in schools with greater amounts of school-level poverty (Dhami et al., 2005). Future research should integrate school-level factors with individual-level factors to examine potential gender differences in delinquency.

In meta-analytic research, findings are only as good as the studies used. Measurement of school delinquency has several limitations. For one, student self-reports may be inaccurate, and data validity screening is not often used in large surveys that examine youth behaviors. Respondents may artificially inflate involvement in delinquency as a way to be socially desirable to their delinquent peers. Recently, surveys such as the California Healthy Kids Survey and the Arizona Youth Survey have begun to use validity assessment procedures, including asking respondents if they have used a fictitious drug to eliminate respondents who do not respond honestly. Additionally, wording and language used in surveys, recall periods, and a focus on severe violence instead of more common forms of delinquency may lead to very different pictures of the extent of violence in schools (Sharkey et al., 2011). Future research should use instruments that are valid and reliable to measure problem behaviors in schools.

Implications for Practice

The finding that situational crime prevention methods did not prevent school delinquency brings the costs and benefits of employing SROs and installing target-hardening devices into question (Fisher & Hennessy, 2016). Since the tragic events at Columbine High School, federal grant money has made over 1 billion dollars available to hire SROs, and $123 million available to purchase metal detectors and cameras for schools. The Stop School Violence Act, which will provide $100 million annually up to the year 2028, was enacted as a response to the Parkland school shooting in 2018 (Wilson, 2020). The finding that high student/teacher ratios are associated with an increase in delinquency suggests that teachers may be better monitors than target-hardening devices and SROs and may help reduce opportunities for delinquency in schools. Schools may be better off by reinvesting crime prevention resources into hiring more teachers in order to reduce the student/teacher ratio in schools. Along with increased monitoring, this may provide academic benefits as well (Whitehurst & Chingos, 2011).

While the protective effect of school cohesion on school delinquency only approached significance (p < .10) in this meta-analysis, it is not to say that fostering efficacy in schools is not important. Previous research has found that programs that enhance school cohesion by improving school climate may prevent delinquency. For example, the School Positive Behavioral Interventions and Support provides training to staff that aims to improve relationships between adults and students on campus. This program has been found to reduce suspensions and problem behaviors (Bradshaw et al., 2010). Additionally, restorative justice practices implemented in schools may also improve school climate by focusing on inclusion and communication, rather than punishment, when problems arise (González, 2012).

Along with addressing school climate, the small and insignificant effect sizes for school-level explanations suggest that individual risk factors must also be addressed. While many of these risk factors exist in the family and home life domains, the school domain can also be targeted at the individual level. For example, a recent meta-analysis found that attachment to school significantly protected boys from victimization (Pusch & Holtfreter, 2021), so focusing on fostering individual attachments to school may be beneficial. One way both school-level and individual-level risk factors can be addressed simultaneously is by using a multi-level programming response (Tillyer et al., 2011). This method uses school-contextual programming plus practices that target at-risk individuals for a whole school comprehensive approach. Especially useful at the individual level are programs that use cognitive behavior therapy that enhances self-control and teaches students to regulate emotions (Gottfredson, 2001). One such multi-level program is the Olweus anti-bullying program, which targets bullying risk at the individual, classroom, and school level (Olweus et al., 1999). Evaluations of the Olweus program have consistently found that it is among the most effective ways to reduce bullying in schools (Gaffney et al., 2019). The same type of program would likely be useful in addressing other types of school delinquency.

Limitations

While this study used variables derived from macro-level theories of crime, this study did not fully test the relationship between criminological theory and school delinquency. For example, in order to test collective efficacy, effect sizes would need to contain measures for both social cohesion and shared norms (Sampson et al., 1997), but the majority of the studies used contained one or the other. A related concern is that the predictor variables may span multiple criminological perspectives. For example, the monitoring component of the student/teacher ratio is consistent with the guardianship component of routine activity theory. However, on the individual level, the monitoring effect is also consistent with social bond theory. Additionally, not every macro-level criminological perspective was included due to lack of prior research. For example, it would have been beneficial to include macro-level strain theory and subculture perspective. However, a sufficient number of effect sizes in previous research could not be located.

Another limitation of this meta-analysis is the small number of datasets used in the analyses, which may potentially reduce statistical power. Unlike other types of analyses, no power analysis test exists to determine the power of three-level meta-analyses, and scholars are not in agreement as to how many unique studies and datasets are needed for the meta-analysis to be sufficiently powered (Assink & Wibbelink, 2016). While this meta-analysis had similar numbers of studies and datasets to other hierarchal meta-analyses (Steffgens et al., 2013; Welmers-Van De Poll et al., 2018), findings should be interpreted with caution.

There is a debate among researchers using meta-analytic methods as to whether it is appropriate to include unpublished work. Unpublished studies have not gone through the peer-review process, and thus methodologically quality has not been assessed (Pratt, 2010). However, including unpublished studies has been considered a best practice of meta-analytic research (McKenzie et al., 2021), and unpublished studies were included in this study for three reasons. First, meta-analyses aim to “take stock” of all research done on a topic. If too many restrictions are placed on inclusion criteria, the meta-analysis cannot examine the full body of evidence (Glass, 2015; Turanovic & Pratt, 2020). Next, excluding unpublished studies would eliminate four datasets that do not appear in any of the published studies. Related to the previous paragraph, this would potentially reduce statistical power even further (Turanovic & Pratt, 2020). Finally, although including unpublished work means that the analysis contains studies of various methodological rigor, this is controlled for with the use of moderating analyses for publication status and issues of methodological quality.

