Strain and Schools

http://yvj.sagepub.com

Youth Violence and Juvenile Justice

DOI: 10.1177/1541204007308430 2008; 6; 115 Youth Violence and Juvenile Justice

Daniel R. Lee and Jeffrey W. Cohen Examining Strain in a School Context

http://yvj.sagepub.com/cgi/content/abstract/6/2/115 The online version of this article can be found at:

Published by:

http://www.sagepublications.com

On behalf of:

Academy of Criminal Justice Sciences

can be found at:Youth Violence and Juvenile Justice Additional services and information for

http://yvj.sagepub.com/cgi/alerts Email Alerts:

http://yvj.sagepub.com/subscriptions Subscriptions:

http://www.sagepub.com/journalsReprints.navReprints:

http://www.sagepub.com/journalsPermissions.navPermissions:

http://yvj.sagepub.com/cgi/content/refs/6/2/115 Citations

at INDIANA UNIVERSITY OF PENNSYLVANIA on October 27, 2008 http://yvj.sagepub.comDownloaded from

http://www.acjs.org

http://yvj.sagepub.com/cgi/alerts

http://yvj.sagepub.com/subscriptions

http://www.sagepub.com/journalsReprints.nav

http://www.sagepub.com/journalsPermissions.nav

http://yvj.sagepub.com/cgi/content/refs/6/2/115


Examining Strain in aSchool ContextDaniel R. LeeJeffrey W. CohenIndiana University of Pennsylvania

General strain theory has accumulated a considerable amount of empirical support. Many ofthese assessments have tested the direct relationship that strain has on crime and delinquency.The research presented here examines the relationship between schools and delinquency withina general strain theory perspective. More specifically, this research examines how schools cannot only act as a source of an individual’s strain and subsequent delinquency but also be a sourcefor mediating or coping with strain and minimizing delinquency. To test the relationship betweenschools and delinquency, data from the National Educational Longitudinal Survey (NELS:88)are analyzed in a model of general strain that specifies sources of school-based strain and sourcesof school-based mechanisms for controlling strain.

Keywords: NELS; school administration; school context; school violence; strain; substanceuse; truancy

Introduction

Schools and delinquency are related in a number of ways. Because of compulsoryeducation laws, young people are legally bound to attend schools for a significant portionof the day and for several months each year. Because younger people are more likely tocommit crimes or delinquent acts than older people, and most criminals are more likely tooffend against those individuals who are most like themselves, schools seem to be not onlya likely place for delinquent acts but also a place for young people to be victimized (seeDeVoe, Peter, Noonan, Snyder, & Baum, 2005; U.S. Department of Education, 2003). Thisscenario of likely offending and likely victimization is one that should be of considerableinterest and importance to the criminological community.

Other than identifying delinquency and victimization, schools are useful for the admin-istration of surveys, and the testing of theories among adolescent samples, and the basictenets and constructs of our most popular criminological theories are conceptually tied to thedaily routines of students. Hirschi’s (1969) theory of social bonding predicts that adoles-cents, who are more committed and involved in prosocial activities (like school) will be lesslikely to commit delinquency acts. Sutherland’s (1947) differential association theory andAkers’ (1977) social learning theory expect that a considerable amount of definitions of andattitudes toward delinquent acts are accepted, shared, or conditioned through adolescent–peer

Youth Violence andJuvenile Justice

Volume 6 Number 2April 2008 115-135

© 2008 Sage Publications10.1177/1541204007308430

http://yvj.sagepub.comhosted at

http://online.sagepub.com

115

Authors’ Note: Correspondence concerning this article should be addressed to Daniel R. Lee, PhD, Department ofCriminology, G-1 McElhaney Hall, Indiana University of Pennsylvania, Indiana, PA 15705; e-mail: [email protected].



associations. Strain theories such as Cohen’s (1955) offer that schools serve as conveyorsof socially prescribed goals and create unique opportunities for these goals to becomeblocked to individuals from lower social and economic classes. Advances in strain theoryhave proposed that schools can be a unique source of social–psychological strain (Agnew,1985, 1992, 2001).

This study examines this social–psychological version of strain and identifies how schoolscan not only act as a source of an individual’s strain but also as a source for mediation orcoping with strain. First, a brief review of the development of strain theory is presented. Then,an application of contemporary strain theory specific to schools is discussed. Finally, anexpanded model of general strain theory is offered as a more complete alternative to under-standing schools as a source of delinquency causation and mediation. This model is assessedwith data drawn from the National Education Longitudinal Study (NELS: 88). From thisassessment, conclusions are presented and direction for continued school-based research andpolicy is offered.

Literature Review

Most versions of strain theory trace their origin to Merton’s (1938) essay about anomieand social structure. In that essay, Merton suggested that personal success and satisfactionderive not only from attaining goals but from surpassing other competitors. This competitivespirit can lead individuals to manipulate different and sometimes illegal means to achievesuccess. These pursuits can also “invite exaggerated anxieties, hostilities . . . and antisocialbehavior” (Merton, 1938, p. 680). To some extent, Merton’s propositions emphasize pecu-niary success, but an expanded interpretation would allow for success to come in manyforms and not be limited to an individual’s financial gains. Cohen (1955) and Cloward andOhlin (1960) elaborated Merton’s thesis and offered specific explanations for juveniledelinquency that included schools as a multifaceted source of strain that might includediminished status and blocked opportunities for social advancement.

For several decades, strain theory was empirically assessed as a macro-level theory;1 thatis, many strain assumptions have been tested through the identification and measurementof socially prescribed goals and achievement of those goals. Some tests have moved awayfrom these aggregate or macro-assessments of the theory and have begun to point towardaspirations to and achievement of individual measures of success (see Figure 1). Thismovement toward an individual model of strain was solidified when Agnew began todevelop General Strain theory (see Agnew, 1985).

General strain theory proposed that crime and delinquency were the result of an individ-ual’s emotional status produced by negative personal relationships (Agnew, 1992). Thestrain from these negative relationships is produced by a greater variety of circumstancesthan those proposed in earlier strain theories, but General Strain theory has included in thesecircumstances a remnant of its theoretical predecessor, namely, the disjunction between anindividual’s aspirations and expectations (see Figure 2). Added to this historic element ofstrain were the removal of positively valued stimuli (e.g., the loss of a boy/girlfriend, thedeath or divorce of parents, or the separation from a group of peers) and the introduction

116 Youth Violence and Juvenile Justice



of negative stimuli (e.g., the presence of a mean-spirited teacher or bully). These strainfulelements can produce within an individual what Agnew has referred to as a “negative affectivestate.” This negative affect is expected to be associated with states of anger, frustration, andrage. Any individual who has developed a negative affect is also likely to experience anincreased likelihood of delinquency or criminality. Despite the increased likelihood ofdelinquency, Agnew proposed that an individual might be able to develop, implement, andenjoy coping strategies that could minimize the likelihood of delinquent responses. Thesecoping strategies can be cognitive, behavioral, or emotional and might include activities likerationalizing stressful events as being temporary, pursuing social support, or participating inexercise or drug abuse.

