m.
%

I

{.r ..
a)“. 0»th

.

u. .m....
nmmww

o
a

.fi
1

5
£35 II).
. .3...
Jury».
‘ .l...
hunk...
. .1
.

gnu“;

.u. 5.31.".

3: Knit}... .
.7 .

 

 

 

 

T} i {j f«

This is to certify that the

thesis entitled

An Application of General Strain Theory
to Korean Adolescents: A Test of the Relation between
Negative Stimuli at School and Delinquency

presented by

Jung-Mi Kim

has been accepted towards fulfillment
of the requirements for

M.S. degree in Criminal Justice

 

 

7/4», 75/
/ mm professor

Date September 2, 2002

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

 

 

 

 

LIBRARY

Michigan State

University

 

 

PLACE IN RETURN Box to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

ML 1 7 2004

 

312104

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 cJCIRC/Dateouopss-ots

.._.__. -_......__-.-.__-.._.—--—-——-— -.--——-—...- -—-—-

An Application of General Strain Theory to Korean Adolescents: A Test of the
Relation between Negative Stimuli at School and Delinquency

BY

Jung-Mi Kim

ATHESIS

Submitted to
Michigan State University
In partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Criminal Justice

2002

ABSTRACT

An Application of General Strain Theory to Korean Adolescents: A Test of the
Relation between Negative Stimuli at School and Delinquency

By

Jung-Mi Kim

This study is designed to test theory to explain the phenomenon of
delinquency in South Korea. To better understand the causation of
delinquency, this study will examine the relationship of students’ negative
experiences related to school with their delinquent behavior. Based on the
general strain theory (GST), this study focuses on the higher-education
oriented school culture as a major source of strain on youths in Korean society.
Hypotheses will be tested with data collected by the Korean National Institute
of Criminology (KIC) from a city-wide sample of adolescents. To collect this
data, the KIC conducted a survey of 1919 middle and high school students in

40 classes within 20 schools that were located in the Seoul area.

Copyright by
Jung-Mi Kim

2002

To Him

iv

ACKNOWLEDGEMENT

I wish to thank the fine faculty of School of Criminal Justice at Michigan
State University for assisting me in the realization of this project. Especially, I am
grateful to Dr. Merry Morash for her impressive support and guidance. Her
guidance and assistance made this research possible. I also would like to thank
the thesis committee members, Dr. Vincent Hoffman and Dr. Sheila Maxwell, for
their advice and being always willing to help.

I extend special thanks to Dr. Byoung-ln Cho and the Korean Institute of
Criminology for their support in providing the data used in this research. Also, I
sincerely would like to thank my best friend, Chang-Hun Lee, who discussed this
study with me many times.

Without the support and encouragement of my family this research could
not have been completed. Especially, I am grateful to my parents. They provided

invaluable love and support.

TABLE OF CONTENTS

LIST OF TABLES ................................................................................... x
LIST OF FIGURES ................................................................................ xii
CHAPTER1

INTRODUCTION ................................................................................... 1
CHAPTER 2

LITERATURE REVIEW .......................................................................... 3

I. Delinquent Youths in Korea and Current Circumstances ......................... 3

II. Theoretical Background ................................................................... 5

1. Overview of Strain Theory ............................................................ 6

2. The Basic Idea of GST ................................................................. 7

3. School as a major source of Strain (Negative Stimuli) ...................... 10

(1 ) School Dissatisfaction ......................................................... 10

(2) Negative Teacher Relationships ............................................. 11

(3) Low Grades ....................................................................... 11

(4) Negative peer relationships ................................................... 11

4. Other Sources of Strain - Loss of Positively Valued Stimuli and Failure

to Achieve Positively Valued Goals .............................................. 13
(1) Strain as the loss of positively valued stimuli ............................. 13
(2) Strain as the Failure to achieve positively valued goals ............... 14
Ill. The Korean Context and Strain Theory ................................................. 14

1. The Emphasis on Higher Education: The Confucianism Explanation...15
2. External and Internal Pressure from Parents and Self-Expectation
as Negative Stimuli ................................................................... 16

3. Parent’s socioeconomic status and negative stimuli ........................ 17

Vi

IV. Moderating Variables ........................................................................ 18

1. Social Support: Positive Adaptation to Strain (Seeking Help) ............ 19
2. Gender Differences in Delinquency .............................................. 19
CHAPER 3

RESEARCH METHODS ........................................................................ 21

I. Data Collection Procedure and Sample .................................................. 21

1. Data Collection Procedure ......................................................... 21

2. Sample .................................................................................. 21

II. Hypotheses .................................................................................... 22

III. Operationalized Variables .................................................................. 23

1. Independent Variables ............................................................... 23

(1) School dissatisfaction ......................................................... 23

(2) Negative Teacher Relationships ........................................... 24

(3) Low Grades ...................................................................... 24

(4) Negative peer relationships ................................................. 24

(5) Internal and External Pressure to achieve academically ............. 24

(6) Parental socioeconomic status ............................................. 25

2. Dependent Variable .................................................................. 25

3. Moderating Variables: Social support ........................................... 26

IV. Data Analysis Procedure ................................................................... 26
CHAPTER 4

DATA ANALYSIS AND FINDINGS ........................................................... 28

I. Reliability / Factor Analysis and Scales ................................................... 28

1. Independent Variables ............................................................... 28

vii

2. Dependent Variable .................................................................. 30

II. Univariate Analysis ........................................................................... 34
1. Demographic Characteristics of Sample ....................................... 34

2. General Characteristics of Sample .............................................. 36

III. Bivariate Analysis ............................................................................. 37
1. School Dissatisfaction and Delinquency ....................................... 37

2. Negative Teacher Relations and Delinquency ................................ 37

3. Grades and Delinquency ........................................................... 38

4. Negative Peer Relations and Delinquency .................................... 38

5. Internal pressure and Delinquency .............................................. 39

6. External pressure and Delinquency ............................................. 39

7. Parents’ Social Status and Delinquency ....................................... 39

8. Social Support and Delinquency ................................................. 40

IV. Multivariate Analysis ......................................................................... 43
1. Multivariate Model Regression .................................................... 43

(1) Regression Model for General Delinquency ............................. 43

(2) Regression Model for Minor/ Serious Delinquency .................... 44

2. Two-way Interaction Analysis between Independent Variables and

Social Support Variables ............................................................ 46
(1) Two-way interaction with the dissatisfaction with family relation....46
(2) Two-way interaction with living with both parents ....................... 49
(3) Significance of the interactions .............................................. 51
(4) The nature of the interactions ................................................ 52

3. Multivariate Model for Each Gender Group — Subgroup Analysis ....... 56
(1) Sub-group Regression Model for General Delinquency .............. 56
(2) Sub-group Regression Model for Minor Delinquency ................. 58

viii

(3) Sub-group Regression Model for Serious Delinquency .............. 59

4. Two-way Interaction Analysis between Independent Variables and

Social Support Variables for Each Gender Group ............................ 61

(1) Two-way interaction with the dissatisfaction with family relation... .61

(2) The nature of the Interactions ................................................ 64

(3) Two-way interaction with living with both parents ....................... 65

(4) The nature of the Interactions ................................................ 68
CHAPTER 5

DISCUSSION AND CONCLUSION .......................................................... 70

I. Summary and Discussion .................................................................... 70

1. Multivariate Analysis ................................................................. 7O

2. Interaction Effects .................................................................... 73

3. Gender Differences .................................................................. 73

ll. Limitations and Recommendations ....................................................... 75

III. Conclusion and Policy Implications ...................................................... 76

APPENDICES ..................................................................................... 79

Appendix A. Scatterplots ............................................................... 80

Appendix B. Correlation Matrix for All Independent Variables ................ 83

BIBLIOGRAPHY .................................................................................. 84

Bibliography ............................................................................... 85

ix

LIST OF TABLES

Table 1. Distribution of the subject per types of school ................................. 22
Table 2. Reliability Coefficients for Scales Used .......................................... 29
Table 3. Scales Used / Variable Names ..................................................... 29
Table 4. Factor Scores for Delinquency Items Based on Promax-Rotated Two-
Factor Solution ........................................................................ 31
Table 5. Reliability Coefficients for Delinquency Scales Used (Alpha) ............. 32
Table 6. Scales Used /Variable Names ..................................................... 33
Table 7. Demographic Characteristics of sample ......................................... 35
Table 8. General Characteristics of Sample ................................................ 36
Table 9. Correlation .............................................................................. 41
Table 10. Regression Analysis results ...................................................... 45

Table 11. Multivariable Analysis for Two-way Interactions between the

Independent Variables and the Dissatisfaction with Family

Relation ................................................................................ 49
Table 12. Multivariable Analysis for Two-way Interactions between the

Independent Variables and the Living with Both Parents ................. 51
Table 13. Subgroup Regression Analysis Results for General

Delinquency ........................................................................... 57
Table 14. Sub-group Regression Analysis Results for Minor

Delinquency ........................................................................... 59
Table 15. Sub-group Regression Analysis Results for Serious

' Delinquency ........................................................................... 60

Table 16. Multivariable Analysis for Two-way Interactions of the Dissatisfaction

with Family Relation in the Female Subgroup Analysis .................... 62
Table 17. Multivariable Analysis for Two-way Interactions of the Dissatisfaction

with Family Relation in the Male Subgroup Analysis ....................... 63
Table 18. Nature of Interaction: Dissatisfaction with Family Relation in Each

Gender Group ........................................................................ 65

Table 19. Multivariable Analysis for Two-way Interactions of the Living with Both
Parents in the Female Subgroup Analysis .................................... 66
Table 20. Multivariable Analysis for Two-way Interactions of the Living with Both
Parents in the Male Subgroup Analysis ....................................... 67
Table 21. Nature of Interaction: Living with Both Parents in Each Gender
Group ................................................................................... 69

xi

LIST OF FIGURES

Figure 1. Model of This Study ................................................................... 9
Figure 2. Independent and Dependent Variables ......................................... 18
Figure 3. Moderating Variables — Help Seeking Behavior .............................. 19
Figure 4. Moderating Variables — Gender Difference .................................... 20
Figure 5. The Equation of the Restricted Multivariate Analysis ....................... 47

Figure 6. The Equation of the Two-way Interaction Analysis between the
Independent Variables and the Dissatisfaction with Family
Relations .............................................................................. 47
Figure 7. The Equation of the Two-way Interaction Analysis between the

Independent Variables and the Living with Both Parents .................. 50
Figure 8. Nature of Interaction: Grades 8 Dissatisfaction with Family
3 Relation ................................................................................ 53
Figure 9. Nature of Interaction: External Pressure & Dissatisfaction with Family
Relations .............................................................................. 54
Figure 10. Nature of Interaction: Grade & Living with Both Parents ................. 55
xii

CHAPTER 1
INTRODUCTION

In the United States, juvenile crime and delinquency has constantly been
a public concern, and has received a great deal of research and media attention
in recent years. A rising amount of crime and violence in the United States has
caught youth in its expanding wake. Youth are intimately acquainted with violence
as both perpetrators and victims. Crime statistics show that youth are responsible
for approximately 29 percent of all crime (FBI, 1992).

Similarly, in Korea incidents of juvenile crime and delinquency primarily by
middle and high school students have become increasingly common, and there
have been traumatic impacts on victims (Kim & Kim, 1996). In addition, as
juvenile crime and delinquency become more serious, several tragic incidents
have happened recently in Korea.

For instances, three students in Seoul were sentenced to up to two years
in juvenile jail for burning a classmate with cigarettes and jabbing him with sharp
objects. The tormentors said they were just having fun with the victim because he
looked “timid” in gym class (800 & Kim, 2000). More recently in 2000, a 16 years
old teenaged boy in Seoul committed suicide. He was one of the academically
distinguished students and a famous swimmer in his middle school, but when he
moved up to a new high school, he was faced with a problematic situation.
Several boys started to bully him, and then they were joined by other classmates.
They repeatedly extorted money and forced him to carry their school bags. They

told police that they picked on him because of his high grades and his

“excellence” (Han, '2000). Without doubt this is a crime not only against a person,
but also a threat overall to youths in Korea.