Conclusions

In sum, school delinquency continues to be a significant issue, and negative consequences can last into adulthood. The finding that situational crime prevention approaches did not prevent school delinquency offers some practical implications. An investment of resources to improve the climate of K-12 public schools, rather than install metal detectors and surveillance cameras, may be beneficial. Creating a climate that fosters trust, respect, and communication may help students feel safer and improve their behaviors, including reducing the occurrence of misbehavior and bullying (VanLone et al., 2019). While this study helped to shed some light on school-level predictors of school delinquency, research must continue, and should focus on examining both individual-level and school-level factors simultaneously. Hopefully, this whole-school approach will reveal the best way to use scarce school resources in order to make schools even safer than before.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Natasha Pusch

Author Biography

Natasha Pusch is an assistant professor in the Department of Sociology, Anthropology, and Social Work at Texas Tech University. She earned her PhD in Criminology and Criminal Justice and Arizona State University. Her research interests include school violence, juvenile offending and victimization, criminological theory, and gender and crime.

References

Adams

Mrug

(2019). Individual-and school-level predictors of violence perpetration, victimization, and perceived safety in middle and high schools. Journal of School Violence, 18, 468–482. https://doi.org/10.1080/15388220.2018.1528551

*Agirdag

Demanet

Van Houtte

Van Avermaet

. (2011). Ethnic school composition and peer victimization: A focus on the interethnic school climate. International Journal of Intercultural Relations, 35, 465–473. https://doi.org/10.1016/j.ijintrel.2010.09.009

*Agnich

L. E.

Miyazaki

(2013). A cross-national analysis of principals’ reports of school violence. International Criminal Justice Review, 23, 378–400. https://doi.org/10.1177/1057567713515273

Aldridge

J. M.

McChesney

Afari

(2018). Relationships between school climate, bullying and delinquent behaviours. Learning Environments Research, 21, 153–172. https://doi.org/10..1007/s10984-017-9249-6

Assink

Wibbelink

C. J.

(2016). Fitting three-level meta-analytic models in R: A step-by-step tutorial. The Quantitative Methods for Psychology, 12, 154–174.

Augustine

M. C.

(2003). A multilevel social control model of everyday school delinquency. [Unpublished doctoral dissertation]. University of Kentucky.

Azeredo

C. M.

Rinaldi

A. E. M.

de Moraes

C. L.

Levy

R. B.

Menezes

P. R

. (2015). School bullying: A systematic review of contextual-level risk factors in observational studies. Aggression and Violent Behavior, 22, 65–76. https://doi.org/10.1016/j.avb.2015.04.006

Bates

(2008). School collective efficacy: Pulling students up or pushing students out. [Unpublished doctoral dissertation]. University of Albany.

Bauer

(2008). School structural characteristics and social capital: The impact of the school environment on criminal and non-criminal incidents in public schools. [Unpublished doctoral dissertation]. American University.

10.

Blosnich

Bossarte

. (2011). Low-level violence in schools: Is there an association between school safety measures and peer victimization? Journal of School Health, 81, 107–113. https://doi.org/10.1111/j.1746-1561.2010.00567.x

11.

Borenstein

Hedges

L. V.

Higgins

J. P.

Rothstein

H. R.

(2011). Introduction to meta-analysis. John Wiley & Sons.

12.

*Boswell

M. A.

(2016). School level predictors of bullying among high school students. [Unpublished doctoral dissertation]. University of Kentucky.

13.

Bradshaw

C. P.

Mitchell

M. M.

Leaf

P. J.

(2010). Examining the effects of schoolwide positive behavioral interventions and supports on student outcomes: Results from a randomized controlled effectiveness trial in elementary schools. Journal of Positive Behavior Interventions, 12, 133–148. https://doi.org/10.1177/1098300709334798

14.

*Bradshaw

C. P.

Sawyer

A. L.

O’Brennan

L. M.

(2009). A social disorganization perspective on bullying-related attitudes and behaviors: The influence of school context. American Journal of Community Psychology, 43, 204–220. https://doi.org/10.1007/s10464-009-9240-1

15.

*Bradshaw

C. P.

Waasdorp

T. E.

Johnson

S. L.

(2015). Overlapping verbal, relational, physical, and electronic forms of bullying in adolescence: Influence of school context. Journal of Clinical Child & Adolescent Psychology, 44, 494–508. https://doi.org/10.1080/15374416.2014.893516

16.

*Brezina

Piquero

A. R.

Mazerolle

(2001). Student anger and aggressive behavior in school: An initial test of Agnew’s macro-level strain theory. Journal of Research in Crime and Delinquency, 38, 362–386. https://doi.org/10.1177/0022427801038004002

17.