Since Agnew’s (1985) initial proposition of the theory and subsequent elaborations andapplications (Agnew, 1992, 1995, 2001), the theory has enjoyed a considerable amount ofempirical support. Although many of these empirical assessments include schools as an indi-rect source for the general strain, a more elaborate model might be necessary to accuratelytest whether generalized strain operates within the school setting.

Agnew and White (1992) provided an initial examination of general strain by testing thelikelihood that family, school, and neighborhood problems could affect delinquency anddrug use. Their analysis provided some confirmation of the general strain propositions, but

Lee, Cohen / Examining Strain in a School Context 117

Strain• Goals/Means• Aspirations/Expectations

Crime andDelinquency

Figure 1Traditional Strain Model (Adapted From Cohen, 1955; Merton, 1938)

Strain• Aspirations/Expectations• Remove Positive Stimuli • Introduce Negative Stimuli

Negative Affect • Anger• Frustration

Coping Strategies• Cognitive• Behavioral• Emotional


Absence ofCrime

Delinquency

Figure 2General Strain Model (Adapted From Agnew, 1985, 1992)



the magnitude of the effect could be considered minimal. Although confirmatory, theseresults could be scrutinized because of use of data that measure the constructs of strain anddelinquency longitudinally with a gap of 3 years between the first and last measurement.An attribute of general strain is that it identifies strainful experiences at an individual level;although it is likely that these negative affective experiences occur as a process over time,it is plausible to assume that this process will be more contemporaneous than what occursduring a 3-year time span. Another possibly confounding issue is that delinquent peers werefound to increase participation in delinquency and drug use and lower the measurement ofself-efficacy. These relationships with delinquent peers introduce the theory to competitionfrom differential association and social learning theories.

Another assessment of this process was conducted by Paternoster and Mazerolle (1994).Using data drawn from the National Youth Survey (see Elliott, Huizinga, & Ageton, 1985),a direct test of the “hypotheses about the conditions of strain under which adverse conditionsof strain may be amplified or muted” (Paternoster & Mazerolle, 1994, p. 246) was conducted.Paternoster and Mazerolle found that peer hassles were significantly related to subsequentdelinquency, and this was second in strength to negative relationships with adults and asstrong as moral beliefs and delinquent peers. Although the National Youth Survey data usedin this analysis presented a shorter (1-year) lag between measurements, distinct possibleproblems with this analysis could be with the scaling and measurement of the theoreticalconstructs. Some of the survey items that were used to measure the strain constructs couldactually be considered measures of other theories (e.g., social disorganization). In addition,some survey items intended to measure strain could actually provide an indication thatschools could be a source of coping with negative affective state.

Other tests of general strain theory have found support for specific components ofgeneral strain (see Brezina, 1996; Mazerolle & Piquero, 1997, 1998) and support amongdifferent populations of offenders (see Broidy, 2001; Piquero & Sealock, 2000; Broidy &Agnew, 1997). The diversity of these tests provides confirmation that the theory should betested more explicitly and completely with appropriate data drawn from an appropriatesample within an appropriate context.

The School-Strain Relationship

Agnew (2000) has argued that general strain theory is particularly appropriate to studythe relationship schools have with delinquency. General strain theory predicts that severalsources of strain can accumulate to produce the negative affective state that leads individualsto delinquency. These sources of strain can include neighborhood, familial, and school-basedrelationships (see Figure 3). Outside school, neighborhood problems such as poverty,racism, inequality, and relative deprivation can be examples of negative stimuli. At thefamilial level, negative stimuli can come from dysfunctional relationships with parents or sib-lings. Although these experiences are likely to occur with some frequency among delinquentyouths, the frequency and duration of school-based relationships makes their study partic-ularly interesting.

The school experience can provide a variety of noxious events. Although positive peerrelationships exist for many students, negative peer relationships (e.g., bullying, teasing,




and peer pressure) are also abundant. Likewise, teachers may represent negative relation-ships by exposing students to their poor temperament, demeaning attitude, or unfair grad-ing practices. Low academic achievement or a learning disability might add to a generaldissatisfaction with the entire school environment, increased levels of boredom, and cancontribute to an attitude that school activities are irrelevant to either immediate or futurelife circumstances. Although some students might cope with a lack of peer support by rev-eling in a “loner” status or rationalizing poor grades as being meaningful only to those ina college preparatory curriculum track, it should be expected that most students would findthese experiences negative life events. That is, to most students, these experiences wouldbe strainful and would contribute to the development of a negative affective state and thelikely progression toward delinquent activity.

Although the use of general strain theory to explain school-based delinquency seemsplausible, it would be inappropriate to discount the impact that other theoretical perspectivesand constructs might have within a school context. For instance, differential association(Sutherland, 1947) and social learning (Akers, 1977) theories suggest that the presenceof delinquent peers could increase the likelihood of offending. To be sure, peers who areexperienced in delinquent activities and willing to model their behavior certainly contributeto the school atmosphere and the availability of inappropriate social networks. In addition,social control (Hirschi, 1969) theory suggests that schools represent an opportunity forstudents to become committed and involved in socially appropriate activities that inhibitparticipation in delinquent acts.

In a tangential line of research over several decades, Kaplan and his colleagues have devel-oped and tested a general theory suggesting that social relationships can facilitate the indi-vidual motivation (e.g., negative self-feelings and self-esteem) necessary to engage indelinquency and deviance (see Kaplan, 1972, 1975, 1980, 1984; Kaplan & Damphouse,1997; Kaplan & Johnson, 2001; Kaplan et al., 1986; Kaplan & Peck, 1992). A recent test


Other Strain• Family• Poverty• Inequality

School Strain• Negative Affective State

o Negative Peer Relations o Negative Teacher Relations o Poor Grades o General Dissatisfaction


Lower Social Control Delinquent Peers

Figure 3School-Based Strain Model (Adapted From Agnew, 2000)



has focused on the intervening and mediating effects of negative self-feelings on the rela-tionship between relative deprivation and crime. Stiles, Liu, and Kaplan (2000) analyzed asingle wave of panel data collected from more than 6,000 subjects, who were surveyedwhen they were in their mid-to-late twenties. They found that the fit of several multivariatemodels that considered self-assessments of deprivation (relative to friends, neighbors, andperceived national averages) improved when negative self-feelings were included as anindependent variable. When negative self-feelings were included, the impact of relativedeprivation across social references was either diminished or became statistically insignif-icant. This suggests that negative self-feelings, an independent parallel to negative affect,should be considered in any assessment of social–psychological relationships to delin-quency and crime.