The Korean Institute for Youth Development found that in the past five
years juvenile crime and minor type of delinquency has constantly increased, and
more than 60 percent of cases were committed by students (KIYD, 2001 ). In
addition, Korean national police agency statistics show that more than 70 percent
of drug offenses were committed by youths (Korean national police agency,
1998)

Although much effort has been invested in keeping youths safe and in
explaining delinquency in Korea, the amount of juvenile delinquency is still
increasing. The basic question, “why do youths commit delinquent behaviors?’ is
still under investigation by academics and practitioners alike. However, so far,
unfortunately there is a lack of empirical cumulated knowledge on delinquency in
Korea. To better understand the causation of delinquency in contemporary Korea,
general strain theory, which explains that school related strain could be a major
cause of delinquency, will be applied in this study, and a previously collected
nation-wide data set will be analyzed to provide evidence of a causal relationship
between strain and delinquency. In addition, since there has been a lack of
academic research on the problems of youth in Korea, this study will make a
contribution to the accumulation of knowledge to further understand possible

reasons for Korean youths’ violent and other delinquent behaviors.

CHAPTER 2
LITERATURE REVIEW
I. Delinquent Youths in Korea and Current Circumstances

According to statistics released by the Department of Education in Korea,
delinquency, including physical and verbal aggression has increased this year by
17 percent over last year (Korean Times, 2000). In addition, the Seoul Family
Court surveyed 2,900 middle and high school students, and the results indicate
that 50 percent of the boys and 39 percent of the girls had suffered extortion,
beatings, and other form of violence (Kim & Park, 1996). These results show that
more youths become exposed to violence as both perpetrators and victims.
Additionally, experts say violent victimization goes unreported because students
fear retaliation, or believe their parents are either not interested or too busy to
care (Kim & Noh, 1996). More than half of the victims reported they never told
parents, teachers, or the police about their victimization.

A series of suicides by young students due to delinquency victimization
prompted experts to take emergency measures, but no effective remedies are in
sight because juvenile delinquency is very deeply rooted and rampant. It is
beyond juvenile experts' capability to solve the problem alone, since a solution
would require coordinated efforts among schools, parents, and professionals
focused on juvenile problems, and government authorities.

However, delinquency is not the only negative feature of school life. As
the major factor causing delinquency, Korean experts point out the breakdown in

the school function (Han, 2000). Criticisms are that in the Korean school system

rules are too strict, pressure to succeed is enormous, and the emphasis on rote
learning for testing is such that schools are commonly referred to as an “exam
hell” (Han, 2000).

Most middle and high schools in Korea bring together a large numbers of
students. They are concentrated in the same place for a great amount of time a
day, and are placed in situations where they compete against each other to
achieve their goals; they are compelled by both parents and teachers to aspire to
the top ranked university. The university entrance exam, therefore, determines
one’s future. Afractlon of those in the top 1 percentile will go to the top ranked
universities, considered the pinnacle of Korea's university hierarchy. Getting into
the top ranked schools gives youths the best shot at getting into a top company
or ministry. There are always more students applying to the university than there
are places available for them. More money, therefore, is spent on private tutoring
and exam institutions than on public education. These school conditions could
present an aversive situation to some youth who are low in their level of school
attachment. The pressure might push some students to feel a great deal of strain,
and respond to strain by committing delinquent behavior.

Clearly, the system favors people who are good at rote memorization and
exam taking. Although some universities have started to base their admissions
partly on abstracts, like volunteer work and extra-curricular activities, instead of
just exam results, it has just put an extra burden on the youth -- youths still need
to spend many hours a day studying. Moreover, the high cost of tutoring means

the system is still not really egalitarian; poorer families just cannot afford to keep

their children in the race.

In this context, the excessive expectations for higher education in Korean
society and the inability of educational institutions to deal with the societal and
cultural expectations of higher education may have resulted in increasing levels
of stress of school life. In other words, high levels of parental pressure, frustration,
and strain on youth may produce the competitive setting of the Korean
educational system. Under these conditions, youths who perform poorly and
those from a low socioeconomic status could marginalize themselves in school,
and become more likely to be committed to delinquent behavior than others who
are outstanding academically and from the middle class.

In summary, experience of negative stimuli related to school could push
some youth towards delinquency, and the excessive pressure to do well on
exams and the difficulties that youths may have in school are possible
explanations that why some youths exhibit violent and other forms of delinquency.
In the next section, a more specific theoretical application of strain theory to this

study will be discussed.

II. Theoretical Background

Criminologists and politicians have argued that delinquency could be
explained by several factors including the effects of mass media violence, an
increase in gangs, the breakdown of the family paradigm, and other possible
sources of delinquency. These explanations are reflected in the several dominant

theories of delinquency. For example, differential association /social learning

theory, assumes that youths learn to engage in delinquency from others such as
the criminal family, delinquent friends, and the violent mass media (Vold, Bernard,
& Snipes, 1998). The differential and social learning explanations of delinquency
have proved their relevance through a great amount of empirical research. These
explanations, however, may not be generalizable to all societies, specifically to
Korea, where there is extreme emphasis on the importance of higher education.
In contrast, strain theory may apply to Korean society. According to this theory,
when people fail to achieve desired goals by legitimate means, they might

become upset and some of them might respond with deviant behavior.

1. Overview of Strain Theorv

The classical strain theorists argue that delinquency results from inability
of individuals to achieve positively valued goals through legitimate channels
(Merton, 1938; Cohen, 1955). In the core assumption of strain theory, individuals
focus on the achievement of monetary success or on middle-class values. When
they are unable to achieve the goals by legitimate means, however, they may
turn to illegitimate means in frustration from strain (Cohen, 1955).

The classic strain theory, however, has been criticized for its narrow
conceptualization of strain (e.g., Agnew, 1985). While strain theories were
criticized as having a limited role in explanations of delinquency, Agnew (1985)
developed a revised version of strain theory, which is referred to “general strain
theory” (hereafter referred to as GST). While the classical strain theorists argued

that individuals try to achieve monetary success or middle-class status, the

general strain theorists argued that youths are not only concerned about the
future goals of monetary success or middle-class status, but are also concerned
about the achievement of more immediate goals such as good grades, popularity
with the opposite sex, and maintaining good peer relationships (Elliot and Voss,

1985).

2. The Basic Idea of GST

The core idea of the GST is that delinquency can occur in response to
noxious circumstances and situations as the result of pressure (Agnew, 1985).
Instead of the narrow conception of strain of classical theorists mentioned in the
previous section, general strain theorists develop and expand the range of
situations and circumstances that produce strain. As a result, for instance, Agnew
identifies three major types of strain, which were not identified in the traditional
strain theory. In particular, strain occurs when (1) others remove positively valued
stimuli that youth possess, (2) youths fail to achieve positively valued goals, or
(3) juveniles are presented with negatively valued stimuli. The first type of strain
is derived from the loss of positive stimuli, including such things as the divorce of
parents, or the death of a parent, or loss of a boyfriend or girlfriend. The second
type of strain involves individuals’ failure to achieve positively valued goals. Next,
the presentation of negative stimuli might include various sources of strain, such
as an assortment of stressful life events, dissatisfaction with school life, and
negative relationships with parents, teachers and others (Agnew, 1992). The

above types of strain, according to Agnew (1987), have a cumulative effect on

delinquency. However, the third type of strain for which youths feel negative
stimuli from the environment would be a primary concern of this study, since it
seems especially applicable to the Korean school context.

Agnew (1992) and his colleagues have argued that strain may result from
exposure to painful situations in which negative stimuli occurs. Similarly, related
psychological research found that exposure to various types of aversive stimuli
may lead to aggression (Elliott, 1989). Agnew argued that delinquency is related
to such aversive stimuli as unfair or inconsistent discipline, and adverse or
negative school experiences (1985). In particular, youths are often placed in
aversive situations from which they cannot legally escape. In this circumstance
youths may try to escape from the situation. However, effort to avoid sources of
negative stimuli may be blocked. This blockage of pain-avoidance behavior
frustrates the youths and may lead to illegal attempts to escape such as drug use
or anger-based delinquency (Agnew, 1992). Similarly, other proponents of GST
argue that individuals who feel more negative stimuli will be strongly motivated to
avoid the aversive situation and some of them may respond with deviance
(Piquero and Sealock, 2000). Youths who find school aversive, for example, may
try to transfer to another school, or they may illegally skip school or engage in
delinquent behaviors.

In addition, one of the distinguishing features of youths is that they lack
power and are often compelled to remain in situations where they experience
aversive stimuli (Agnew, 1992). They are compelled to go to a certain school, to

interact with the same group of peers, and live with their family in a certain

neighborhood. If any of these contexts are aversive, there is little the youth can
do legally to escape. Most adults, in contrast, have many legal choices of escape
available, such as divorce, quitting one’s job, or moving to another neighborhood
(Agnew, 1992). In this connection, youths located in adverse environments may
turn to certain behaviors, such as delinquent behaviors, to alleviate their stress

that stems from the aversive environment.

Figure1. Model of This Study

 

 

Independent Variable Dependent Variable
Strain —> Delinquency
(Negative Stimuli) (Y)

(X)

 

 

 

 

 

 

Since this study will focus primarily on school related strain, several
variables which measure negative stimuli related to school will be considered.
Because of the interest in negative stimuli in the school and also because of
limitations of the available data, the full GST model will not be tested. Figure 1
shows the model to be tested in the proposed study. This figure explains that
when youths find negative stimuli in the environment where they are located,
they may feel strain, which is positively related to delinquency. The theory has
suggested that particularly the youths who lack positive help seeking behaviors
may turn to delinquency. This study’s emphasis on school-related strain seems to
be appropriate because of the problems of violence and the societal emphasis on

higher education in the Korean schools. Even though this study cannot test the

full model that is suggested by strain theory, it can still make a contribution to
theory by showing the strength of the relationship between one type of strain,

negative stimuli in Korean schools, and delinquency.

3. School as a maior source of Strain (Negative Stimuli)

General strain theory suggests that some individuals are more likely to
engage in delinquent activities, both within school and outside of school, because
they experience more strain and are more likely to respond to strain with
delinquency (Farnworth, & Leiber, 1989). Related studies suggest that individuals
are more likely to engage in delinquent activities when they experience the
following types of strain: child abuse or neglect, criminal victimization, negative
relationships with teachers and peers, and a wide range of stressful life events
such as the dissatisfaction with school life, and changing schools (Agnew, 1992).

This study suggests four categories of strain in the school: (1) general
dissatisfaction with school, (2) negative teacher relationships, (3) low grades, and
(4) negative peer relationships (e.g., Agnew, 1992). In recent research, theorists
suggest that individuals who experience these types of strain are more likely to

engage in delinquency (e.g., Agnew, 1992).

(1) School Dissatisfaction
Many students experience a general type of strain at school. They find
school boring and a “waste of time,” partly because they have trouble

understanding what is going on in class and keeping up with school-work (Agnew,

10

1987). They have little involvement in school activities, and they would rather be
elsewhere. In sum, Agnew’s research (1987) suggests that youths who are not
involved in productive school performance and are generally dissatisfied with
school activities may feel stress in their school life, and these negative stimuli

may lead some of them to turn to delinquent behavior.

(2) Negative Teacher Relationships

Agnew (1992) suggests that negative relations with teachers are
positively related to delinquency. In school settings, teachers could present
students with negative stimuli. One study found that delinquency was more likely
to occur among students who reported that their teachers often lost their tempers,

made negative comments, and talked down to students (Agnew, 1987).

(3) Low Grades

Students might become upset when they receive poor grades, and
research suggests that the receipt of poor grades is an important source of strain
and delinquency in certain cases (Farnworth et al., 1989). Some criminologists
also suggest that the impact of academic stressors is great among students, who
have a large gap between educational aspirations and actual achievement

(Fitzpatrick, 1999).