*Broll

Lafferty

(2018). Guardians of the hallways? School resource officers and bullying. Safer Communities, 14, 202–212. https://doi.org/10.1108/SC-06-2018-0018

18.

*Byrd

K. M.

Kahle

L. L.

Peguero

A. A.

Popp

A. M.

(2015). Social control and intersectionality: A multilevel analysis of school misconduct, location, race, ethnicity, and sex. Sociological Spectrum, 35, 109–135. https://doi.org/10.1080/02732173.2014.1000552

19.

Cecen-Celik

Keith

(2019). Analyzing predictors of bullying victimization with routine activity and social bond perspectives. Journal of Interpersonal Violence, 34, 3807–3832. https://doi.org/10.1177/0886260516672941

20.

*Chen

(2008). Communities, students, schools, and school crime: A confirmatory study of crime in US high schools. Urban Education, 43, 301–318. https://doi.org/10.1177/0042085907311791

21.

*Chen

Jacobson

K. C.

(2013). Impulsivity moderates promotive environmental influences on adolescent delinquency: A comparison across family, school, and neighborhood contexts. Journal of Abnormal Child Psychology, 41, 1133–1143. https://doi.org/10.1007/s10802-013-9754-8

22.

Chui

W. H.

Chan

H. C. O.

(2015). Self-control, school bullying perpetration, and victimization among Macanese adolescents. Journal of Child and Family Studies, 24, 1751–1761. http://doi.org/10.1007/s10826-014-9979-3

23.

Clarke

R. V.

(1983). Situational crime prevention: Its theoretical basis and practical scope. In Tonry

Morris

(Eds.), Crime & justice: A annual review of research (Vol. 4, pp. 225–256). University of Chicago Press.

24.

*Coffey

B. S.

(2017). Environmental factors and school disorder: The role of urbanicity. [Unpublished master’s thesis]. East Tennessee State University.

25.

Cohen

L. E.

Felson

(1979). Social change and crime rate trends: A routine activity approach. American Sociological Review, 44, 588–608.

26.

*Cornell

Huang

(2016). Authoritative school climate and high school student risk behavior: A cross-sectional multi-level analysis of student self-reports. Journal of Youth and Adolescence, 45, 2246–2259. https://doi.org/10.1007/s10964-016-0424-3

27.

*Cuellar

M. J.

(2018). School safety strategies and their effects on the occurrence of school-based violence in US high schools: An exploratory study. Journal of School Violence, 17, 28–45. https://doi.org/10.1080/15388220.2016.1193742

28.

*Demanet

Van Houtte

(2011). Social-ethnic school composition and school misconduct: Does sense of futility clarify the picture? Sociological Spectrum, 31, 224–256. https://doi.org/10.1080/02732173.2011.541343

29.

*Demanet

Van Houtte

(2012). School belonging and school misconduct: The differing role of teacher and peer attachment. Journal of Youth and Adolescence, 41, 499–514. https://doi.org/10.1007/s10964-011-9674-2

30.

*Demanet

Van Houtte

(2019). Socioeconomic status, economic deprivation, and school misconduct: An inquiry into the role of academic self-efficacy in four European cities. Social Psychology of Education, 22, 1–22. https://doi.org/10.1007/s11218-018-9470-x

31.

*Devlin

D. N.

Gottfredson

D. C.

(2018). The roles of police officers in schools: Effects on the recording and reporting of crime. Youth Violence and Juvenile Justice, 16, 208–223. https://doi.org/10.1177/1541204016680405

32.

Dhami

M. K.

Hoglund

W. L.

Leadbeater

B. J.

Boone

E. M

. (2005). Gender-linked risks for peer physical and relational victimization in the context of school-level poverty in first grade. Social Development, 14, 532–549. https://doi.org/10.1111/j.1467-9507.2005.00315.x

33.

Dhungana

K. J.

(2012). Transcending beyond the schoolyard: A multilevel examination of the environmental influences and prevelance of traditional and cyber bullying perpetration. [Unpublished doctoral dissertation]. Florida State University.

34.

Diliberti

Jackson

Correa

Padgett

(2019). Crime, violence, discipline, and safety in U.S. public schools: Findings from the school survey on crime and safety: 2017–18 (NCES 2019-061). U.S. Department of Education. National Center for Education Statistics.

35.

*DiPietro

S. M.

Slocum

L. A.

Esbensen

F. A.

(2015). School climate and violence: Does immigrant status matter? Youth Violence and Juvenile Justice, 13, 299–322. https://doi.org/10.1177/1541204014547589

36.

Elsass

H. J.

Schildkraut

Stafford

M. C.

(2016). Studying school shootings: Challenges and considerations for research. American Journal of Criminal Justice, 41, 444–464. https://doi.org/10.1007/s12103-015-9311-9

37.

*Fink

Patalay

Sharpe

Wolpert

(2018). Child- and school-level predictors of children’s bullying behavior: A multilevel analysis in 648 primary schools. Journal of Educational Psychology, 110, 17–26. https://doi.org/10.1037/edu0000204

38.

Fisher

B. W.

Hennessy

E. A.

(2016). School resource officers and exclusionary discipline in US high schools: A systematic review and meta-analysis. Adolescent Research Review, 1, 217–233. https://doi.org/10.1007/s40894-015-0006-8

39.