Whether alongside or completely aside from these theoretical perspectives, a moreelaborate or extended school-based model should be able to more accurately define howschools can be a source of both the negative affective state and the coping strategies thatare vital components of general strain theory (see Figure 4). That is, some of the constructsthat previous research has identified as school-based problems that create the negativeaffective state might also be able to contribute to the coping strategies mediating strain. Forinstance, in some individuals, participating in sports or required physical education couldproduce strainful circumstances when the student is unable to succeed athletically. For others,athletics might provide the behavioral coping that Agnew discussed as mediating the strainattributable to poor academic achievement.


School Problems

Strain

Negative Affect

Delinquency

Coping

SchoolMechanisms

School Problems

Strain

Negative Affect

Time 2Time 1

NoDelinquency

Delinquency

Figure 4Conceptual School-Based General Strain Model



A careful survey and measurement of these constructs could provide more explicit evi-dence that individual differences can alter the impact and direction of certain activities. Anability to manipulate these negative affective behaviors into potential coping behaviorscould establish a fruitful school-based prevention program. By allowing the same con-structs to fluctuate between risk and protective factors, delinquency could be seen moreprecisely as an individual phenomenon.

The current study addresses some of the issues discussed above. Specifically, this studyanalyzes data from two waves of a national representative sample separated by a 2-yearinterval. In addition, the current study tests a specified model of school-based strain (seeFigure 5). This model includes school mechanisms as both a contributing and mediatingfactor, so this is an assessment of the impact that schools can have in terms of both increas-ing and decreasing an individual’s level of delinquency. In the present analyses, twohypotheses are tested:

Hypothesis 1: Strainful school-based experiences are positively related to delinquency.Hypothesis 2: School mechanisms are negatively related to delinquency and can mediate delinquency.

Methods

Data

Data for this study were originally collected as part of the National Longitudinal EducationStudy (NELS). The first wave of the NELS survey was administered to a national probabil-ity sample of students in 1988. Follow-up surveys were administered at 2-year intervals


Time 1

Delinquency• Violence• Truancy• Substance use

School Strain • Safety• Crime Exposure • Affect

Delinquency• Violence• Truancy• Substance use

School Strain • Safety• Crime Exposure • Affect

School Mechanisms • Admin. Recognition • Involvement• Atmosphere

School Mechanisms • Admin. Recognition • Involvement• Atmosphere

Time 2

No Delinquency No Delinquency

Figure 5School-Based General Strain Model as Analyzed With NELS Data



beginning in 1990. This study analyzed data from the first and second follow-up surveys,in 1990 and 1992, respectively.2 The NELS survey included items that measured a largerange of social phenomena; however, this study was primarily concerned with items thatmeasured delinquency, school-based strain, students’ affective state, and school mecha-nisms. The total sample size for both follow-ups used in this analysis is 12,144 individuals.Descriptive statistics and sample characteristics are presented in Table 1.

School strain. Two measures of school-based strain were included in this analysis. First,each respondent reported perceptions of safety while at school. Respondents indicatedhow strongly they agreed with the statement “I don’t feel safe at this school” on a 4-pointLikert-type scale. Responses were reversed from their original coding so that a higher scoreindicated greater perceptions of being unsafe increased school-based strain.


Table 1Descriptive Statistics

Mean Standard Deviation N Coding/Range

Fighting1 0.18 0.46 11,074 0 = neverFighting2 0.12 0.38 10,612 1 = once or twice

2 = more than twiceTruancy1 0.61 1.02 11,112 0 = neverTruancy2 0.97 1.31 10,585 1 = 1 to 2 times

2 = 3 to 6 times3 = 7 to 9 times4 = 10 times or more

Substances1 2.80 2.87 10,048 Reported usesSubstances2 2.94 3.12 8,605 Range = 0-19Safety1 1.63 0.69 10,903 1-4Safety2 1.67 0.72 10,602Exposure1 0.57 0.49 11,031 0 = no exposureExposure2 0.43 0.49 10,600 1 = exposureAffect1 27.77 5.96 10,026 14-56Affect2 24.92 5.67 9,380Recognition1 0.61 0.48 10,158 0 = noneRecognition2 0.63 0.48 10,227 1 = anyInvolvement1 0.81 0.39 9,504 0 = noneInvolvement2 0.81 0.38 9,461 1 = anyAtmosphere1 11.31 1.83 10,775 4-16Atmosphere2 11.45 1.90 10,537Sex 0.52 0.50 12,081 0 = male 52% Female

1 = femaleRace 0.31 0.46 12,030 0 = White 54% Non-White

1 = non-WhiteAge 15.61 0.60 11,769 YearsSES 2.54 1.12 11,809 Within sample quartiles based on a

composite score developed by theNELS authors



The second measure of school-based strain was exposure to criminal behavior. This wasmeasured through three survey items. Respondents were asked to report the frequency ofthree acts occurring during the first half of the current school year. These acts included hav-ing something stolen from them at school, having someone offer to sell them drugs atschool, and having someone threaten to hurt them at school. For this analysis, responseswere coded as either never occurring or occurring more than once. These responses werethen aggregated into an additive measure of exposure to crime and victimization while atschool. Our use of this construct and its coding is based on the expectation that any rela-tionship with this exposure (rather than the frequency of exposure) is a strainful event andcontributes to the school-based environmental strain.

Affective state. A total of 14 items were used to assess each respondent’s affective state(see appendix for a specification of each scale used in this analysis). Items were coded sothat a higher score indicates a greater degree of negative affect.

School mechanisms. School mechanisms are a construct that measures the school’s abil-ity to integrate activities that might promote coping or mediation of school-based strain.These mechanisms were divided into three separate categories. First, school recognitionwas measured with a nine-item scale that asked respondents to indicate whether they hadreceived any awards or other types of recognition. Measures of school recognition werecoded as 0 for no recognition and 1 for one or more instances of official recognition.

The second school mechanism included was perceptions of the school atmosphere. Thisvariable was measured using a four-item scale. Although there were a number of otheritems that measured the atmosphere of the school in the NELS survey, these four itemswere selected for two reasons. First, only those sentiments that could be manipulated by theschool administration or faculty were included. Other items measuring school atmospheredealt with sentiments induced by other students, not the school faculty or administration.Second, only items that were included in both the first and second follow-ups wereincluded in the analysis. These items were measured with a 4-point scale of agreement andhave been coded so that higher values indicate a more positive school atmosphere.

The third type of school mechanism included in this study was school involvement.Although the types of activities included in the school involvement scale were the same inboth survey administrations, the items were organized differently (see appendix). Forinstance, in the first follow-up survey, respondents could indicate participation in a varietyof team and individual sports (e.g., baseball, football, soccer, etc.), but in the second fol-low-up, respondents could indicate participation in either team sports and/or individualsports, that is, in the second follow-up administration of the NELS survey, several individ-ual items had been grouped together. These items have been coded as 0 to indicate noinvolvement with any activities and 1 to indicate involvement in one or more activities. It isexpected that any involvement could be just as meaningful to the predicted relationship asfrequent involvement.

Delinquency. Three measures of delinquency are analyzed. The first measure of delin-quency is violence, and is measured through a single item that reports the frequency of




involvement in physical fights at school during the first half of the current school year.Responses to this item were coded as 0 = never, 1 = once or twice, and 2 = more than twice.