(4) Negative peer relationships

Schools, particularly middle schools and high schools, usually bring

11

together large numbers of students. These students have different interests and
personalities, and they do not have much experience coping with difficult
situations. Nevertheless, they are concentrated in the same space for several
hours a day, and they are often placed in situations where they compete against
one another for such things as good grades and status in the adolescent world. It
is not surprising that many of students get Into conflicts with other classmates;
they sometimes treat each other in an extremely competitive manner.

One common example of the negative peer relations in school is
“bullying,” which occurs when a person is exposed repeatedly to negative acts
(Olweus, 1991 ). Negative peer relations such as bullying could be a major source
of strain, and one of the few studies of this suggests that such problems increase
the likelihood of delinquency (Brezina, 1997). In this connection, many of the
students who had committed delinquent activity at school were in part motivated
by the fact that they had been bullied by others, or had experienced other
Interpersonal problems with peers. Although the majority of students do not react
to peer problems in such an extreme manner, negative peer relationships could
be one positive influence on delinquency (Brezina, 1997). Surprisingly,
criminologists, however, have not devoted much of their attention to the impact of

such problems on delinquency.

12

4. Other Sources of Strain -st of Positively Valued Stimuli and Failure to
Achieve Positively Valued Goals

(1) Strain as the loss of positiveILvalued stimuli

As indicated in the previous section, strain occurs when youths lose
positively valued stimuli. A psychological literature on aggression and stress
suggest that youth who have experienced life events involving the loss of
positively valued stimuli were more likely to turn to delinquent behavior.
Numerous examples of such loss can be found in the inventories of stressful life
events. For example, one recent study of adolescent stress employed a life
events list that focuses on such items as the loss of a boyfriend/girlfriend, moving
to a new school district, the divorce or separation of parents, and suspension
from school (Williams, & Uchiyama, 1989). In this connection, the actual or
anticipated loss of positively valued stimuli may lead to delinquency as the youth
tries to prevent the loss of the positive stimuli, retrieve the loss stimuli or obtain
substitute stimuli, seek revenge against those responsible for the loss, or
manage the negative effect caused by the loss by taking illicit drugs.

Although the primary focus of the proposed study is on negative stimuli in
the school, one important measure of loss of positively valued stimuli will be
considered. Loss of one parent living in the home (single parent/both parents) will
be considered as a control variable so that it is possible to see whether negative
stimuli related to school explains more variation in delinquency than it does, and
also so that it is possible to see whether the relationship of negative stimuli in the

school is spurious due to parental status.

13

(2) Strain as the Failure to achieve positively valued goals

General strain theorists in criminology argue that adolescent strain results
when youths fail to achieve positively valued goals (ideal goals) (Agnew, 1986).
The justice literature also suggests that the failure to achieve such expectation
may lead to such emotions as anger, dissatisfaction, disappointment, and
unhappiness — that is, all the emotions customarily associated with strain (Agnew,
1988). The available data to be used for the present research does not include a
complete measure of gaps between goals and achievement of those goals.
However, it does include a measure of grades, which could be assumed to
indicate some gap between educational aspirations and achievement. However,
it is recognized that the absence of a direct measure of aspirations, is a limitation

in the proposed research.

Ill. The Korean Context and Strain Theory

General strain theorists argue that strain occurs when youth experience
negative emotions as a result of extreme pressure to achieve positively valued
goals (Agnew, 1992). Youth who are pressed to achieve certain goals, but do not
have legitimate means, may react to the gap between expectations and
achievements with deviant behaviors. As a good example, when there is a gap
between parental expectations and actual achievement of youth, this situation
might be a source of strain. In this sense, external pressure would refer to
pressure caused from others around youths. Similarly, internal pressure refers to

pressure that stems from self-expectation about achievement in this proposal.

14

Obviously, external parental pressure for academic achievement and the
internally felt burden of study could be a source of negative stimuli. It is possible
that the extreme emphasis on academics that is part of middle and high school
culture in Korea is manifested in both very high parental expectations for
academic achievement (external pressure) and students’ very high expectations
for themselves (internal pressure), and that these pressures become major
sources of strain in for adolescents in Korean society.

In this section, the historical background for the Korean cultural emphasis
on educational achievement will be presented. Then, higher education oriented

school culture, as a major source of strain in Korean society, will be addressed.

1. The linphasis on Higher Education: The Confucianism Explanation

The Korean cultural emphasis on higher education stems from a
traditional philosophy based on Confucian values (Lee, 1999). Originating from
the teaching of Confucius, Confucianism was introduced from China in the same
period as Buddhism, about 1000 years ago during the Three Kingdoms period of
Korea. By the time of the Choson Dynasty (1392 — 1910), Confucian thought
became the ruling ideology of the nation as well as the heart of the educational
system (Lee, 1999) Throughout the time of the Choson Dynasty, Confucianist
values were highly developed, and internalized by the people. These ideals,
however, have been exaggerated and distorted in contemporary beliefs about
higher education.

Confucianist values have a positive influence as the spiritual pillar of the

15

today’s Korean society, but the results of widespread distorted beliefs about
education may at the same time have negative results in current Korean society
(Lee, 1999). Most upper class (Yang-ban: the highest social class in old Korean
society) respected education as a one of greatest Confucianism values, but even
they thought education should be limited to the upper class.

On the other hand, people thought that physical activities including labor
should be carried out by people in the lower class (Sang-min: the lowest social
class in old Korean society) (Lee, 1999). Currently however the belief is that
higher education should be the goal of every person, and there is a singular

emphasis on all youths preparing themselves for the same top universities.

2. _E_xternal and Internal Pressure from Parents and Self-Expeflation - as
Negative Stimuli

In part as a result of Confucian values, contemporary Koreans put great
stock in a college education. It is the most important factor in deciding one's
career. Another reason for such a focus on education is the devotion of Korean
parents to their children. Many parents expect that their children will achieve the
success that they dreamed of, but were unable to achieve. When there is a gap
between parental expectations and actual achievement of adolescents, this
situation might be a source of strain, which leads adolescents to turn to
delinquency.

Youths who grow up in the Korean social environment where there is so

much emphasis on higher education may themselves place great value on higher

16

education. In this proposal, internal pressure refers to pressure that is caused by
high self-expectations about achievement during middle and high school due to

the importance of gaining admission to highly ranked universities or colleges.

3. Parent’s socioeconomic status and negative stimuli

As mentioned in a previous section, socioeconomic status of parents could
cause negative stimuli to youths in Korean schools. In a Korean study regarding
school violence, Kim found that one of major causes of one form of delinquency,
bullying among youths, was poverty (Kim et al., 1996). Specifically, the high cost
of private tutoring means that poor families cannot afford to give extra tutoring to
their children. Youths from poor families, then, may feel they are discriminated by
their peer or by teachers because of their parents’ low socio-economic status. In
addition, since Korean society has been open to the Western culture, one of the
dominant ideologies the Korean culture has is capitalism. The basic criticism on
the establishment of capitalism in Korea is that contemporary Korean culture
places too much emphasis on monetary power of the individual rather than an
individual’s personality and aptitude (Kim et al., 1996). Consequently, youths who
are lower than others in socioeconomic status could be poorly treated and could
marginalize themselves in school, and become more likely to be committed to

delinquent behavior than others who are from the middle class.

17

Figure2. Independent and Dependent Variables

 

Independent Variables

 

 

[ Strain Negative Stimuli I

 

1. School dissatisfaction (X1)

2. Negative teacher relationship
(X2)

3. Grade (>6)

4. Negative peer relationship (X4)
5. lnternal(Xs)

6. External Pressure (X6)

7. Parental Socioeconomic status
(X7)

 

 

 

IV. Moderating Variables

 

Dependent Variables

 

 

 

I Delinquency ]

 

- General delinquency

- Minor delinquency
e.g. smoking, alcohol use,
shop lifting etc.

- Serious delinquency
e.g. drug use, robbery,
vandalism, gang activity etc.

 

Several studies suggest that when they feel strain from various sources,

some youths respond with delinquent behavior. The majority of students,

however, do not react in such an extreme manner. It is difficult to explain why a

few students go to such extremes when nondelinquent responses are also

possible. Although there are wide ranges of factors affecting a person’s methods

for coping with strain and their disposition to delinquency, this study has

limitations in of the variables included in the available data set. Thus, level of

conventional social support and gender will be employed as moderating variables

that might influence the connection of school related strain to delinquency.

18

 

1. Social Support: Positive Adaptation to Strain (Seeking Help)

Leone, (2000) and his colleagues suggest that individuals with high level
of conventional social support such as strong family bond are more resistant to
stress. He also found that social support is significantly related to delinquency
because it helps youths to resist strain. Youths with conventional social supports,
then, should be better able to respond to objective strains in a nondelinquent
manner. GST predicts differential responses to strain for individuals varying in
their level of social support regarding deviant behavior. In this view, individuals
with strong conventional social support would be less likely to respond to strain
with delinquency than individuals with weak social support (Leone, 2000). Figure
3 shows how the moderating variables, social support, would be incorporated

into the model for testing GST.

Figure3. Moderating Variables — Help Seeking Behavior

 

 

 

Moderating Variable Independent Variable
Social Su ort X Strain —> Delinquency
pp (Negative Stimuli)

 

 

 

 

 

 

 

 

 

2. Gender Differences in Delinquency

Despite an increase in research on general strain theory, very little
empirical analysis has investigated the potential differences between males and
females. GST offers a useful framework for assessing similarities and differences

in the risks and processes leading males and females to delinquency. GST

l9

related research has shown that there are many differences between males and
female on the various coping responses they employ in the face of strain. For
example, it is possible that males engage in more delinquency than females
because females cope with stress and strain better or differently than males
(Mazerolle. 1998). As another example, males may engage in violent delinquent
acts at a higher rate because of factors related to gender stereotypes, physical
size (Felson, 1996), socialization to social norms, and differences in self-control
(Gottfredson and Hirschi, 1990). Looking at the effects of gender on delinquency
both directly and through its effect on other independent variables in the model
can contribute to the development of theory to explain delinquency. Also,
considering gender as a moderating variable, that affects the existence and
strength of relationships of independent variables to delinquency can contribute
to theory development. Figure 4 shows how gender would be incorporated into
the model to compare the effect of various independent variables for girls and for

boys.

Figure4. Moderating Variables — Gender Difference

 

Girls — X1 + X2 + X3+ X4+ X5+ X6 —> Delinquency

Boys — X1 + X2 + X3+ X4+ X5+ X5 —> Delinquency

 

 

 

20

CHAPER 3
RESEARCH METHODS

I. Data Collection Procedure and Sample
1. _D_a_t_a Collection Procedure

This data were gathered for the research, “A study on violence around the
school”, funded by the Korean Institute of Criminology. The survey was
conducted from June 12, 1996 to June 28, 1996. The study subjects included
920 middle school students and 999 high school students in Seoul, South Korea.

Specifically, the researchers divided Seoul into 5 areas based on
geographical locations (north, south, east, west, and middle). Each area contains
4 or 5 municipal districts. Based on the population of each area, the researchers,
then, randomly selected 5 boy’s middle schools, 5 girl’s middle schools, 4 boy’s
general high schools. 4 girl’s general high schools, 1 boy’s business high school
and 1 girl’s business high school. Finally, the survey was conducted in 40 classes

of the 20 schools, that is, in two classes in each of the 20 schools.

2. Sample

The total number of the middle school students was 920 (484 male
students and 436 female students) (See Table 1). The total number of general
high school students was 794 (387 male students and 407 female students),
and the total number of business high school students was 205 (108 male

students and 97 female students).

21

Table 1. Distribution of the subject per types of school

 

. High school
middle Total
general business subtotal
Male 484(25.2) 387(20.2) 108(5.6) 495(25.8) 979(51.0)
female 436(22.7) 407(21 .2) 97(5.1) 504(26.3) 940(49.0)

total 920(47.9) 79409.5) 205(205) 999(52.1) 1,919(100)

 

Il. Hypotheses
This study was designed to examine the relationship between an aversive

school environment and delinquent behavior. Several hypotheses will be tested.