Fisher

B. W.

Higgins

E. M.

Homer

E. M.

(2019). School crime and punishment and the implementation of security cameras: Findings from a national longitudinal study. Justice Quarterly, 38, 22–46. https://doi.org/10.1177/0011128719875702

40.

*Fisher

B. W.

Mowen

T. J.

Boman

J. H.

(2018). School security measures and longitudinal trends in adolescents’ experiences of victimization. Journal of Youth and Adolescence, 47, 1221–1237. https://doi.org/10.1007/s10964-018-0818-5

41.

*Fissel

E. R.

Wilcox

Tillyer

M. S.

(2019). School discipline policies, perceptions of justice, and in-school delinquency. Crime & Delinquency, 65, 1343–1370. https://doi.org/10.1177/0011128718794186

42.

Gaffney

Ttofi

M. M.

Farrington

D. P

. (2019). Evaluating the effectiveness of school-bullying prevention programs: An updated meta-analytical review. Aggression and Violent Behavior, 45, 111–133. https://doi.org/10.1016/j.avb.2018.07.001

43.

Gerlinger

J. C.

(2016). Preventing school bullying: Should schools prioritize an authoritative school discipline approach over security measures? Journal of School Violence, 15, 133–157. https://doi.org/10.1016/j.avb.2019.01.010

44.

Gibbs

M. C.

(2002). The role of school context in the development of aggressive behavior in children. [Unpublished doctoral dissertation] Texas A&M University.

45.

Glass

G. V

. (2015). Meta-analysis at middle age: A personal history. Research Synthesis Methods, 6, 221–231. https://doi.org/10.1002/jrsm.1133

46.

Gonzalez

J. M. R.

Jetelina

K. K.

Jennings

W. G.

(2016). Structural school safety measures, SROs, and school-related delinquent behavior and perceptions of safety. Policing: An International Journal of Police Strategies & Management, 39, 438–454. https://doi.org/10.1108/PIJPSM-05-2016-0065

47.

González

(2012). Keeping kids in schools: Restorative justice, punitive discipline, and the school to prison pipeline. Journal of Law & Education, 41, 281–335.

48.

Gottfredson

D. C.

(2001). Schools and delinquency. Cambridge University Press.

49.

*Gottfredson

D. C.

DiPietro

S. M.

(2011). School size, social capital, and student victimization. Sociology of Education, 84, 69–89. https://doi.org/10.1177/0038040710392718

50.

Gottfredson

G. D.

Gottfredson

D. C.

Payne

A. A.

Gottfredson

N. C.

(2005). School climate predictors of school disorder: Results from a national study of delinquency prevention in schools. Journal of Research in Crime and Delinquency, 42, 412–444. https://doi.org/10.1177/0022427804271931

51.

*Gregory

Cornell

Fan

Sheras

Shih

T. H.

Huang

(2010). Authoritative school discipline: High school practices associated with lower bullying and victimization. Journal of Educational Psychology, 102, 299–309. https://doi.org/10.1037/a0018562

52.

Hemphill

S. A.

Tollit

Herrenkohl

T. I.

(2014). Protective factors against the impact of school bullying perpetration and victimization on young adult externalizing and internalizing problems. Journal of School Violence, 13, 125–145. https://doi.org/10.1080/15388220.2013.844072

53.

Hirschfield

P. J.

(2008). Preparing for prison? The criminalization of school discipline in the USA. Theoretical Criminology, 12, 79–101. https://doi.org/10.1177/1362480607085795

54.

*Hoffmann

J. P.

Dufur

M. J.

(2008). Family and school capital effects on delinquency: Substitutes or complements? Sociological Perspectives, 51, 29–62. https://doi.org/10.1525/sop.2008.51.1.29

55.

*Jennings

W. G.

Khey

D. N.

Maskaly

Donner

C. M.

(2011). Evaluating the relationship between law enforcement and school security measures and violent crime in schools. Journal of Police Crisis Negotiations, 11, 109–124. https://doi.org/10.1080/15332586.2011.581511

56.

*Johnson

C. L.

Wilcox

Peterson

(2019). Stressed out and strapped: Examining the link between psychological difficulties and student weapon carrying and use. Criminal Justice and Behavior, 47, 980–998. https://doi.org/10.1177/0093854819826110

57.

Johnson

S. L

. (2009). Improving the school environment to reduce school violence: A review of the literature. Journal of School Health, 79, 451–465. https://doi.org/10.1111/j.17461561.2009.00435.x

58.

Johnson

S. L.

Waasdorp

T. E.

Cash

A. H.

Debnam

K. J.

Milam

A. J.

Bradshaw

C. P.

(2017). Assessing the association between observed school disorganization and school violence: Implications for school climate interventions. Psychology of Violence, 7, 181–191. https://doi.org/10.1037/vio0000045

59.

*Jones

S. E.

(2003). Control theories and school antisocial behaviors: Understanding the interdependency of self-control, social bonds, and school collective efficacy [Unpublished doctoral dissertation]. University of Kentucky.

60.