The second category of delinquency measures truancy. This is also measured through asingle item that asked respondents to report the number of times they had cut or skippedclasses within the first half of the current school year. Responses to this item were codedas 0 = never, 1 = 1 to 2 times, 2 = 3 to 6 times, 3 = 7 to 9 times, and 4 = more than 10 times.

The final measure of delinquency is substance use/abuse. This was measured with fouritems that report respondents’ use of cigarettes, alcohol, marijuana, and cocaine. The itemmeasuring cigarette use asked respondents to report the number of cigarettes they usuallysmoked in a day. Responses were coded as 0 = I don’t smoke at all, 1 = less than 1 ciga-rette a day, 2 = 1 to 5 cigarettes a day, 3 = about 1/2 pack a day, 4 = more than 1/2 packa day but less than 2 packs a day, and 5 = two packs a day or more. In the first follow-upsurvey administration, alcohol, marijuana, and cocaine use were measured in terms of life-time use. In the second follow-up administration, alcohol, marijuana, and cocaine use wasmeasured in terms of previous 12 months. The responses were coded as 0 = never, 1 = 1 to2 occasions, 2 = 3 to 19 occasions, and 3 = 20+ occasions.

Results

Analyses were conducted over two stages. First, bivariate correlations were computedfor all the variables included in the analyses, and a correlation matrix is presented in Table2. In addition to this, a series of ordinary least squares (OLS) regression models were esti-mated for each measure of delinquency. In each model, the dependent variables are the var-ious forms of delinquency measured during the second follow-up administration of thesurvey. In addition, each model estimates the impact of demographic characteristics and allindependent variables measured during the first follow-up of the NELS survey. Becausethis is an exploratory study, the models were estimated in a series of steps that progressivelyinclude blocks of prior measures of school-based strain or prior measures of school mech-anism variables and then blocks of contemporaneous measures of school-based strain andschool mechanism variables. In the final full model, all measures of prior and contempora-neous strain and school mechanisms were included.

Bivariate Correlations

School mechanisms. Table 2 presents the correlation matrix. Overall, the correlationsreported here are remarkably low. The largest correlation coefficient, excluding associa-tions of the same variable measured across two waves, estimates the association of sub-stance use and truancy both measured during the first follow-up survey (r = .429).Collectively, this matrix indicates that many relationships exist at a bivariate level in theo-retically expected directions, and the small-to-moderate magnitude of these relationshipsindicates that problems with colinearity are not likely to exist in the multivariate modelsdescribed below.




125

Tabl

e 2

Cor

rela

tion

s

Var

iabl

e1

23

45

67

89

1011

1213

1415

1617

1819

2021

22

1. S

afet

y 11

2. S

afet

y 2.3

00**

13.

Fig

ht1

.113

**.0

54**

14.

Fig

ht2

.076

**.1

05**

.377

**1

5. T

ruan

cy1

.107

**.0

86**

.226

**.1

39**

16.

Tru

ancy

2.0

76**

.111

**.1

32**

.189

**.4

50**

17.

Sub

stan

ce U

se1

.074

**.0

30**

.243

**.1

57**

.429

**.2

61**

18.

Sub

stan

ce U

se2

.046

**.0

36**

.187

**.1

89**

.290

**.3

52**

.625

**1

9. E

xpos

ure 1

.150

**.1

00**

.215

**.1

23**

.165

**.1

25**

.202

**.1

53**

110

. Exp

osur

e 2.1

10**

.155

**.1

42**

.238

**.1

31**

.190

**.1

69**

.216

**.2

77**

111

. Rec

ogni

tion 1

–.07

0**

–.05

1**

–.03

5**

–.04

5**

–.15

0**

–.09

5**

–.17

6**

–.14

9**

–.02

8**

–.02

1**

112

. Rec

ogni

tion 2

–.04

8**

–.06

3**

–.05

9**

–.04

7**

–.11

4**

–.12

9**

–.12

7**

–.14

8**

–.02

9**

–.02

9**

.308

**1

13. A

ffec

t 1.2

77**

.159

**.1

02**

.052

**.1

57**

.095

**.1

74**

.138

**.1

22**

.109

**–.

154*

*–.

135*

*1

14. A

ffec

t 2.1

81**

.210

**.0

55**

.074

**.0

72**

.122

**.0

95**

.136

**.0

77**

.129

**–.

107*

*–.

146*

*.5

47**

115

. Inv

olve

men

t 1–.

098*

*–.

084*

*–.

044*

*–.

023*

–.15

6**

–.06

7**

–.13

7**

–.06

9**

–.00

6–.

018

.239

**.1

72**

–.11

7**

–.08

9**

116

. Inv

olve

men

t 2–.

079*

*–.

092*

*–.

060*

*–.

064*

*–.

142*

*–.

105*

*–.

146*

*–.

111*

*–.

007

–.03

8**

.182

**.2

67**

–.11

7**

–.09

8**

.357

**1

17. A

tmos

pher

e 1–.

278*

*–.

143*

*–.

170*

*–.

112*

*–.

236*

*–.

165*

*–.

241*

*–.

185*

*–.

163*

*–.

113*

*.1

31**

.080

**–.

259*

*–.

181*

*.1

27**

.116

**1

18. A

tmos

pher

e 2–.

171*

*–.

258*

*–.

100*

*–.

126*

*–.

134*

*–.

198*

*–.

136*

*–.

172*

*–.

118*

*–.

169*

*.0

88**

.112

**–.

188*

*–.

233*

*.0

59**

.097

**.4

23**

119

. Sex

–.02

7**

–.00

6–.

213*

*–.

174*

*–.

010

–.07

0**

–.05

2**

–.13

8**

–.11

9**

–.12

9**

.014

.054

**.0

80**

.040

**.0

15.0

39**

.013

–.00

21

20. R

ace

.085

**.1

31**

.015

.041

**.0

62**

.069

**–.

131*

*–.

158*

*–.

015

.015

.044

**–.

019

–.00

9.0

16–.

053*

*–.

054*

*.0

06–.

019

.018

*1

21. S

ES

–.10

4**

–.11

6**

–.08

6**

–.07

9**

–.06

0**

.001

–.04

1**

.018

.001

.002

.100

**.1

40**

–.13

9**

–.12

2**

.171

**.1

68**

.060

**.0

76**

–.04

3**

–.21

1**1

22. A

ge–.

068*

*–.

070*

*–.

107*

*–.

077*

*–.

043*

*–.

013

–.05

3**

–.03

9**

–.03

2**

–.02

5*.0

32**

.049

**–.

089*

*–.

065*

*.0

61**

.070

**.0

44**

.005

.107

**–.

079*

*.2

37**

1

*p<

.05

(tw

o-ta

iled)

. **p

< .0

1 (t

wo-

taile

d).