Hypothesis 1: Dissatisfaction with school life is positively related to delinquent
behavior.

Hypothesis 2: Negative relationship with teachers is positively related to
delinquent behavior.

Hypothesis 3: Low grades are positively related to delinquent behavior.
Hypothesis 4: Negative relationship with peer is positively related to delinquent
behavior.

Hypothesis 5: Internal and external pressures are positively related to
delinquent behavior.

Hypothesis 6: Parental socioeconomic class is negatively related to delinquent
behavior.

Hypothesis 7: The interaction of conventional social support with indicators of

strain (dissatisfaction with school, negative relations with teachers, low grades,

22

negative relationships with peers, pressure to achieve in school, and parental
socio economic status) is related to delinquent behavior.

Hypothesis 8: Gender moderates the relationships of independent variables
(dissatisfaction with school, negative relations with teachers, low grades,
negative relationships with peers, pressure to achieve in school, and parental

socio economic status) to delinquent behavior.

lll. Operationalized Variables
1. Independent Variables

The Independent variables in this study are divided into the following
seven categories: (1) School dissatisfaction, (2) negative teacher relationships,
(3) low grades, (4) negative peer relationships, (5) internal and external pressure

to achieve academically, and (6) parental socioeconomic status.

(1) School dissatisfaction

School dissatisfaction (X1) is the ‘Level of interest in school’. In order to
measure school dissatisfaction, three items are employed: 1) I am not interested
in school life or study in school; 2) I feel alienation from school including teacher
and classmates; and 3) I think going school is waste of time. Responses for
these variables are: disagree=1, some what disagree=2, neutral =3, mostly

agree= 4, agree= 5.

23

(2) Negative Teacher Relationships

Another independent variable (X2) is negative relationship with teacher. To
measure this variable, the three items are: 1) l have been neglected by teachers
because of poor grade, 2) I have been beaten by teachers, 3) l have experienced
sexual victimization by a teacher including sexual harassment. Possible
responses to items are: never in the past year = 1, one or two times: 2, three or

four times= 3, or more than four times= 4.

(3) Low Grades
The measure of grades (X3) will be agreement with the item, ‘I usually
receive good grades (above 10%)’. Possible responses are: disagree=1, some

what disagree=2, neutral =3, mostly agree= 4, agree= 5.

(4) Mggative peer relationships

To measure the relationship with peer In school (X4), three items were
employed: 1) l have experience of bulling in school, 2) I have experience of
physical victimization by peers, 3) I have experience of sexual victimization by
peers including sexual harassment. Possible responses to items are: never in the
past year = 1, one or two times= 2, three or four times= 3, or more than four

times: 4.

(5) Internal and External Pressure to achieve academically

To measure the internal pressure, the question ‘Gaining admission of

24

college or university is important to me and my future’ was employed. The
question used to measure external pressure was “My mother wants me to receive
good grade for every exam.’ Possible responses to the items are: disagree=1,

some what disagree=2, neutral =3, mostly agree= 4, agree= 5.

(6) Parental socioeconomic status

Parental socioeconomic status was measured by monthly income and
type of occupation of parents. The range of monthly income was 1) less than
500,000 Won (less than $400), 2) between 500,000 Won and 1,000,000 Won ($
400 — $800), 3) between 1,000,000 Won and 1500.000 Won ($ 800 —$1200), 4)
between 1,500,000Won -2,000,000 Won ($1200- $ 1600), 5) between
2,000,000Won -3,000,000 Won ($1600- $2400), 6) above 3,000,000 Won ($
above 2400). Secondly, type of occupations consists 6 categories: professionals,
managers, office workers, sales person, factory or construction worker, and

farmer or fishman.

2. Dependent Variable

Self-reports of the following were used to measure delinquent behavior:

. Drinking in a pub or bar.

. Cheating on exam

. Absence without permission

. Run away

0 Activities in delinquent informal circles

25

. Drinking alcohol

. Smoking
0 Drug use including toxic glue, profane gas
0 Sexual intercourse

. Shop lifting
0 Robbery
- Vandalism
0 Gang fight
0 Threatening classmate to get money
The possible responses to these items were never in the past year, one or two

times, three or four times, or more than four times.

3. Moderating Variables: Social supgm

Moderating variable will be included in the multivariate model. The
variables, social support from parents and siblings is used as moderating
variables. The measure of conventional social support was agreement with the
items: ‘I am not satisfied in relationships with parents.’ and ‘I am not satisfied in

relationships with siblings.’

IV. Data Analysis Procedure
At the beginning of the data analysis, descriptive statistics will be
obtained on all variables in this study to show the demographic characteristics of

students, and the means and the variances or the frequencies for variables.

26

When more than one item is expected to be part of a scale, reliability analysis will
be conducted, and appropriate decisions will be made about which items to
include in each scale.

The research hypotheses will be tested by using correlation and
regression analyses. Correlations will be used to test for linear relationship
between the independent variables and the dependent variable. Finally, to test
the main hypotheses, multivariate regression method will be used. The reason
for use of regression is that the variables, which used in this study are at least
ordinal level. The .05 alpha level is the criterion for significance level in analysis.

The Statistical Package for the Social Science (SPSS) is used to analyze the

data.

27

CHAPTER 4
DATAANALYSIS AND FINDINGS
I. Reliability / Factor Analysis and Scales
1. Independent Variables

Reliability and Factor analysis were used to develop several scales. For
all scales, reliability coefficients were calculated to determine if individual items
formed a reliable scale.

The alpha level for school dissatisfaction indicates that the school
dissatisfaction scale has weak reliability (OF .56). A school dissatisfaction scale
was formed by adding together response values for three items.

Unfortunately, one indicator of negative teacher relations, a question
about sexual harassment, was only used with girls, and it decreased the
reliability for the scale. Thus, it was treated as a separate item in the model
tested with just girls. The alpha statistics for the remaining items on negative
teacher relations was low, but acceptable (.60). Thus, these two items were
combined into a scale.

Similarly, the sexual harassment by peers was only used with girls, and it
decreased the reliability for the scale. Thus, it was treated as a separate item in
the model tested with just girls. The remaining two items, bullying and physical
victimization, produced a scale with the low reliability of .48. However, since the
level of alpha is unacceptably low, all of these items were treated as separate
variables.

The alpha level for the dissatisfaction with family scale is acceptable

28

(a=.79). Based on the reliability coefficient, item dissatisfaction with parents and

dissatisfaction with siblings were computed into a scale.

Table 2. Reliability Coefficients for Scales Used

 

 

Subscales of Negative Strain Number of Items Alpha
School dissatisfaction 3 .56
Negative teacher relation 2 .60
Dissatisfaction with family 2 .79

 

Based on the above reliability test results, three categories of items were
combined into scales. Table 3 summarized the items used in each scale. Also

included is an explanation of what the scale measures.

Table 3. Scales Used / Variable Names

 

School Dissatisfaction Scale (3-15: 3=very satisfied, 15=not at all satisfied)
Interested in school
Alienation
Going to school is a waste of time

Negative Teacher Relation Scale (0-6: 0=very positive relation, 6=very
negative relation)

Neglecting

Beafing

Dissatisfaction with Family Scale (2-10: 2=very satisfied, 10=not at all
safisfied)

Dissatisfaction w/ parents

Dissatisfaction w/ siblings

 

29

In addition, for this investigation, several other independent variables were
formed from the existing data. To measure social status of parents, two variables,
‘FJOB’ and ‘MJOB’ were used to create the variable, ‘STATUS’, in the following
way: if father’s job status is greater than or equal to mother’s job status,

‘STATUS’ was equal to ‘FJOB‘; and if mother’s job status is greater than father’s
job status, ‘STATUS’ was equal to ‘MJOB’. If there is only one parent in the home,
that person’s job status was used to indicate STATUS.

The variable ‘FAMILY (Living with)’ which measures whether a student is
living with either biological parents or step / single parent, was recoded into a
dummy variables, ‘RBOTH’. The dummy variables was created for ‘Iiving with

both biological parent’, and has two categories, no=0 and yes=1.

2. Dependent Variable

Factor analysis is used to determine whether a set of items measures
one or more concept, and also to determine which items should be combined
together into a scale. Since the number of delinquency items is very extensive
(14 items), to reduce the number of observed variables to a smaller number of
I latent variables, identified as factors, factor analysis was employed. For the
factor analysis of delinquency items, promax-rotation was used, because types of
delinquency are likely to be correlated. Table 3 summarized the results of the
factor analysis on the 14 items. Factor 1 in Table 4 contains items related to
minor delinquency items: drinking in a pub or bar, cheating on exam, absence

without permission, drinking alcohol, smoking, shoplifting. Factor 2, referred to as

30

serious delinquency, contains: run away, activities in delinquent informal circles,
drug use including toxic glue and propane gas, sexual intercourse, robbery,

vandalism, gang fight and threatening a classmate to get money.

Table 4. Factor Scores for Delinquency Items Based on Promax-Rotated Two-
Factor Solution

 

 

 

Items Factor loading
1 2
Drinking in a pub or bar .82 .45
Cheating on exam .50 .33
Absence without permission .65 .57
Run away .39 .52
Activities in delinquent informal circles .22 .43
Drinking alcohol .81 .36
Smoking .77 .52
Drug use including toxic glue, propane gas .23 .42
Sexual intercourse .43 .51
Shop lifting .52 .43
Robbery .14 .42
Vandalism .35 .50
Gang fight .46 .61
Threatening classmate to get money .41 .58

 

Note. Boldface indicates highest factor loadings.

Table 5 summarizes the reliability test results for the delinquency scales.
The alpha level for the general delinquency scale is acceptable (or: .84). Based
on the reliability coefficient, a scale, ‘general delinquency; was formed. In
addition, the factor analysis results formed delinquency items into two sub-scales,
minor delinquency scale and serious delinquency scale. The reliability coefficient

for each scale was .83 and .72, which both are acceptable.

31

Table 5. Reliability Coefficients for Delinquency Scales Used (Alpha)

 

Scale Alpha

 

General Delinquency Scale
Including all items of delinquency

Standardized item Alpha .84
Minor Delinquency Scale
Drinking in a pub or bar. .78
Cheating on exam .82
Absence without permission .81
Drinking alcohol .78
Smoking .78
Shop lifting .82
Standardized item Alpha .83
Serious Delinquency Scale
Run away .66
Activities in delinquent informal circles .68
Drug use including toxic glue, profane gas .68
Sexual intercourse .67
Robbery .69
Vandalism .67
Gang fight .64
Threatening classmate to get money .65
Standardized item Alpha .72

 

Based on the reliability and factor analysis, three new delinquency scales
were created. Table 6 summarized the variables used in each delinquency scale.

Also included is an explanation of what the scale measures.

32

Table 6. Scales Used / Variable Names

 

General Delinquency Scale (0-42: 0 = non-delinquent, 42 = extreme delinquent)
Drinking in a pub or bar.
Cheating on exam
Absence without permission
Run away .
Activities in delinquent informal circles
Drinking alcohol
Smoking
Drug use including toxic glue, profane gas
Sexual intercourse
Shop lifting
Robbery
Vandalism
Gang fight
Threatening classmate to get money

Minor Delinquency Scale (0-18: 0 = non-delinquent, 18 = extreme delinquent)
Drinking in a pub or bar.
Cheating on exam
Absence without permission
Drinking alcohol
Smoking
Shop lifting

Serious Delinquency Scale (024: 0 = non-delinquent, 24 = extreme delinquent)
Run away
Activities in delinquent informal circles
Drug use including toxic glue, profane gas
Sexual intercourse
Robbery
Vandalism
Gang fight
Threatening classmate to get money

 

33

ll. Univariate Analysis
1. Demographic Characteristics of Sample

Table 2 shows a description of the sample used in this study. There are
more male respondents (51.0% of the total cases) than female respondents
(49.0% of the total cases), and there are no missing values for this variable. The
sample consists of 920 (47.9% of the total cases) middle school students and
999 (52.1% of the total cases) high school students (see table2).