Jonson

C. L.

(2017). Preventing school shootings: The effectiveness of safety measures. Victims & Offenders, 12, 956–973. https://doi.org/10.1080/15564886.2017.1307293

61.

Khoury-Kassabri

Benbenishty

Avi Astor

Zeira

(2004). The contributions of community, family, and school variables to student victimization. American Journal of Community Psychology, 34, 187–204. https://doi.org/10.1007/s10464-004-7414-4

62.

*Klein

Cornell

(2010). Is the link between large high schools and student victimization an illusion? Journal of Educational Psychology, 102, 933–946. https://doi.org/10.1037/a0019896

63.

*Koo

D. J.

Peguero

A. A.

Shekarkhar

(2012). The “model minority” victim: Immigration, gender, and Asian American vulnerabilities to violence at school. Journal of Ethnicity in Criminal Justice, 10, 129–147. https://doi.org/10.1080/15377938.2011.609405

64.

Kurlychek

M. C.

(2004). Victimization: Examining the influence of student, school, and community characteristics. [Unpublished doctoral dissertation]. Pennsylvania State University.

65.

*Larochette

A. C.

Murphy

A. N.

Craig

W. M.

(2010). Racial bullying and victimization in Canadian school-aged children: Individual and school level effects. School Psychology International, 31, 389–408. https://doi.org/10.1177/0143034310377150

66.

*Latimore

T. L.

Peguero

A. A.

Popp

A. M.

Shekarkhar

Koo

D. J.

(2018). School-based activities, misbehavior, discipline, and racial and ethnic disparities. Education and Urban Society, 50, 403–434. https://doi.org/10.1177/0013124517713603

67.

*Lee

G. P.

Martins

S. S.

Pas

E. T.

Bradshaw

C. P

. (2014). Examining potential school contextual influences on gambling among high school youth. The American Journal on Addictions, 23, 510–517. https://doi.org/10.1111/j.1521-0391.2014.12142.x

68.

*Lehman

(2014). Gender differences in bullying victimization: The role of academics and school context. Sociological Spectrum, 34, 549–570. https://doi.org/10.1080/02732173.2014.947454

69.

*Lesneskie

Block

(2017). School violence: The role of parental and community involvement. Journal of School Violence, 16, 426–444. https://doi.org/10.1080/15388220.2016.1168744

70.

*Lleras

(2008). Hostile school climates: Explaining differential risk of student exposure to disruptive learning environments in high school. Journal of School Violence, 7, 105–135. https://doi.org/10.1080/15388220801955604

71.

*Martinez

McMahon

S. D.

Treger

(2016). Individual-and school-level predictors of student office disciplinary referrals. Journal of Emotional and Behavioral Disorders, 24, 30–41. https://doi.org/10.1177/1063426615588289

72.

*Maskaly

Donner

C. M.

Lanterman

Jennings

W. G.

(2011). On the association between SROs, private security guards, use-of-force capabilities, and violent crime in schools. Journal of Police Crisis Negotiations, 11, 159–176. https://doi.org/10.1080/15332586.2011.587381

73.

*Maume

M. O.

Kim-Godwin

Y. S.

Clements

C. M.

(2010). Racial tensions and school crime. Journal of Contemporary Criminal Justice, 26, 339–358. https://doi.org/10.1177/1043986210369283

74.

McKenzie

J. E.

Brennan

S. E.

Ryan

R. E.

Thomson

H. J.

Johnston

R. V.

Thomas

(2021). Defining the criteria for including studies and how they will be grouped for the synthesis. In Higgins

J. P. T.

Thomas

Chandler

Cumpston

Page

M. J.

Welch

V. A.

(Eds.), Cochrane handbook for systematic reviews of interventions 6.2. www.training.cochrane.org/handbook

75.

Moon

Alarid

L. F.

(2015). School bullying, low self-control, and opportunity. Journal of Interpersonal Violence, 30, 839–856. https://doi.org/10.1177/0886260514536281

76.

Moule

R. K.

Jr Fox

(2021). Belief in the code of the street and individual involvement in offending: A meta-analysis. Youth Violence and Juvenile Justice, 19, 227–247. https://doi.org/10.1177/1541204020927737

77.

Myers

Turanovic

J. J.

Lloyd

K. M.

Pratt

T. C.

(2020). The victimization of LGBTQ students at school: A meta-analysis. Journal of School Violence, 19, 421–432. https://doi.org/10.1080/15388220.2020.1725530

78.

*Na

Gottfredson

D. C.

(2013). Police officers in schools: Effects on school crime and the processing of offending behaviors. Justice Quarterly, 30, 619–650. https://doi.org/10.1080/07418825.2011.615754

79.

Nakagawa

Santos

E. S.

(2012). Methodological issues and advances in biological meta-analysis. Evolutionary Ecology, 26, 1253–1274. https://doi.org/10.1007/s10682-012-9555-5

80.

Nakamoto

Schwartz

(2010). Is peer victimization associated with academic achievement? A meta-analytic review. Social Development, 19, 221–242. https://doi.org/10.1111/j.1467-9507.2009.00539.x

81.