Multivariate Results

To test the causal relationships between the variables, as well as the mediating impact ofschool mechanisms on the relationship between school strain and delinquency, a series ofOLS multiple regression models were estimated. As described above, four models wereestimated for each of the three measures of delinquency included in this study (i.e., thebaseline model, school mechanisms model, school strain model, and full model).

Fighting. Table 3 presents the findings of the four OLS regression models that estimatedthe impact of school-based strain and school mechanisms on fighting. In model 1, all threemeasures of prior delinquency had a significant impact on current levels of fighting. Inaddition, previous exposure to crime and victimization at school was the only measure ofschool strain that showed a significant relationship. Finally, both prior recognition and priorperceived atmosphere were significantly and negatively related to fighting. This modelindicates that various forms of prior delinquency and exposure to delinquency and victim-ization increase subsequent delinquency, whereas administrative recognition and positiveatmosphere reduce subsequent delinquency.

In Model 2, all the baseline measures remained, and the contemporaneous school-basedstrain measures were added. In this model, all three measures of prior delinquency againhad a significant impact on subsequent fighting. Also, prior administrative recognition wasnegatively related to subsequent fighting. In this model, the relationship between prioraffective state and subsequent fighting is significant and negative. Perceptions of safety,exposure to delinquency and victimization, and negative affect also had significant and pos-itive relationships with contemporaneous levels of fighting, indicating that current strainincreases fighting, while controlling for past measures of strain and school mechanisms.

In Model 3, all the baseline measures again remained in the model, but the three mea-sures of contemporaneous school mechanisms replaced the contemporaneous measures ofschool-based strain. As in the baseline, Model 1, both prior fighting and prior truancy aresignificantly and positively related to subsequent fighting; however, the relationshipbetween prior substance use and subsequent fighting is negative. Also, the relationshipbetween exposure to crime and violence in school and subsequent fighting continued to bethe only significant relationship between the measures of school strain and fighting. In thismodel, the impact of school atmosphere on subsequent fighting was not significant; how-ever, the impact of prior administrative recognition on subsequent fighting remainedsignificant and negative. In addition, the only contemporaneous measure of school mecha-nisms to have a significant impact on fighting was school atmosphere, and this was in theexpected direction of reducing fighting.

Finally, the full model included all the baseline measures and all the contemporaneousschool mechanisms and school strain measures. In this model, the impact of prior delinquencyand prior school-based strain is assessed simultaneously with the measures of both contem-poraneous school-based strain and contemporaneous school mechanisms. The impact of priordelinquency on subsequent fighting remained significant and positive. The impact of prioraffect and recognition also remained significant and negatively related to fighting, but allother measures of prior strain and school mechanisms are not significant. The only contem-poraneous measures of school mechanisms to show a significant impact on fighting wasperceived school atmosphere, and this was in the expected negative direction.




127

Mod

el 1

Mod

el 2

Mod

el 3

Fu

ll M

odel

Stan

dard

St

anda

rdSt

anda

rdSt

anda

rdb

Err

orβ

bE

rror

βb

Err

orβ

bE

rror

β

Sex

–.06

2***

.008

–.09

0–.

050*

**.0

08–.

075

–.06

3***

.009

–.09

3–.

053*

**.0

09–.

079

Rac

e.0

23**

*.0

09.0

30.0

17*

.009

.022

.023

**.0

09.0

31.0

17*

.010

.022

Age

.012

*.0

08.0

18.0

08.0

08.0

13.0

08.0

08.0

12–.

009

.008

.013

SES

–.00

8**

.004

–.02

4–.

004

.004

–.01

5–.

008*

*.0

04–.

026

–.00

5.0

04–.

017

Figh

ting 1

.228

***

.010

.268

.218

***

.011

.260

.221

***

.011

.263

.214

***

.011

.254

Tru

ancy

1.0

16**

*.0

05.0

44.0

16**

*.0

05.0

44.0

17**

*.0

05.0

48.0

16**

*.0

05.0

45Su

bsta

nces

1.0

08**

*.0

02.0

66.0

06**

*.0

02.0

51–.

008*

**.0

02.0

62.0

07**

*.0

02.0

52Sa

fety

1.0

06.0

06.0

12–.

005

.007

–.00

9.0

05.0

07.0

09–.

005

.007

–.00

9E

xpos

ure 1

.021

**.0

08.0

30–.

007

.009

–.01

0.0

21**

.009

.031

–.00

1.0

09–.

002

Aff

ect 1

.000

.001

–.00

3–.

002*

*.0

01–.

043

–.00

1.0

01–.

012

–.00

3***

.001

–.05

2R

ecog

nitio

n 1–.

022*

**.0

08–.

031

–.01

8**

.009

–.02

6–.

016*

.009

–.02

4–.

016*

.009

–.02

3In

volv

emen

t 1.0

13.0

11.0

14.0

14.0

11.0

15.0

12.0

12.0

13.0

16.0

12.0

17A

tmos

pher

e 1–.

006*

*.0

02–.

029

–.00

3.0

02–.

016

–.00

1.0

03–.

008

–.00

2.0

03–.

009

Safe

ty2

.029

***

.006

.060

.026

***

.007

.053

Exp

osur

e 2.1

00**

*.0

09.1

48.0

96**

*.0

09.1

43A

ffec

t 2.0

03**

.001

.043

.003

**.0

01.0

49R

ecog

nitio

n 2.0

02.0

09.0

03.0

04.0

09.0

06In

volv

emen

t 2–.

008

.012

–.00

8–.

007

.012

–.00

8A

tmos

pher

e 2–.

010*

**.0

02–.

057

–.00

5**

.003

–.02

8C

onst

ant

.000

.135

–.05

1.1

38.1

52.1

42–.

001

.145

R2

.132

.154

.133

.154

F78

.524

***

69.2

21**

*57

.027

***

53.9

68**

*df

1316

1619

*p<

.10.

**p

< .0

5. *

**p

< .0

1.

Tabl

e 3

OL

S R

egre

ssio

n of

Fig

htin

g



Truancy. Table 4 presents the findings of the four OLS models that include truancy as thedependent variable. For Model 1, both prior truancy and substance use/abuse had a signif-icant impact on subsequent truancy. As in the previous models for fighting, prior exposureto crime and violence at school was the only measure of school-based strain that had asignificant impact on increasing truancy. Also, prior administrative recognition and priorperceptions of school atmosphere had a significant impact on reducing truancy.

The results for Model 2 indicate that the relationship between prior delinquency (specifi-cally, truancy and substance use) and subsequent truancy remained consistently significantand positive in this model as well. The impact of prior affective state on subsequent truancyis significant and negative, but the impact of negative affect on truancy is positive and ofgreater magnitude. The other measures of strain, feelings of safety and exposure to violenceand crime at school, also were significant and positive indicating increases in truancy.Perceived atmosphere had a significant and negative impact on subsequent levels of truancy.