The majority of the students (91.7% of the total cases) live with biological
parents, and 8.3 percent of students live with stepparents or single parent (see
Table 2). Most students (55.7% of the total cases) report that their family monthly
income ranges from 1,500,000 Won ($1200) to 3,000,000Won ($ 2400), and
students who belong to a higher income family — above 3,000,000 Won (above
$2400) - comprise 14.7 percent of the total students, while students who belong
to a lower Income family — less than 1,500,000 Won ($1200) - were 29.6 percent
of the total students (see table2).

Table2 shows types of occupation of parents, which would reflect social
status of parents. As described in the table, 26.8 percent of students reported
that the type of occupation of their father or mother is professional, or manager —
medical doctor, pharmacist, lawyer, professor, researcher, school teacher,
journalist, writer, engineer, pilot, chief executive officer, and high-ranked public
official such as police officer or military personnel. The majority of students
(63.8%) reported that the type of occupation of their father or mother is office

workers, or sales person — public official, bank teller, self-employed (number of

34

employer are less than 5), real estate agent, security, driver, janitor, delivery
person, etc. Finally, 9.5 percent of students reported that the type of occupation
of their father or mother is factory worker or construction worker, or farmer, or

fisherman (e.g., Kim, et al., 1996).

Table 7. Demographic Characteristics of sample (N=1919)

 

 

n (0/0)
Sex
Male 979 51.0
Female 940 49.0
Type of school
Middle school 920 47.9
High School 999 52.1
Parents Social Status
Parents monthly income* (1 - 6)
Less than $400 29 1.5
$400 - $800 119 6.2
$800 - $1200 391 20.4
$1200 - $1600 530 27.6
$1600 - $2400 485 25.3
More than $2400 267 13.9
Parents’ job status scale (1 — 6)
Professionals 245 12.8
Managers 239 12.5
Office workers 576 30.0
Sales 576 30.0
Factory or construction workers 157 8.2
Farmer or fisherman 15 .8
Social Support
Living with
Both parents 1755 91.7
A parent and step parent / single parent 159 8.3
/ no parent

 

* Korean Won is calculated based on the money exchange rate (1 .00$ = 1,300
Won).

35

2. General Characteristics of Sample

The general characteristics of students are depicted in Table 8. The table
includes the mean scores for school dissatisfaction, negative teacher relations,
grades and negative peer relations. The mean for the school dissatisfaction (3 —
15) is 6.37 and the mean of the grades (1 — 5) is 2.98. Both variables seem to be
normally distributed around the means. However, the negative teacher relations
measures and negative peer relations measures seem to be skewed toward the
left of each mean. The internal and external pressure variables seem to be
normally distributed around the means (3.98 and 3.89 respectively). The
dissatisfaction with family scale and the delinquency measures seem to be left

skewed in their distributions.

Table 8. General Characteristics of Sample (N=1919)

 

 

Mean St. D.

School Satisfaction Scale (3 — 15) 6.37 2.49
Negative Teacher Relation

Teacher relation scale (0 - 6) .80 1.42

Sexual harassment (0 — 3) .13 .46
Grades (1 - 5) 2.98 1.19
Negative Peer Relation

Bullying (0 — 3) .39 76

Physical victimization (0 — 3) .38 .79

Sexual harassment (O — 3) .09 .41
Internal Pressure

Importance of college admission (1 — 5) 3.98 1.17
External Pressure

Parents want good grade (1 — 5) 3.89 1.24
Social Support

Dissatisfaction with Family Scale (2-10) 2.66 1.29
Delinquency

General delinquency scale (0 — 42) 5.08 5.84

Minor delinquency scale (0 - 18) 4.25 4.50

Serious delinquency scale (0 — 24) .83 1.96

 

36

Ill. Bivariate Analysis.

Correlations were computed between independent variables (school
dissatisfaction, negative teacher relations, grade, negative peer relations,
internal/external pressure, parents social status and social support) and
dependent variables (general delinquency, minor delinquency and serious

delinquency).

1. School Dissatisfaction and Delinquency

Table 9 shows that there is a significant correlation between school
dissatisfaction and delinquency. When Pearson’r is calculated for 1) school
dissatisfaction and general delinquency; 2) school dissatisfaction and minor
delinquency; and 3) school dissatisfaction and serious delinquency, its value
was .338, .338 and .225 respectively. Thus, there is positive linear relationship
between school dissatisfaction and each type of delinquency (general, minor and
serious). Also important is that these results are statistically significant at

the .01 level.

2. Negative Teacher Relations and Delinquencv

As table 9 summarized, there are significant relationships between 1)
negative teacher relation scale and general / minor / serious delinquency; and 2)
sexual harassment and general / minor / serious delinquency. In the table 9,
Pearson’s r value shows that the correlations between each independent variable

and each dependent variable are positive in direction. These relationships are

37

statistically significant at the significance level of .01.

3. Grades and Delinquency

Table 9 shows that there is a significant correlation between grades and
delinquency. When Pearson’r is calculated for 1) grade and general delinquency;
2) grade and minor delinquency; and 3) grade and serious delinquency, its value
was .178, .171 and .142 respectively, reflecting that there is positive linear
relationship between grade and each types of delinquency (general, minor and

serious). Again, the results are statistically significant at the .01 level.

4. Mative Peer Relations and Delinquency

Table 9 shows the correlation between negative peer relation measures
and delinquency. Based on the result, there is a positive correlation between (for
girls) sexual harassment and general / minor / serious delinquency. The
relationship is statistically significant at the significance level of .01. However,
there are no significant correlations between bullying and general / minor/
serious delinquency (p=.29 / p=.39 / p= .13), and between physical victimization
and general / minor / serious delinquency (p=.10 / p=.10 / p= .19) Also Table 8
shows that the Pearson’s r is value for 1) the bullying and general / minor/
serious delinquency, were near 0 (r=.01, r=.01 and r= .03), and for 2) physical
victimization and general / minor/ serious delinquency, were near 0 (r=.04, r=.04
and r= .03). The low correlations and the lack of significance suggest that these

independent and dependent variables are not related.

38

5. Internal pressure and Delinguency

Table 9 shows that there is a significant correlation between internal
pressure and delinquency. When Pearson’r is calculated for 1) internal pressure
and general delinquency; 2) internal pressure and minor delinquency; and 3)
internal pressure and serious delinquency, its value was -.14, -.12 and -.15.
There is a negative linear relationship between internal pressure and each type
of delinquency (general, minor and serious), and the correlations are statistically

significant at the .01 level.

6. External pressure and Delinquency

Table 9 shows that there is a significant correlation between external
pressure and delinquency. The Pearson’s r values for 1) external pressure and
general delinquency; 2) external pressure and minor delinquency; and 3) external
pressure and serious delinquency, were -.17, -.16 and -.13. This results shows
that there is negative linear relationship between external pressure and

delinquency (general, minor and serious), at the .01 level of significance.

7. ParerLs’ Social Status and Delinquency

To measure the parents’ social status, two variables (income and parents’
job status) were employed. As shows in the table 9, there are significant
correlations between income and delinquency (general / minor / serious). The
Pearson’s r values for 1) income and general delinquency; 2) income and minor

delinquency; and 3) income and serious delinquency, were .08, .08 and .05.

39

There are slightly positive linear relationships between income and delinquency
(general, minor and serious). The level of significance of these correlations
were .01 level (for general and minor delinquency) and .05 level (for serious
delinquency).

For the parent’s job status, the results indicate that correlations between
1) parents’ job status and general delinquency; and 2) parents’ job status and
minor delinquency, are statistically significant at the .05 and .01 level. However,
there is no significant correlation between the parents’ job status and serious
delinquency. Pearson’s r values support that there are positive correlation
between 1) parents’ job status and general delinquency (r=.07); and 2) parents”

job status and minor delinquency (r=.08).

8. Social Support and Delinguency

Social support was measured by two variables, dissatisfaction with family
scale and living with both parents. Table 8 shows that there are significant
correlations between dissatisfaction with family scale and each type of
delinquency, at the .01 levels. The Pearson’s values for each correlation
were .23, .21 and .20. The results indicate that there are positive relationships
between dissatisfaction with family scale and each levels of delinquency (general,
minor and serious).

For the living with both parents variable, the results indicate that there are
no significant correlation between living with both parents and general / minorl

serious delinquency (p=.05, p=.06, p=.08).

4O

Table 9. Correlation

 

 

General Minor Serious
Delinquency Delinquency Delinquency

School dissatisfaction .338" .338“ .225**
Negative teacher relation

Negative teacher relation scale .381** .376** .270"

Sexual harassment .194“ .184** .164**
Grades .178** .171** .142"
Negative peer relation

Bullying .018 .010 .039

Physical Victimization .043 .042 .030

Sexual harassment .135** .129** .1 08**
Internal Pressure

Importance of college admission -.143** -.124** -.155**
External Pressure

Parents want good grade -.173** -.167** -.135**
Parents social status

Parents monthly income .085** .085** .058*

Parents’ job status .079* .082“ .044
Social support

Dissatisfaction with family .231** .214" .208“

Living with both parents -.O56 -.O52 -.048

 

* Correlation is significant at the .05 level (1-tailed).
** Correlation is significant at the .01 level (1-tailed).

Consequently, most independent variables were significantly related to
delinquency. The school dissatisfaction and the negative teacher relations both
were most strongly correlated to the delinquency among correlated independent
variables. However, bullying, physical victimization by peers and living with both
parents seemed to be not significantly related to any levels of delinquency. In
addition, minor and serious delinquency seemed to have different correlations
with the independent variable, parents’ job status. In turn, the parents’ job status
was significantly related to general and minor delinquency, but not significantly

related to serious delinquency. Overall, correlation statistics shows that none of

41

correlations in this data are exceptionally strong, but correlations of delinquency
with 1) school dissatisfaction and general / minor delinquency; and 2) negative

teacher relations and general / minor delinquency are moderate.

42

IV. Multivariate Analysis

Before starting the regression analysis, to check distributions of the
skewed variables (see Table 8) and regression assumptions, scatterplots were
generated and examined. As the results show (see Appendix A), there seem to
be straight linear relationships between the left skewed independent variables
and the dependent variables. Since the OLS regression is robust enough to
handle the skewed data, the regression analysis was used in this analysis.

In addition, to investigate whether there is problem with multicollinearity
among independent variables, the correlation matrix was employed (see
APPANDIX 2). The correlation matrix shows that the intercorrelations among
independent variables are very high. However, collinearity diagnostics statistics
revealed that there are not problems with multicollinearity among independent
variables, because of high tolerance and low VIF vales.

1. Multivariate Model Regression I

( 1) Regression Model for General Delinquency

Table 10 presents the results of the multivariate regression analysis of
delinquency on the independent variables. As the table shows, school
dissatisfaction (p=. 000), negative teacher relations (p =.000), low grades (p
=.000), low family income (p =.007) and dissatisfaction with family relations (p
= .002) were positively related to general delinquency. However, sex, bullying

experiences (p =.001), physical victimization (p=. 042) and living with both

 

' Collinearity diagnostics analysis was conducted to investigate whether there Is problem with
multicollinearity in multivariate regression analysis. The collinearity diagnostics statistics revealed
that there are not problems with multicollinearity in multivariate regression analysis.

43

parents (p =.017) were negatively related to general delinquency. In turn,
students who have many bullying experiences and physical victimization were
less likely to commit delinquency, and students who live with both parents were
less likely to commit delinquency.

Internal and external pressure for school success, and parents’ job status
were not significantly related to the general delinquency. Moreover, the entire

model explains 25% (R2: .25) of the variation in general delinquency.