*Nickerson

A. B.

Martens

M. P.

(2008). School violence: Associations with control, security/enforcement, educational/therapeutic approaches, and demographic factors. School Psychology Review, 37, 228–243. https://doi.org/10.1080/02796015.2008.12087897

82.

*Olmstead

G. T.

(2005). Violence prevention programs and zero tolerance policies: Effectiveness for school safety as reported in the SSOCS [Unpublished doctoral dissertation]. University of Arkansas.

83.

Olsson

Låftman

S. B.

Modin

(2017). School collective efficacy and bullying behaviour: A multilevel study. International Journal of Environmental Research and Public Health, 14, 1607–1619. https://doi.org/10.3390/ijerph14121607

84.

Olweus

Limber

Mihalic

(1999). Blueprints for violence prevention (Vol. 9). The Bullying Prevention Program. Institute of Behavioral Science, University of Colorado.

85.

*O’Neill

McGloin

J. M

. (2007). Considering the efficacy of situational crime prevention in schools. Journal of Criminal Justice, 35, 511–523. https://doi.org/10.1016/j.jcrimjus.2007.07.004

86.

*Parker

E. M.

Johnson

S. L.

Debnam

K. J.

Milam

A. J.

Bradshaw

C. P.

(2017). Teen dating violence victimization among high school students: A multilevel analysis of school-level risk factors. Journal of School Health, 87, 696–704. https://doi.org/10.1111/josh.12538

87.

Patchin

J. W.

Hinduja

(2011). Traditional and nontraditional bullying among youth: A test of general strain theory. Youth & Society, 43, 727–751. https://doi.org/10.1177/0044118X10366951

88.

*Payne

A. A.

(2008). A multilevel analysis of the relationships among communal school organization, student bonding, and delinquency. Journal of Research in Crime and Delinquency, 45, 429–455. https://doi.org/10.1177/0022427808322621

89.

*Payne

A. A

. (2009). Girls, boys, and schools: Gender differences in the relationships between school-related factors and student deviance. Criminology, 47, 1167–1200. https://doi.org/10.1111/j.1745-9125.2009.00172.x

90.

*Payne

A. A.

(2012). Communal school organization effects on school disorder: Interactions with school structure. Deviant Behavior, 33, 507–524. https://doi.org/10.1080/01639625.2011.636686

91.

Payne

A. A.

Gottfredson

D. C.

Gottfredson

G. D

. (2003). Schools as communities: The relationships among communal school organization, student bonding, and school disorder. Criminology, 41, 749–778. https://doi.org/10.1111/j.1745-9125.2003.tb01003.x

92.

*Peguero

A. A.

(2011). Immigration, schools, and violence: Assimilation and student misbehavior. Sociological Spectrum, 31, 695–717. https://doi.org/10.1080/02732173.2011.606726

93.

Peguero

A. A.

Hong

J. S.

(2019). Are violence and disorder at school placing adolescents within immigrant families at higher risk of dropping out? Journal of School Violence, 18, 241–258. https://doi.org/10.1080/15388220.2018.1459632

94.

*Peguero

A. A.

Jiang

. (2014). Social control across immigrant generations: Adolescent violence at school and examining the immigrant paradox. Journal of Criminal Justice, 42, 276–287. https://doi.org/10.1016/j.jcrimjus.2014.01.003

95.

*Peguero

A. A.

Jiang

(2016). Backlash for breaking racial and ethnic breaking stereotypes: Adolescent school victimization across contexts. Journal of Interpersonal Violence, 31, 1047–1073. https://doi.org/10.1177/0886260514564063

96.

Peterson

R. A.

Brown

S. P.

(2005). On the use of beta coefficients in meta-analysis. Journal of Applied Psychology, 90, 175–181. https://doi.org/10.1037/0021-9010.90.1.175

97.

*Pigott

Stearns

A. E.

Khey

D. N.

(2018). School resource officers and the school to prison pipeline: Discovering trends of expulsions in public schools. American Journal of Criminal Justice, 43, 120–138. https://doi.org/10.1007/s12103-017-9412-8

98.

Plank

S. B.

Bradshaw

C. P.

Young

(2009). An application of “broken-windows” and related theories to the study of disorder, fear, and collective efficacy in schools. American Journal of Education, 115, 227–247. https://doi.org/10.1086/595669

99.

*Popp

A. M.

Peguero

A. A.

(2011). Routine activities and victimization at school: The significance of gender. Journal of Interpersonal Violence, 26, 2413–2436. https://doi.org/10.1177/0886260510383021

100.

Popp

A. M.

Peguero

A. A.

(2012). Social bonds and the role of school-based victimization. Journal of Interpersonal Violence, 27, 3366–3388. http://doi.org/10.1177/0886260512445386

101.

Pratt

T. C.

(2010). Meta-analysis in criminal justice and criminology: What it is, when it’s useful, and what to watch out for. Journal of Criminal Justice Education, 21, 152–168. https://doi.org/10.1080/10511251003693678

102.

Pratt

T. C.

Cullen

F. T.

(2005). Assessing macro-level predictors and theories of crime: A meta-analysis. Crime and Justice, 32, 373–450. https://doi.org/10.1086/655357

103.