In Model 3, prior truancy and substance use are again significantly and positively relatedto subsequent truancy. Also, the impact of prior administrative recognition on current truancyremained significant and negative. Unlike Model 3 for fighting, a significant and positive rela-tionship between prior school involvement and subsequent truancy was found in this model.Finally, in this model, all three measures of contemporaneous school mechanisms (adminis-trative recognition, involvement, and atmosphere) were significantly and negatively related tocurrent truancy indicating that school mechanisms can reduce contemporaneous truancy.

In the full model, prior truancy and substance use again had a significant impact on subse-quent truancy. Also, the impact of prior affect on subsequent truancy is again significantly andnegatively related to truancy, as it was in Model 2. Prior and current administrative recognitionand current school atmosphere have a significant negative impact on truancy in the full model,but prior involvement in school activities has a significant positive impact. More importantly,the measures of school-based strain, (perceptions of safety, exposure to crime and violence,and affect) all showed a consistent significant and positive impact on current levels of truancy.

Substance use. Table 5 presents the findings of the four OLS models that included substanceuse as the dependent variable. The findings from Model 1 show that both prior truancy andsubstance use have a significant positive impact on subsequent substance use. Unlike thefirst model for fighting and truancy, the estimates for this model show that prior negativeaffect has a significant positive impact on subsequent substance use. This finding is notunusual considering that general strain predicts that substance use can be a form of copingwith strain. Similar to the findings from Model 1 for truancy, both prior administrativerecognition and prior perceived school atmosphere showed a significant negative impact oncurrent levels of substance use.

In Model 2, the measures of current school strain were included, and the estimates forthe impact of prior truancy and prior substance use on subsequent substance use remainedsignificant and positive. Counter to what could be anticipated, both prior feelings of beingunsafe and prior negative affect showed a significant negative impact on subsequent substanceuse. As expected, prior perceived atmosphere was significantly and negatively related tosubsequent substance use. Finally, two measures of school-based strain, exposure to crimeand violence at school and affect, showed a significant and positive impact on contempo-raneous levels of substance use.




129

Tabl

e 4

OL

S R

egre

ssio

n of

Tru

ancy

and

Tar

dine

ss

Mod

el 1

Mod

el 2

Mod

el 3

Fu

ll M

odel

Stan

dard

Stan

dard

St

anda

rdSt

anda

rdb

Err

orβ

bE

rror

βb

Err

orβ

bE

rror

β

Sex

–.14

3***

.029

–.05

5–.

109*

**.0

30–.

043

–.14

3***

.030

–.05

6–.

111*

**.0

31–.

043

Rac

e.2

09**

*.0

32.0

72.2

22**

*.0

33.0

77.2

25**

*.0

33.0

78.2

18**

*.0

34.0

75A

ge–.

022

.027

–.00

9–.

023

.028

–.00

9–.

022

.028

–.00

9–.

021

.029

–.00

8SE

S.0

41**

*.0

13.0

36.0

49**

*.0

13.0

42.0

57**

*.0

14.0

49.0

61**

*.0

14.0

53Fi

ghtin

g 1–.

009

.037

–.00

3.0

11.0

38.0

03.0

21.0

39.0

06.0

20.0

40.0

06T

ruan

cy1

.534

***

.016

.393

.524

***

.017

.389

.521

***

.017

.387

.507

***

.018

.376

Subs

tanc

es1

.047

***

.006

.099

.044

***

.006

.093

.042

***

.006

.089

.042

***

.006

.089

Safe

ty1

.021

.022

.011

–.02

4.0

24–.

013

.013

.023

.007

–.01

2.0

25–.

006

Exp

osur

e 1.0

75**

.029

.029

.007

.030

.003

.059

**.0

30.0

23.0

08.0

31.0

03A

ffec

t 1.0

04.0

03.0

18–.

007*

*.0

03–.

034

.001

.003

.006

–.00

8***

.003

–.04

0R

ecog

nitio

n 1–.

120*

**.0

30–.

045

–.11

6***

.030

–.04

4–.

076*

*.0

32–.

029

–.07

6**

.032

–.02

9In

volv

emen

t 1.0

35.0

39.0

10.0

46.0

40.0

13.0

82*

.043

.024

.090

**.0

44.0

26A

tmos

pher

e 1–.

031*

**.0

08–.

042

–.02

4**

.009

–.03

4.0

02.0

09.0

02.0

00.0

10.0

00Sa

fety

2.0

88**

*.0

22.0

48.0

55**

.023

.030

Exp

osur

e 2.2

71**

*.0

30.1

05.2

57**

*.0

32.1

00A

ffec

t 2.0

19**

*.0

03.0

86.0

16**

*.0

03.0

71R

ecog

nitio

n 2–.

171*

**.0

33–.

064

–.17

2***

.034

–.06

4In

volv

emen

t 2–.

078*

.043

–.02

2–.

067

.044

–.01

9A

tmos

pher

e 2–.

077*

**.0

09–.

113

–.06

3***

.009

–.09

2C

onst

ant

1.15

6**

1.98

4.7

07.4

911.

825*

**.4

991.

351*

**.5

14R

2.2

36.2

60.2

56.2

69F

160.

02**

*13

3.86

***

128.

33**

*10

8.67

***

df13

1616

19

*p <

.10.

**p

< .0

5. *

**p

< .0

1.



130

Tabl

e 5

OL

S R

egre

ssio

n of

Sub

stan

ces

Mod

el 1

Mod

el 2

Mod

el 3

Fu

ll M

odel

Stan

dard

Stan

dard

Stan

dard

Stan

dard

bE

rror

βb

Err

orβ

bE

rror

βb

Err

orβ

Sex

–.52

2***

.065

–.08

4–.

467*

**.0

66–.

075

–.49

5***

.067

–.08

0–.

443*

**.0

68–.

072

Rac

e–.

494*

**.0

73–.

070

–.54

6***

.074

–.07

7–.

520*

**.0

76–.

073

–.54

9***

.077

–.07

7A

ge.0

10.0

62–.

002

.012

.063

.002

.036

.064

.006

.028

.065

.005

SES

.087

**.0

29.0

31.0

98**

.030

.035

.104

***

.031

.037

.108

***

.031

.038

Figh

ting 1

.012

.084

.001

–.03

7.0

86–.

005

–.01

2.0

88–.

002

–.05

0.0

89–.

006

Tru

ancy

1.1

19**

*.0

37.0

36.1

01**

*.0

38.0

31.1

16**

*.0

39.0

35.0

99**

.039

.030

Subs

tanc

es1

.683

***

.013

.593

.675

***

.014

.585

.672

***

.014

.582

.664

***

.014

.575

Safe

ty1

–.06

7.0

51–.

014

–.11

4**

.053

–.02

4–.

067

.053

–.01

4–.

089*

.055

–.01

9E

xpos

ure 1

.081

.065

.013

–.06

2.0

67–.

010

.076

.068

.012

–.06

6.0

70–.

011

Aff

ect 1

.016

**.0

06.0

30–.

016*

*.0

07–.