(2) Regression Model for Minor/ Serious Delinquency

Similar to the above results, school dissatisfaction, negative teacher
relation, low income, and dissatisfaction with family relations were positively
related to both minor and serious delinquency, while bullying is negatively related
to both minor and serious delinquency. However, when dependent variable was
divided into minor and serious delinquency, other results were moderated.

Minor delinquency - Table 10 presents a positive relationship between low
grade and minor delinquency (p=.000), and a negative significant relationship
between physical victimization and minor delinquency (p=.023), and between
living with both parents and minor delinquency (p =.009). In turn, students who
have low grade were less likely to commit minor delinquency. Also, students who
are frequently victimized physically and who are living with both biological
parents were less likely to commit minor delinquency. In addition, the entire
model explains 24% (R2: .24) of the variation in minor delinquency.

Serious delinquency - Interestingly, internal pressure, which measures

44

importance of college admission, was negatively related to serious delinquency

(p =.006). In other words, students who believe that gaining college admission is

important were less likely to commit delinquency. Also, the entire model explains

16% (R2: .16) of the variation in serious delinquency.

Table 10. Regression Analysis results

 

 

 

General Minor Serious
delinquency delinquency delinquency
B Beta B Beta B Beta
Sex -.16 -.14** 1.64 -.12** -.52 -.13**
School dissatisfaction .54 .23" .38 .22" .14 .18“
Negative teacher relation 1.17 .28" .88 .28“ .28 .21**
Grades .38 .07“ .31 .08“ .06 .05
Negative peer relation
Bullying -.52 -.07** -.33 -.05** -.19 -.07**
Physical victimization -.33 -.04* -.29 -.02* -.04 -.01
Internal Pressure *
Importance of college admission “-17 "-03 '06 '-01 ‘-11 -.06
External Pressure
Parents want good grade -.14 -.03 -.13 -.03 -.01 -.00
Parents social status * *
Parents monthly income 29 05* -18 05* ~11 -06
Parents’ job status .10 .02 .13 .03 -.02 -.01
Social support * * "
Dissatisfaction with family 35 07* 23 06* 42 -07
Constant -1.52 -.67 -.85
R2 .23 .15

 

*p<.05,**p<.01

45

Overall, school related items, including school dissatisfaction, negative
teacher relations, and experiences of bullying, were significantly related to all
types of delinquency (general / minor / serious). In addition, although there are
two variables (income and job status), which measure parents’ socio-economic
status, delinquency was related to monthly income rather than parents’ job status.
Moreover, dissatisfaction with family was significantly related to all types of
delinquency (general, minor and serious). However, living with parents was not
significantly related to delinquency when the type of delinquency is serious. In
this context, emotional attachment to family could be more important than

whether students are living with both parents.

2. Two-way Interaction Analysis between Independent Variables and Social
Support Variables

(1) Two-way interaction with the dissatisfaction with family relation

To investigate effects of the moderating variables, which are social support
variables, two-way interaction analysis was conducted. Figure 5 present the
equation of the restricted multivariate analysis. By employing a moderating
variable, which is the dissatisfaction with family relation, the equation of the two-

way interaction was derived (see Figure 6).

46

Figure 5. The equation of the restricted multivariate analysis

 

 

Y (Delinquency) = a + b1School dissatisfaction + szegative teacher relations +
b3 Grades + D. Bulling + b5 Physical victimization + be Internal pressure + b7
External pressure + be Income + b9 Status + b1o Dissatisfaction with family
relations + b11Living with both parents + e

 

Figure 6. The equation of the two-way interaction analysis between the
independent variables and the dissatisfaction with family.

 

 

Y (Serious Delinquency) = a + b1School dissatisfaction + szegative teacher
relations + b3 Grades + D. Bulling + b5 Physical victimization + be Internal
pressure + by External pressure + be Income + be Status + bioLiving with both
parents + b11lDWFRSD + b12lDWFRNT + b13lDWFGRAD + b14IDWFNP1 +
b15IDWFNP2 + b16lDWFIP + bnlDWFEP + b18IDWFINCO + b1ngWFSTAT + e

 

* In the above equations, each interaction variables were created by following
way:

IDWFRSD = Dissatisfaction with family relations * School dissatisfaction
IDWFRNT= Dissatisfaction with family relations * b2Negative teacher relations
IDWFGRAD = Dissatisfaction with family relations * Grades

IDWFNP1 = Dissatisfaction with family relations * Bulling

IDWFNP2 = Dissatisfaction with family relations * Physical victimization
IDWFIP = Dissatisfaction with family relations * Internal pressure

IDWFEP = Dissatisfaction with family relations * External pressure
IDWFINCO = Dissatisfaction with family relations * Family Income
IDWFSTAT= Dissatisfaction with family relations * Status

As Table 11 indicates, there are two significant interactions between the
independent variables and dissatisfaction with family relations. The moderating
effect of the dissatisfaction with family relations, however, exists only when
serious delinquency is considered. Specifically, the effect of the grade variable on

the serious delinquency changes as dissatisfaction with family relation varies. As

dissatisfaction with family relations increases, the net effect of the grade on the

47

 

 

serious delinquency increases. In addition, the effect of external pressure on
serious delinquency changes with variation in dissatisfaction with family relations.
As dissatisfaction with family relation increases, the net effect of the external
pressure on serious delinquency decreases. Moreover, the entire model explains
17% (R2: .17) of the variation in serious delinquency when the model includes

dissatisfaction with family relations.

48

Table 11. Multivariable analysis for two-way interactions between the
independent variables and the dissatisfaction with family relation (N=1919)

 

 

 

 

 

General Minor Serious
delinquency delinquency delinquency
B Beta B Beta B Beta

School dissatisfaction .63 .27** .48 .27** .15 .04**
Negative teacher relation 1.37 .34“ 1.10 .35** .27 .20**
Grades .28 .06 .37 .10* -.09 -.05
Negative peer relation

Bullying -.15 -.02 -.09 -.02 -.06 —.02

Physical victimization -.29 -.04 -.32 -.06 .04 .01
Internal Pressure -.14 -.03 .04 .01 -.17 -.10
External Pressure .18 .04 .03 .01 .15 .09
Parents social status

Parents monthly income .14 .03 .10 .03 .03 .02

Parents’ job status .16 .03 .06 .02 .10 .06
Social support

Dissatisfaction with family .83 .18 .80 .23 .03 .02

Living with both parents -1.19 -.05* -.99 -.06** -.20 -.03
Two-way interactions of the dissatisfaction of family relation (DWF) with:
School dissatisfaction -.05 -. 11 -.04 -.13 -.01 -.04
Negative teacher relation -.07 -.06 -.07 -.08 .01 .02
Grades .07 .07 -.00 -.00 .07 .20*
Negative peer relation

Bullying -.15 -.06 -.11 -.06 -.05 -.06

Physical victimization .06 .03 .06 .04 .00 .00
Internal Pressure -.03 -.O3 -.05 -.06 .02 .05
External Pressure -.13 -.12 -.07 -.08 -.06 -.17*
Parents social status

Parents monthly income .06 .06 .03 .04 .03 .08

Parents’ job status -.03 -.03 .02 .02 -.05 -.14
Constant -2.81 -2.19 -.62
R2 .26 .24 .17

 

*p<.05 **p<.01

(2) Two-way interaction with living with both parents

By the same token, to investigate the effect of the moderating variable,

which is the living with both parents, two—way interaction analysis was conducted.

The equation of the two-way interaction between the living with both parents and

the independent variables was derived (see Figure 7).

49

Figure 7. The equation of the two-way interaction analysis between the
independent variables and the living with both parents.

 

 

Y (Serious Delinquency) = a + b1School dissatisfaction + szegative teacher
relations + b3 Grades + D. Bulling + b5 Physical victimization + be Internal
pressure + by External pressure + b8 Income + 69 Status + bio Dissatisfaction with
family relations + b11 IRBRSD + bileBRNT + b13lRBGRAD + b14IRBNP1 +
b15|RBNP2 'l' b15|RBIP '1' banBEP "l" b13|RBINCO + b1glRBSTAT '1' e

 

* In the above equations, each interaction variables were created by following
way:

IRBRSD = Living with both parents * School dissatisfaction
IRBRNT= Living with both parents * szegative teacher relations
IRBGRAD = Living with both parents * Grades

IRBNP1 = Living with both parents * Bulling

IRBNP2 = Living with both parents * Physical victimization

IRBIP = Living with both parents * Internal pressure

IRBEP = Living with both parents * External pressure

IRBINCO = Living with both parents * Family Income

IRBSTAT= Living with both parents * Status

As Table 12 indicates, there is only one significant interaction between the
independent variables and living with both parents. The moderating effect of
living with both parents exists only for serious delinquency. Specifically, the effect
of grades on serious delinquency varies depending on whether the youth is living
with both parents. When students live with both biological parents, the net effect
of grades on serious delinquency decreases. Moreover, the entire model explains

17% (R2: .17) of the variation in serious delinquency when moderating for

dissatisfaction with family relations.

50

 

Table 12. Multivariable analysis for two-way interactions between the

independent variables and the Living with both parents (N=1919)

 

 

 

 

 

General Minor Serious
delinquency delinquency delinquency
B Beta B Beta B Beta

School dissatisfaction .50 .21“ .37 .21* .13 .16
Negative teacher relation 1.24 .30** .79 .25“ .45 .32**
Grades -.21 -.O4 .02 .00 -.22 -.13
Negative peer relation

Bullying .18 .02 .08 .01 .10 .04

Physical victimization -.39 -.05 -.14 -.02 -.26 -.10
Internal Pressure .07 .02 -.14 -.04 .21 .12
External Pressure .35 .08 .36 .10 -.O1 -.00
Parents social status

Parents monthly income .16 .03 .10 .03 .06 .03

Parents’ job status .64 .13 .65 .17 -.00 -.00
Social support

Dissatisfaction with family .32 .07** .21 .06" .11 .07"

Living with both parents 2.01 .09 1.54 .09 .47 .06
Two-way interactions of the living with both parents (RBOTH) with:
School dissatisfaction .01 .01 .00 .00 .01 .07
Negative teacher relation -.04 -.01 .13 .04 -.16 -. 11
Grade .71 .17 .38 .12 .34 .24*
Negative peer relation

Bullying -.79 -.10 -.47 -.08 -.32 -.12

Physical victimization .30 .04 -.00 -.00 .30 .12
Internal Pressure -.30 -.08 .05 .02 -.35 -.27
External Pressure -.55 -.15 -.53 -.19 -.01 -.01
Parents social status

Parents monthly Income .14 .04 .09 .03 .05 .04

Parents’ job status -.61 -.15 -.57 -.18 -.03 -.02
Constant —4.60 -3. 14 -1.45
R2 .26 .24 .17

 

*p<.05 **p<.01

(3) ggnificance of the interactions

To test the significance of the interactions and the additional increment in

explained variance in serious delinquency, the F test for the comparison between

the restricted model and the full model was employed. The equation for the F test

51

and the result are as follows:

F:(R:-R:)/(KF-K,,)= (.17—.16)/(20—12) =286
(Web/(N151) (1—.17)/(1919—20—1) '

 

where K1: is the number of independent variables in the full model, KR is the
number of independent variables In the restricted model, and N is the number of
cases. As the coefficient of determination for both the restricted and full model
regression shows (see Table 10, 11, and 12), the full model explains more of the
variation in the serious delinquency than the restricted model. Specifically, the full
model explains 17 % of the variation In the serious delinquency, whereas 16% of
the variation in the serious delinquency can be explained by the restricted model.
In addition, this increase (1%) in the explained variation is statistically significant,
because the significance value of the F (2.86) Is less than .05, where the
numerator degree of freedom is 8, and the denominator degree of freedom is

1898.