Pratt

T. C.

Turanovic

J. J.

Fox

K. A.

Wright

K. A.

(2014). Self-control and victimization: A meta-analysis. Criminology, 52, 87–116. https://doi.org/10.1111/1745-9125.12030

104.

Pusch

Holtfreter

(2021). Sex-based differences in criminal victimization of adolescents: A meta-analysis. Journal of Youth and Adolescence, 50, 4–28. https://doi.org/10.1007/s10964-020-01321-y

105.

*Rezapour

Khanjani

Mirzai

. (2019). Exploring associations between school environment and bullying in Iran: Multilevel contextual effects modeling. Children and Youth Services Review, 99, 54–63. https://doi.org/10.1016/j.childyouth.2019.01.036

106.

*Richard

J. F.

Schneider

B. H.

Mallet

(2012). Revisiting the whole-school approach to bullying: Really looking at the whole school. School Psychology International, 33, 263–284. https://doi.org/10.1177/0143034311415906

107.

Rosenthal

(1979). The file drawer problem and tolerance for null results. Psychological Bulletin, 86, 638–641.

108.

Rosenthal

(1994). Parametric measures of effect size. In Cooper

Hedges

(Eds.), The handbook of research synthesis (pp. 231–244). Russell Sage Foundation.

109.

Rowan

Bossert

S. T.

Dwyer

D. C.

(1983). Research on effective schools: A cautionary note. Educational Researcher, 12, 24–31.

110.

Saarento

Garandeau

C. F.

Salmivalli

(2015). Classroom-and school-level contributions to bullying and victimization: A review. Journal of Community & Applied Social Psychology, 25, 204–218. https://doi.org/10.1002/casp.2207

111.

Sampson

R. J.

Raudenbush

S. W.

Earls

(1997). Neighborhoods and violent crime: A multilevel study of collective efficacy. Science, 277, 918–924.

112.

Sapouna

(2010). Collective efficacy in the school context: Does it help explain victimization and bullying among Greek primary and secondary school students? Journal of Interpersonal Violence, 25, 1912–1927. https://doi.org/10.1177/0886260509354509

113.

Schreck

C. J.

Miller

J. M.

Gibson

C. L.

(2003). Trouble in the school yard: A study of the risk factors of victimization at school. Crime & Delinquency, 49, 460–484. https://doi.org/10.1177/0011128703049003006

114.

Seldin

Yanez

(2019). School crime supplement (SCS) to the national crime victimization survey, 2005 through 2017. U.S. Department of Justice, Bureau of Justice Statistics.

115.

Servoss

T. J.

(2017). School security and student misbehavior: A multi-level examination. Youth & Society, 49, 755–778. https://doi.org/10.1177/0044118X14561007

116.

*Sevigny

E. L.

Zhang

(2018). Do barriers to crime prevention moderate the effects of situational crime prevention policies on violent crime in high schools? Journal of School Violence, 17, 164–179. https://doi.org/10.1080/15388220.2016.1275657

117.

Sharkey

J. D.

Dowdy

E. T.

Twyford

Furlong

M. J.

(2011). An overview of measurement issues in school violence and school safety research. In Jimerson

S. R.

Nickerson

A. B.

Mayer

M. J.

Furlong

M. J.

(Eds.), The handbook of school violence and school safety: international research and practice (2nd ed., pp. 259–272). Routledge.

118.

Shaw

C. R.

McKay

H. D.

(1942). Juvenile delinquency and urban areas. University of Chicago Press.

119.

Smith-Adcock

Lee

S. M.

Kerpelman

Majuta

Young

C. B.

(2013). Reputation enhancement and school delinquency: A prospective study using the National Educational Longitudinal Survey [NELS: 88]. Youth & Society, 45, 201–220. https://doi.org/10.1177/0044118X11410895

120.

Steffgen

Recchia

Viechtbauer

. (2013). The link between school climate and violence in school: A meta-analytic review. Aggression and Violent Behavior, 18, 300–309. https://doi.org/10.1016/j.avb.2012.12.001.

121.

Stickle

W. P.

(2009). Predictors of in-school weapon carrying. [Unpublished doctoral dissertation]. University of Maryland, College Park.

122.

*Stretesky

P. B.

Hogan

M. J.

(2005). Segregation and school disorder. The Social Science Journal, 42, 405–420. https://doi.org/10.1016/j.soscij.2005.06.007

123.

*Swartz

Wilcox

Ousey

G. C.

(2017). Culture as values or culture in action? Street codes and student violent offending. Victims & Offenders, 12, 868–890. https://doi.org/10.1080/15564886.2017.1307294

124.

*Tanner-Smith

E. E.

Fisher

B. W.

Addington

L. A.

Gardella

J. H.

(2018). Adding security, but subtracting safety? Exploring schools’ use of multiple visible security measures. American Journal of Criminal Justice, 43, 102–119. https://doi.org/10.1007/s12103-017-9409-3

125.

*Theriot

M. T

. (2009). School resource officers and the criminalization of student behavior. Journal of Criminal Justice, 37, 280–287. https://doi.org/10.1016/j.jcrimjus.2009.04.008

126.