030

.008

.006

.015

–.01

9***

.007

–.03

6R

ecog

nitio

n 1–.

356*

**.0

67–.

056

–.35

5***

.067

–.05

6–.

292*

**.0

72–.

046

–.31

4***

.072

–.04

9In

volv

emen

t 1.0

43.0

89.0

05.0

27.0

90.0

03.0

59.0

96.0

07.0

57.0

97.0

07A

tmos

pher

e 1–.

048*

*.0

19–.

028

–.04

4**

.019

–.02

6–.

001

.021

–.00

1–.

006

.022

–.00

3Sa

fety

2.0

51.0

49.0

11–.

013

.052

–.00

3E

xpos

ure 2

.616

***

.068

.099

.648

***

.070

.104

Aff

ect 2

.052

***

.007

.099

.047

***

.007

.087

Rec

ogni

tion 2

–.29

6***

.074

–.04

6–.

235*

**.0

75–.

036

Invo

lvem

ent 2

–.06

5.0

98–.

008

–.08

0.0

98–.

009

Atm

osph

ere 2

–.13

1***

.019

–.07

9–.

101*

**.0

20–.

061

Con

stan

t1.

913*

1.08

61.

200

1.09

92.

848*

*1.

135

2.04

3*1.

151

R2

.431

.445

.432

.446

F33

2.59

***

273.

79**

*25

0.47

***

215.

67**

*df

1316

1619

*p <

.10.

**p

< .0

5.**

*p <

.01.



In Model 3, both prior truancy and prior substance use continued to show a significantpositive impact on subsequent substance use. Also, prior administrative recognition continuedto show a significant negative impact on subsequent substance use. Only two of the threemeasures of school mechanisms showed a significant contemporaneous relationship withsubstance use, administrative recognition, and perceived school atmosphere.

In the final full model, the impact of prior truancy and substance use continued to besignificantly and positively related to subsequent substance use. The impact of prior feel-ings of safety and prior negative affect are significantly and negatively related to subsequentsubstance use and counter to what would be predicted by strain theory. Consistent with theother models for substance use, prior administrative recognition showed a significantnegative impact on substance use in the full model. No other measure of prior school mech-anisms was significant, but both administrative recognition and school atmosphere havesignificant, negative contemporaneous relationships with substance use. In this full model,exposure to violence and crime at school and negative affect display significant and posi-tive contemporaneous relationships with substance use, whereas administrative recognitionand school atmosphere are significantly and negatively related, indicating that school activ-ities might be able to contribute to reducing substance use among students.

Discussion

This study focused on two main hypotheses. First, it was expected that strainful school-based experiences would be positively related to various forms of delinquency. These multi-variate models have indicated partial support for this hypothesis. Exposure to violence andcrime while at school was consistently related to higher contemporaneous levels of fighting,truancy, and substance use, while controlling for other measures of strain such as a moregeneral negative affect. Feeling unsafe was also significantly related to higher contempora-neous levels of fighting and truancy, although it was not significantly related to substance use.

The second hypothesis considered the relationship between school mechanisms and delin-quency. It was expected that certain school mechanisms (i.e., administrative recognition,involvement, and atmosphere) would reduce involvement in delinquent acts. In the fullestmodels, school atmosphere was negatively related to each measure of delinquency, but the onlyother school-based measure related to lower levels of delinquency was administrative recogni-tion, and this relationship was limited to significant reductions in truancy and substance use.Contrary to expectations, involvement in school-based activities did not reduce participation inthe measures of delinquency considered here, but it is apparent that certain school experiencescan act to reduce delinquency, whereas others can act to increase delinquency.

These findings do offer some important information for school policy. They suggest thatschools can decrease the involvement in delinquency among students, first, by producing amore positive atmosphere and promoting recognition of those students who are committedto or doing well in the school setting. This expectation is consistent with evaluations thathave indicated some success with promoting positive atmospheres in schools to controldelinquency and other problem behaviors (Gottfredson, 1986). Furthermore, expectingstudents to simply find opportunities for positive coping through school activities is not aneffective approach to controlling delinquency. Schools should develop more direct ways to




decrease levels of school strain and create a more positive affective state among students.The traditional model of passively encouraging participation may not be effective.

This research has been conducted with the recognition that further analyses shouldexplicitly consider the dynamic processes that exist within schools. These processes cancontribute to and control delinquent behavior. This research has tried to specify some ofthese dynamic relationships, but it is likely others exist. An additional attribute to thisresearch is that it has been done longitudinally at a time in adolescents’ lives when schoolis an important contributor to social development. It is possible that these relationshipsmight not exist in the same way at different times in adolescents’ lives, and criminologistsand school administrators could benefit by examining different age groups or school gradesthan what was considered here. Also, these analyses are based on a 2-year gap betweensurvey administrations. Although this is more temporally restricted than other past investi-gations of general strain theory among adolescents, it is more open than others. Becauseadolescents develop at such a quick pace, it is possible that these analyses have not com-pletely captured the nuances of the relationship between school-based strain, schoolmechanisms, and delinquency. In addition, we have included one general measure of negativeaffect, but other measures of negative affect and other contributors to strain should furtherspecify the intricate relationship that likely exists with crime and delinquency. Futureresearchers should continue to refine these measures of strain and the social mechanismsthat might mediate the impact of strain on delinquent and criminal behaviors.

Appendix Scale Items

Affective State• I feel good about myself• I do not have enough control over the direction my life is taking (R)• In my life, good luck is more important than hard work for success (R)• I feel I am a person of worth, the equal of other people• I am able to do things as well as most other people (R)• Every time I try to get ahead, something or somebody stops me (R)• My plans hardly ever work out, so planning only makes me unhappy (R)• On the whole, I am satisfied with myself• I feel useless at times (R)• At times, I think I am no good at all (R)• When I make plans, I am almost certain I can make them work• I feel I do not have much to be proud of (R)• Chance and luck are very important for what happens in my life (R)• I feel emotionally empty most of the time (R)

1 = strongly agree; 2 = agree; 3 = disagree; 4 = strongly disagreeHigher values indicate more negative affective stateFollow-up 1: Range = 14-56, α = .8652Follow-up 2: Range = 13-52, α = .8615

(continued)




Appendix (continued)

School Recognition• Elected officer of a school class• Won an academic honor• Participated in science or math fair• Received special recognition for good attendance• Received special recognition for good grades or honor roll• Received special recognition for writing an essay or poem• Named most valuable player on a sports team• Received a community service award• Participated in vocational/technical skills competition

0 = none; 1 = anyFollow-up 1: α = .5563Follow-up 2: α = .5424

School Atmosphere• There is real school spirit• The teaching is good• Discipline is fair• Teachers are interested in students

1 = strongly agree; 2 = agree; 3 = disagree; 4 = strongly disagreeHigher values indicate a more positive school atmosphereRange = 4-16Follow-up 1: α = .5922Follow-up 2: α = .6520

School Involvement

Follow-up 1:

• Baseball/softball• Soccer• Football• Basketball• Other team sport• Swim team• Other individual sport• Cheerleading• Pom-pom, drill team• Band, orchestra, chorus, choir, or other music group• School play or musical• Student government• NHS or other academic honor society• School yearbook, newspaper, or literary magazine• Service clubs

(continued)




Appendix (continued)

• Academic clubs• Hobby clubs• FTA, FHA, FFA or other vocation, education, or professional club

0 = none; 1 = anyα = .5564

Follow-up 2:

• A team sport (baseball, basketball, football, soccer, hockey, etc.)• An individual sport (cross-country, gymnastics, golf, tennis, track, wrestling, etc.)• Cheerleading, Pom-pom, drill team• Band, orchestra, chorus, choir, or other music group• Drama club, school play, or musical• Student government• NHS or other academic honor society• School yearbook, newspaper, or literary magazine• Service clubs• Academic clubs• Hobby clubs• FTA, FHA, FFA or other vocation, education, or professional club• An intramural team sport• An intramural individual sport

0 = none; 1 = anyα = .5755

Notes

1. For a more complete review of the history and empirical status of strain theories, see Akers and Sellers,(2004) or Vold, Bernard, and Snipes (2002).

2. By analyzing these waves of data, we restrict the analysis to the years where respondents are most likelyto be in high school grades; this restriction removes the likelihood that the transition from middle or junior highschool to high school contributes to any student’s amount of strain. Any school-based strain or school-basedmechanism for controlling strain is limited to what is experienced within one school.

References

Agnew, R. (1985). A revised strain theory of delinquency. Social Forces, 64, 151-167.Agnew, R. (1992). Foundation for a general strain theory of crime and delinquency. Criminology, 30, 47-88.Agnew, R. (1995). Controlling delinquency: Recommendations from general strain theory. In H. D. Barlow

(Ed.), Crime and public policy: Putting theory to work (pp. 43-70). Boulder, CO: Westview.Agnew, R. (2000). Strain theory and school crime. In S. S. Simpson (Ed.), Of crime and criminality: The use

of theory in everyday life (pp. 105-120). Boston: Pine Forge Press.Agnew, R. (2001). Building on the foundation of general strain theory: Specifying the types of strain most likely

to lead to crime and delinquency. Journal of Research in Crime and Delinquency, 38, 319-361.Agnew, R., & White, H. (1992). An empirical test of general strain theory. Criminology, 30, 475-500.




Akers, R. (1977). Deviant behavior: A social learning approach. Belmont, CA: Wadsworth.Akers, R., & Sellers, C. (2004). Criminological theories: Introduction, evaluation, and application. Los Angeles:

Roxbury.Brezina, T. (1996). Adapting to strain: An examination of delinquent coping responses. Criminology, 34, 39-60.Broidy, L. (2001). A test of general strain theory. Criminology, 39, 9-35.Broidy, L., & Agnew, R. (1997). Gender and crime: A general strain theory perspective. Journal of Research in

Crime and Delinquency, 34, 275-306.Cloward, R., & Ohlin, L. (1960). Delinquency and opportunity. Glencoe, IL: Free Press.Cohen, A. K. (1955). Delinquent boys. Glencoe, IL: Free Press.DeVoe, J. F., Peter, K., Noonan, M., Snyder, T. D., & Baum, K. (2005). Indicators of school crime and safety:

2005 (NCES 2006-001/NCJ 210697). U.S. Departments of Education and Justice. Washington, DC: U.S.Government Printing Office.

Elliott, D., Huizinga, D., & Ageton, S. (1985). Explaining delinquency and drug use. Beverly Hills, CA: Sage.Gottfredson, D. (1986). An empirical test of school-based environmental and individual interventions to reduce

the risk of delinquent behavior. Criminology, 24, 705-731.Hirschi, T. (1969). Causes of delinquency. Berkeley: University of California Press.Kaplan, H. (1972). Toward a general theory of psychosocial deviance: The case of aggressive behavior. Social

Science and Medicine, 6, 593-617.Kaplan, H. (1975). Self-attitudes and deviant behavior. Pacific Palisades, CA: Goodyear.Kaplan, H. (1980). Deviant behavior in defense of self. New York: Academic Press.Kaplan, H. (1984). Patterns of juvenile delinquency. Beverly Hills, CA: Sage.Kaplan, H., & Damphouse, K. (1997). Negative social sanctions, self-derogation, and deviant behavior: Main

and interactive effects in longitudinal perspective. Deviant Behavior, 18, 1-26.Kaplan, H., & Johnson, R. (2001). Social deviance: Testing a general theory. New York: Kluwer Academic.Kaplan, H., Johnson, R., & Bailey, C. (1986). Self-rejection and the explanation of deviance: Refinement and

elaboration of a latent structure. Social Psychological Quarterly, 49, 110-128.Kaplan, H., & Peck, B. (1992). Self-rejection, coping style, and mode of deviant response. Social Science

Quarterly, 73, 903-919.Mazerolle, P., & Piquero, A. 1997. Violent Responses to Strain: An Examination of Conditioning Influences.

Violence and Victims, 12, 323-343.Mazerolle, P., & Piquero, A. (1998). Linking exposure to strain with anger: An investigation of deviant adaptations.

Journal of Criminal Justice, 26, 195-211.Merton, R. K. (1938). Social structure and anomie. American Sociological Review, 3, 672-682.Paternoster, R., & Mazerolle, P. (1994). General strain theory and delinquency: A replication and extension.

Journal of research in crime and delinquency, 31, 235-263.Piquero, N. L., & Sealock, M. (2000). Generalizing general strain: An examination of an offending population.

Justice Quarterly, 17, 449-484.Stiles, B., Liu, X., and Kaplan, H. (2000). Relative deprivation and deviant adaptations: The mediating effects of

negative self-feelings. Journal of Research in Crime and Delinquency, 37, 64-90.Sutherland, E. (1947). Principles of criminology (4th ed.). Philadelphia: J.B. Lippincott.U.S. Department of Education, National Center for Education Statistics. (2003). In A. K. Miller (Ed.), Violence

in U.S. Public Schools: 2000 school survey on crime and safety (NCES 2004-314). Washington, DC:National Center for Education Statistics.

Vold, G., Bernard, T., & Snipes, J. (2002). Theoretical criminology (5th ed.). New York: Oxford University Press.

Daniel R. Lee, PhD, is an assistant professor and Master of Arts program coordinator in the criminology department atIndiana University of Pennsylvania. His research interests include the measurement and validity of criminological theory,assessing the fear of crime and its impact on behavior and attitudes, and evaluating criminal justice policies.

Jeffrey W. Cohen is a doctoral candidate and temporary faculty member in the criminology department at Indiana Universityof Pennsylvania. He is currently completing his dissertation research that examines the measurement and conceptualizationof gender across social science disciplines. His recent research has been published in the Journal of Men’s Studies.




Strain and Schools

Documents

Transcript of Strain and Schools