(4) The nature of the Interactions
Figure 8 shows the nature of the interaction effects of dissatisfaction with family
relations by grades on serious delinquency. For youth with both low grades and
low level of dissatisfaction with family relations, the mean score for serious
delinquency Is .52. The mean for serious delinquency increases (from .52 to .68)
when dissatisfaction with family relations increases, but grades are held constant

in the low range. By the same token, the mean for serious delinquency increases

52

(from .52 to .80) when grades increase, but dissatisfaction with family relations
remains low. For youth with both high dissatisfaction with family relations and
high grades, the mean for serious delinquency increases dramatically (.52 to
1.32). The combination of high grades and a high level of dissatisfaction with

family relations are most strongly related to serious delinquency.

Figure 8. Nature of interaction: Grades & Dissatisfaction with family relation

 

 

Dissatisfaction with family relation

 

 

Low High
Low .52 .68
Grades
High .80 1 .32

 

 

 

 

 

Figure 9 shows the nature of the Interaction effects for dissatisfaction with
family relation and external pressure on serious delinquency. For youth with high
dissatisfaction with family relation but low external pressure, the mean for serious
delinquency increases (from 1.10 to 1.61). As the external pressure increases,
however, the mean for serious delinquency decreases by half (from 1.10 to .55)
when dissatisfaction with family relations remains low. When both dissatisfaction

with family relations and the external pressure increase, the mean for serious

53

 

delinquency decreases slightly. The combination that is most strongly related to

serious delinquency is high dissatisfaction with family and low external pressure.

Figure 9. Nature of Interaction: External Pressure & Dissatisfaction with family

 

 

 

 

relations
Dissatisfaction with family relation
Low High
Low 1 .10 1 .61
External
Pressure
High .55 .88

 

 

 

 

 

Lastly, Figure 10 shows the nature of the interaction effects of living with
both parents and grades on serious delinquency. When students do not live with
both biological parents and have low grades, the mean for serious delinquency is
1.03. When the students live with a single or step parent rather than both
biological parents and have high grades, the mean for serious delinquency
increases slightly. This combination is most strongly related to serious
delinquency in this figure, but the difference between 1.03 and 1.08 is too small
to consider important. When students live with both biological parents and have

low grades, the mean for serious delinquency is lower than any other

54

 

combination. Interestingly, when the students living with both parents have high
grades, the mean for serious delinquency increases (from .52 to .96). Regardless
of whether a youth’s grade is low or high, not living with both parents is related to
high delinquency. However, when youths live with both parents, if their grades

are high, they tend to exhibit higher delinquency than when their grades are low.

Figure 10. Nature of Interaction: Grade & Living with both Parents

 

 

Living with both parents

 

 

No Yes
Low 1 .03 .52

Grades
High 1 .08 .96

 

 

 

 

 

55

 

3. Multivariate Model for Each Gender Group — Subgroup Analysis

To study possible interaction effects of gender with the independent
variables, subgroup analysis was employed and conducted. Since the male
students were not asked to answer the questions about “sexual harassment”,
and since only the female students answered the questions, the subgroup

analysis is most appropriate for the interaction analysis.

(1) Sub-group Regression Model for General Delinguency

Table 13 presents the results of the multivariate regression analysis
separately for each gender. As the table indicates, in both female and male group,
school dissatisfaction and negative teacher relation were significantly related to
general delinquency.

Sexual harassment by teacher was significantly related to general
delinquency among girls. More specifically, girls who ever experienced sexual
harassment by teacher were .63 times (8:63) more likely to engage In general
delinquency than girls who never experienced sexual harassment by a teacher.
(Since the original data does not ask boys about sexual harassment, the
relationship between sexual harassment and delinquency could not analyzed in
this study.) Similarly, parent’s monthly Income was significantly related to general
delinquency among girls. According to the results presented in table 13, girls who
belong to a high-Income family were more likely to engage in delinquency. Also,
relationship with family was more Important to girls than boys. Girls who are not
satisfied with their relationship with family were more likely to engage in

delinquency than girls who are satisfied. In other words, emotional attachment is

56

more important than whom girls are living with.

Finally, the results show that boys who receive good grade are more likely

to engage in general delinquency than others. In addition, boys who had

experienced bulling were less likely to engage in delinquency. Moreover, the

regression model for girls explains 27% (R2: .27) of the variation in general

delinquency, while the regression model for boys explains 22% (R2: .22) of the

variation in general delinquency.

Table 13. Subgroup Regression analysis results for General Delinquency

 

 

 

 

Female Male
B Beta B Beta

School dissatisfaction .36 .18** .62 .24**
Negative teacher relation

Negative teacher relation 1.19 .27** 1.07 .27**

Sexual harassment .63 .06* - -
Grade .23 .05 .37 .06*
Negative peer relation

Bullying -.37 -.05 -.57 .07*

Physical Victimization -.40 -.04 -.34 -.05

Sexual harassment .42 .03 - -
Internal Pressure

Importance of college admission -.23 -.O5 -.14 -.02
External Pressure

Parents want good grade -.18 -.O4 -.11 -.02
Parents social status

Parents monthly income .42 .10** .17 .03

Parents’ job status .21 .05 .00 .00
Social support

Dissatisfaction with family scale .50 .12** .25 .05

Living with both parents -.95 -.04 -1.00 -.04
R2 .27 .22
* p < .05
** p < .01

57

 

(2) Sub-group Regression Model for Minor Delinquency

As Table 14 Indicates, compared to Table 13, the results of the multiple
regression analysis for minor delinquency for each gender group do not differ
from the results of the previous analysis. Specifically, like the results of the
analysis for the general delinquency, school dissatisfaction and negative teacher
relations were significantly related to both boys’ and girls’ minor delinquency.

For boys, as described above grades was a significant factors related to
minor delinquency. Physical victimization was also significantly related to boys’
delinquency. Boys, who had more experiences in physical victimization, were
less likely to engaged in delinquency. Otherwise, bullying, parents’ monthly
Income and the dissatisfaction with family scale were significantly related to girls’
delinquency. In case of bulling, girls who had more experiences In bulling were
less likely to engaged In delinquency. In addition, the regression model for girls
explains 26% (R2: .26) of the variation in minor delinquency, while the regression

model for boys explains 20% (R2: .20) of the variation in minor delinquency.

58

Table 14. Sub-group Regression analysis results for Minor Delinquency

 

 

 

Female Male
B Beta B Beta

School dissatisfaction .31 .19** .40 .21**
Negative teacher relation

Negative teacher relation .98 .27" .79 .27**

Sexual harassment .42 .05 - -
Grade .17 .05 .33 .08*
Negative peer relation

Bullying -.37 .06* -.28 -.05

Physical Victimization -.30 -.O4 -.33 -.06*

Sexual harassment .30 .03 - -
Internal Pressure

Importance of college admission -.1O -.03 -.03 -.00
External Pressure

Parents want good grade -.17 -.05 -.10 -.02
Parents social status

Parents monthly Income .33 .10** .02 .00

Parents’ job status .18 .05 .05 .01
Social support

Dissatisfaction with family scale .35 .11** .16 .04

Living with both parents -.75 -.04 -.94 -.05
R2 .26 .20

 

*p<.05, **p<.01

(3) Sub-qroup Regression Model for Serious Delinquency

As Table 15 indicates, compared to Table 13 -14, the results of the multiple
regression analysis for serious delinquency for each gender group do not differ
from the results of the previous analysis. Specifically, like the results of the
analysis for the general and minor delinquency, school dissatisfaction and
negative teacher relation were significantly related to both boys’ and girls’ serious
delinquency. However, parents’ monthly income became significant among boys.
In turn, boys who belong to a high-income family were more likely to engage in

delinquency. Similarly, grades were not significantly related to the boys’ serious

59

delinquency.

For girls, sexual harassment by teacher, parents’ monthly income and

dissatisfaction with family were still significantly related to serious delinquency. In

addition, Importance of college admission was significantly related to girls’

serious delinquency. Girls who believe that gaining college admission Is

Important were less likely to engage in serious delinquency. The full models for

both girls and boys explains 15% (R2: .15) of the variation In serious delinquency.

Table 15. Sub-group Regression analysis results for Serious Delinquency

 

 

 

 

Female Male
B Beta B Beta

School dissatisfaction .05 .08* .22 .22**
Negative teacher relation

Negative teacher relation .20 .17** .28 .19**

Sexual harassment .21 .07” - -
Grade .06 .05 .04 .01
Negative peer relation

Bullying -.O1 -.00 -.29 -.10*

Physical Victimization -.10 -.04 -.O1 -.00

Sexual harassment .12 .03 - -
Internal Pressure

Importance of college admission -.12 -.11* -.10 -.05
External Pressure

Parents want good grade -.01 -.00 -.00 -.00
Parents social status

Parents monthly income .09 .O8* .15 .07*

Parents’ job status .02 .02 -.05 -.02
Social support

Dissatisfaction with family scale .14 .14** .09 .04

Living with both parents -.20 -.O4 -.06 -.00
R2 .15 .15
* p < .05
** p < .01

6O

4. Two-way Interaction Analysis between lndemient Variables and Social
Support Variables for Each Gender Group

(1) Two-way interaction with the dissatisfaction with family relation

To investigate moderating effects of the social support variables in the
subgroup analysis, two-way interaction analysis was conducted. As Table 16 and
Table 17 indicate, five significant Interactions were found in the female group and
one significant interaction was found in the male group. More specifically, the
interaction between negative teacher relations and dissatisfaction with family
relations Is significantly related to serious delinquency for the female group. The
interaction between sexual harassment by teachers and dissatisfaction with
family relations is significantly related to all measures of delinquency for girls.
The interaction between external pressure and dissatisfaction with family
relations Is significantly related to serious delinquency for the female group.
Finally, for the male group, only the grades and dissatisfaction with family

relations interact significantly, affecting serious delinquency.

61

Table 16. Multivariable analysis for two-way interactions of the dissatisfaction
with family relation in the female subgroup analysis (N=940)

 

 

 

 

 

 

General Minor Serious
delinquency delinquency delinquency
B Beta B Beta B Beta
School dissatisfaction scale .48 .24** .41 .25** .07 .14
Negative teacher relation
Negative teacher relation scale 1.19 .28** 1.17 .33** .02 .02
Sexual harassment by teachers 3.41 .33“ 2.31 .27** 1.10 .41**
Grade .31 .07 .34 .10 -. 03 -. 03
Negative peer relation
Bullying .41 .06 .20 .04 .21 .11
Physical victimization -. 79 -. 09 -. 68 -. 09 -. 11 -. 05
Sexual harassment -. 27 -. 02 -. 52 -. 06 .25 .08
Internal Pressure -. 45 -. 11 -. 24 -. 07 -. 20 -. 19*
External Pressure .45 .11 .27 .08 .20 .19
Parents social status
Parents monthly income .11 .03 .14 .04 -. 02 -. 02
Parents’ job status .17 .04 .14 .04 .04 .03
Social support
Dissatisfaction with family 1.19 .30 1.03 .32 .16 .15
scale
Two-way interactions of the dissatisfaction of family relation (DWF) with:
School dissatisfaction scale -. 04 -. 10 -. 03 -. 10 -. 01 -. 09
Negative teacher relation scale .02 .02 -. 04 -. 05 .07 .22*
Sexual harassment by teachers -. 98 -.31** -. 66 -.25** -. 32 -.39**
Grade -. 02 -. 02 -. 06 -. 07 .04 .15
Negative peer relation
Bullying -. 29 -. 13 -. 21 -. 11 -. 08 -. 12
Physical victimization .12 .05 .11 .05 .01 .01
Sexual harassment .26 .07 .30 .10 -. 05 -. 05
Internal Pressure .09 .11 .06 .08 .04 .16
External Pressure -. 23 -. 27 -. 16 -. 22 -. 07 -. 33*
Parents social status
Parents monthly Income .11 .12 .07 .09 .04 .18
Parents’ job status .01 .01 .02 .02 -. 01 -. 02
Constant -3.82 -3.27 -.55
R2 .28 .27 .18
*p<.05 **p<.01

62

Table 17. Multivariable analysis for two-way interactions of the dissatisfaction
with family relation in the male subgroup analysis (N=979)

 

 

 

 

 