*Tillyer

M. S.

Fisher

B. S.

Wilcox

(2011). The effects of school crime prevention on students’ violent victimization, risk perception, and fear of crime: A multilevel opportunity perspective. Justice Quarterly, 28, 249–277. https://doi.org/10.1080/07418825.2010.493526

127.

*Tillyer

M. S.

Wilcox

Fissel

E. R.

(2018). Violence in schools: Repeat victimization, low self-control, and the mitigating influence of school efficacy. Journal of Quantitative Criminology, 34, 609–632. https://doi.org/10.1007/s10940-017-9347-8

128.

*Tillyer

M. S.

Wilcox

Gialopsos

B. M

. (2010). Adolescent school-based sexual victimization: Exploring the role of opportunity in a gender-specific multilevel analysis. Journal of Criminal Justice, 38, 1071–1081. https://doi.org/10.1016/j.jcrimjus.2010.07.010

129.

Ttofi

M. M.

Bowes

Farrington

D. P.

Lösel

(2014). Protective factors interrupting the continuity from school bullying to later internalizing and externalizing problems: A systematic review of prospective longitudinal studies. Journal of School Violence, 13, 5–38. https://doi.org/10.1080/15388220.2013.857345

130.

Turanovic

J. J.

Pratt

T. C.

(2020). Meta-analysis in criminology and criminal justice: Challenging the paradigm and charting a new path forward. Justice Evaluation Journal, 1–27. https://doi.org/10.1080/24751979.2020.1775107

131.

VanLone

Freeman

LaSalle

Gordon

Polk

Rocha Neves

(2019). A practical guide to improving school climate in high schools. Intervention in School and Clinic, 55, 39–45. https://doi.org/10.1177/1053451219832988

132.

Viechtbauer

(2010). Conducting meta-analyses in R with the metafor package. Journal of Statistical Software, 36, 1–48.

133.

*Vogel

Barton

M. S.

(2013). Impulsivity, school context, and school misconduct. Youth & Society, 45, 455–479. https://doi.org/10.1177/0044118X11421941

134.

Wang

Chen

Zhang

Oudekerk

B.A.

(2020). Indicators of school crime and safety: 2019 (NCES 2020-063/NCJ 254485). National Center for Education Statistics, U.S. Department of Education, and Bureau of Justice Statistics, Office of Justice Programs, U.S. Department of Justice.

135.

*Wang

(2013). Bullying among US school children: An examination of race/ethnicity and school-level variables on bullying [Unpublished doctoral dissertation]. Clemson University.

136.

*Way

S. M.

(2003). For their own good? The effects of school discipline and disorder on student behavior and academic achievement [Unpublished doctoral dissertation]. University of Arizona.

137.

*Wei

H. S.

Williams

J. H.

Chen

J. K.

Chang

H. Y

. (2010). The effects of individual characteristics, teacher practice, and school organizational factors on students’ bullying: A multilevel analysis of public middle schools in Taiwan. Children and Youth Services Review, 32, 137–143. https://doi.org/10.1016/j.childyouth.2009.08.004

138.

Welmers-Van

Poll

M. J.

Roest

J. J.

Van der Stouwe

Van den Akker

A. L.

Stams

G. J. J. M.

Escudero

Overbeek

G. J.

De Swart

J. J. W.

(2018). Alliance and treatment outcome in family-involved treatment for youth problems: A three-level meta-analysis. Clinical Child and Family Psychology Review, 21, 146–170. https://doi.org/10.1007/s10567-017-0249-y

139.

Whitehurst

G. J.

Chingos

M. M.

(2011). Class size: What research says and what it means for state policy. Brown Center on Education Policy at Brookings. The Brookings Institution.

140.

*Wilcox

Augustine

M. C.

Clayton

R. R.

(2006). Physical environment and crime and misconduct in Kentucky schools. Journal of Primary Prevention, 27, 293–313. https://doi.org/10.1007/s10935-006-0034-z

141.

*Wilcox

Clayton

R. R.

(2001). A multilevel analysis of school-based weapon possession. Justice Quarterly, 18, 509–541. https://doi.org/10.1080/07418820100095001

142.

Wilson

. (2004). The interface of school climate and school connectedness and relationships with aggression and victimization. The Journal of School Health, 74, 293–300. https://doi.org/10.1111/j.1746-1561.2004.tb08286.x

143.

Wilson

J. Q.

Kelling

G. L.

(1982). Broken windows. Atlantic Monthly, 249, 29–38.

144.

Wilson

(2020). At what cost? A review of school police funding and accountability across the U.S. south. Newsletter by the Intercultural Development Research Association. https://www.idra.org/resource-center/at-what-cost-a-review-of-school-police-funding-and-accountability-across-the-u-s-south/

145.

Zaykowski

Gunter

(2012). Youth victimization: School climate or deviant lifestyles? Journal of Interpersonal Violence, 27, 431–452. https://doi.org/10.1177/0886260511421678

146.

*Zhao

(2018). Impacts of the school context on victimization in a sample of Chinese adolescents: A multilevel approach. Journal of Contemporary Criminal Justice, 34, 196–218. https://doi.org/10.1177/1043986218761964