General Minor Serious
delinquency delinquency delinquency
B Beta B Beta B Beta

School dissatisfaction scale .77 .30** .52 .28** .25 .25**
Negative teacher relation

Negative teacher relation 1.09 .28** .84 .29" .25 .17*
scale
Grade -. 07 -. 01 .19 .05 -. 26 -. 12
Negative peer relation

Bullying -. 40 -. 05 -. 20 -. 04 -. 20 -. 07

Physical victimization -. 58 -. 09 -. 56 -. 11 -. 02 -. 01
Internal Pressure .05 .01 .20 .05 -. 15 -. 07
External Pressure .15 .03 -. 01 -. 00 .16 .09
Parents social status

Parents monthly income .40 .07 .21 .05 .19 .09

Parents’ job status .27 .05 .08 .02 .20 .10
Social support

Dissatisfaction with family 1.21 .25 .10 .28 .24 .13
scale
Two-way Interactions of the dissatisfaction of family relation (DWF) with:
School dissatisfaction scale -. 04 -. 09 -. 03 -. 11 -. 01 -. 04
Negative teacher relation

Negative teacher relation .01 .01 -. 01 -. 01 .02 .04
scale
Grade .20 .18 .07 .09 .13 .30"
Negative peer relation

Bullying -. 08 -. 04 -. 05 -. 03 -. 04 -. 04

Physical victimization .09 .04 .08 .06 .00 .00
Internal Pressure -. 07 -. 06 -. 08 -. 11 .01 .04
External Pressure -. 10 -. 08 -. 03 -. 04 -. 06 -. 13
Parents social status

Parents monthly income -. 10 -. 11 -. 09 -. 12 -. 02 -. 05

Parents’ job status -. 11 -. 09 -. 01 -. O1 -. 10 -. 22
Constant ~3. 12 -1.76 «1.36
R2 .22 .19 .16

 

*p<.05 **p<.01

63

(2) The nature of the Interactions

Table 18 shows the nature of the interaction effects that were found in the
previous multivariate interaction analysis for each gender group. For female
students with low negative teacher relation and low level of dissatisfaction with
family relations, the mean score for the serious delinquency is .17. When both
relationships with teachers and family become worse, the mean of the serious
delinquency increases dramatically (1.22).

By the same token, for female students, as the frequency of the sexual
harassment by teachers and the dissatisfaction with family relations Increase, the
means for all of the general, minor and serious delinquency Increase dramatically
(8.07, 6.86 and 1.21 respectively). In addition, the combination of low external
pressure and high dissatisfaction with family relations are most strongly related to
serious delinquency.

For male students with low grades and low dissatisfaction with family
relations, the mean for serious delinquency is .99. However, when both grades
and dissatisfaction with family relations increase, the mean for serious
delinquency increases to 1.51. Thus, the combination of high grades and high
level of dissatisfaction with family relations is most strongly related to serious

delinquency for male students.

64

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 18. Nature of Interaction: Dissatisfaction with family relation in each gender
group
Dissatisfaction with family
relation
Low High

Female Negative teacher relation scale Low .17 .60

in Serious Delinquency High .68 1.22

Sexual harassment by teachers Low 2.81 5.20

in General Delinquency High 5.71 8.07

Sexual harassment by teachers Low 2.55 4.45

in Minor Delinquency High 4.97 6.86

Sexual harassment by teachers Low .26 .75

In Serious Delinquency High .74 1.21

' External pressure in Serious Low .60 1.25

Delinquency High .20 .60

Male Grades In Serious Delinquency Low .99 1.07

High 1.09 1.51

 

 

 

 

 

(3) Two-way interaction with living with both parents

As Table 19 and Table 20 indicate, four significant interactions were found

only in the female group. More specifically, the interactions between the negative

teacher relations and living with both parents are significant In the general and

minor delinquency for the female group. The interaction between the sexual

harassment by teachers and living with both parents is significantly connected to

serious delinquency for girls. The interaction between sexual harassment by

peers and living with both parents is significantly related to girls’ minor

delinquency.

65

 

Table 19. Multivariable analysis for two-way interactions of the living with both
parents in the female subgroup analysis (N=940)

 

General Minor Serious
delinqpency delinquency delinquency

 

B Beta B Beta B Beta

 

 

 

School dissatisfaction scale .32 .16 .21 .13 .11 .21
Negative teacher relation
Negative teacher relation -. 17 -. 04 -. 24 -. 07 .07 .06
scale -1.61 -. 16 -. 70 -. 08 -. 91 -. 34
Sexual harassment by
teachers
Grade .41 .10 .34 .10 .07 .06
Negative peer relation
Bullying -. 49 -. 07 -. 19 -. 03 -. 29 -. 16
Physical victimization .50 .06 -. 11 -. 01 .61 .26
Sexual harassment 4.10 .36 4.76 .51* -. 66 -. 22
Internal Pressure -. 16 -. 03 -. 01 -. 00 -. 11 -. 10
External Pressure -. 56 -. 14 -. 49 -. 15 -. 07 -. 07
Parents social status
Parents monthly income .44 .11 .35 .11 .08 .08
Parents’ job status 1.11 .26 .87 .25 .24 .21
Social support
Living with both parents -. 20 -. 01 -. 56 -. O4 .36 .07
Two-way interactions of the living with both parents (RBOTH) with:
School dissatisfaction scale .16 .09 .19 .14 -. 04 -. 08

Negative teacher relation scale 1.54 .34* 1.38 .37* .16 .13
Sexual harassment by teachers 2.43 .23 1.24 .14 1.19 .43*

Grade -. 07 -. 02 -. 09 —. 03 .02 .02
Negative peer relation
Bullying .07 .01 -. 20 -. 03 .28 .14
Physical victimization -. 95 -. 10 -. 22 -. 03 -. 73 -. 30
Sexual harassment -3.87 -. 33 -4.64 -. 48* .77 .25
Internal Pressure -. 15 -. 05 -. 12 -. 05 -. 03 -. 03
External Pressure .38 .12 .31 .12 .08 .10
Parents social status
Parents monthly income -. 04 -. 01 -. 03 -. 01 -. 01 -. 01
Parents’ job status -. 97 -. 27 -. 74 -. 25 -. 23 -. 25
Constant -.96 -.38 -.58
R2 .27 .26 .15

 

*p<.05 **p<.01

66

Table 20. Multivariable analysis for two-way interactions of the living with
parents in the male subgroup analysis (N=979)

both

 

General Minor Serious
delinweng/ delinguency delinquency

 

 

 

 

B Beta B Beta B Beta

School dissatisfaction scale .53 .21 .30 .16 .23 .23
Negative teacher relation

Negative teacher relation 2.02 .51** 1.43 .49** .60 .40**
scale
Grade -. 19 -. 03 .02 .00 -. 20 -. 10
Negative peer relation

Bullying .65 .09 .51 .09 .14 .05

Physical victimization -1.06 -. 16 -. 61 -. 12 -. 45 -. 17
Internal Pressure .10 .02 -. 29 -. 07 .39 .19
External Pressure -. 15 -. O3 -. 08 -. 02 -. 07 -. 04
Parents social status

Parents monthly income -. 14 -. 03 -. 14 -. 03 -. 01 -. 00

Parents’ job status .62 .12 .65 .17 -. 03 -. 01
Social support

Living with both parents -1.29 -. 06 -1.76 -. 10 .46 .05
Two-way interactions of the living with both parents (RBOTH) with:
School dissatisfaction scale .17 .08 .15 .10 .02 .02
Negative teacher relation

Negative teacher relation -. 95 -. 23 -. 63 -. 21 -. 31 -. 20
scale
Grade .72 .16 .42 .13 .30 .17
Negative peer relation

Bullying -1.34 -. 17 -. 88 -. 15 -. 46 —. 16

Physical victimization .78 .11 .31 .06 .47 .18
Internal Pressure -. 26 -. 06 .27 .09 -. 53 -. 34
External Pressure -. 01 -. 00 -. 07 -. 02 .05 .04
Parents social status

Parents monthly income .36 .09 .19 .06 .17 .11

Parents’ job status -. 65 -. 15 -. 63 -. 19 -. 02 -. 01
Constant .90 2.20 -1.30
R2 .22 .20 .16

 

*p<.05 **p<.01

67

(4) The nature of the Interactions

Table 21 shows the nature of the Interaction effects for girls. The female
students who live with both parents and have low negative teacher relations are
least likely to commit general delinquency (mean of general delinquency = 2.75).
For female students, the negative teacher relationship slightly more affects
general delinquency than living with both parents does (6.19 vs. 4.85). Thus, the
combination of the high negative teacher relations and the not living with both
parents most characterizes girls highest in general delinquency.

Similarly, negative teacher relationship slightly related to minor
delinquency than living with both parents does (5.36 vs. 4.17). The combination
of high negative teacher relations and the not living with both parents Is most
strongly related to minor delinquency for girls (mean score=6.58).

When female students who lIve with both parents have low frequency of
sexual harassment by teachers, they are less likely to commit serious
delinquency. Not living with both parents and sexual harassment by teachers are
almost equally related to serious delinquency for girls (.92 vs. .97).

Finally, female students who have low frequency of sexual harassment by
peers and who live with both parents are least likely to commit minor delinquency.
On the other hand, the combination of high sexual harassment by peers and
living with a single parent or stepparents are most strongly related to minor

delinquency for the female group (mean = 12.00).

68

Table 21.

Nature of Interaction: Living with both parents in each gender group

 

Living with both parents

 

 

 

 

 

 

 

 

 

 

 

 

No Yes

Female Negative teacher relation scale Low 4.85 2.75
in General Delinquency High 8.00 6.19
Negative teacher relation scale Low 4.17 2.48

in Minor Delinquency High 6.58 5.36

Sexual harassment by teachers Low .97 .37

in Serious Delinquency High .57 .92

Sexual harassment by peers in Low 4.69 3.11

Minor Delinquency High 12.00 5.18

 

 

 

69

 

CHAPTER 5
DISCUSSION AND CONCLUSION
l. Summary and Discussion

1. Multivariate Analysis

Several important findings emerged from this study. The multivariate
analysis revealed that youths who feel negative strain from school dissatisfaction
and negative teacher relations are more likely to commit both minor and serious
delinquency. These findings are consistent with the results of several other
studies that examined negative strain and delinquency (e.g. Agnew 1985; 1992).
However, the current study does not support the hypothesis that internal and
external pressure related to academic achievement significantly affects
delinquency.

This study also found that youths who are living with both biological
parents were less likely to commit general and minor delinquency. This finding is
consistent with Leone’s (2000) finding that youths with strong parents attachment
would be less likely to be Involved in delinquency. Moreover, this finding also is
consistent with another research result that parental monitoring of youth is
negatively related to delinquency (Wright & Cullen, 2001). It may be that such
monitoring and supervision are not possible when there is only one parent in the
home. Thus, living with both parents may decrease the possibility of committing
delinquency.

Several other unexpected findings also emerged from this study. As

expected, this study revealed that grades, bullying, internal pressure (in serious

70

delinquency) and parents’ income were significantly related to delinquency.
However, the relationships of these variables to delinquency were in the opposite
direction than what was expected. For example, in hypothesis 3, low grades were
expected to be positively related to delinquency. The regression statistics,
however, found that the direction of relationship was negative. In turn, students
who receive higher grade were more likely to commit delinquency. One of the
possible explanations for this unexpected finding is that the grade could be a part
of the outcome of parents’ socio-economic status. In Korea, a great number of
students are having private tutoring outside of schools and private tutoring costs
a great amount of money. Those students who have private tutoring will spend
most of their free hours studying instead of spending time with their parent(s).
Consequently, parents may be satisfied with the students’ high grades, but the
students may not be satisfied with their relationship with parents. This
dissatisfaction with their parents may cause delinquency in general. (See
Interaction between grades and social support). However, this study does not
include variables such as time spent with parents or time spent with tutors.
Future research needs to measure such variables to try to figure this out.
Similarly, the present study found that bullying experiences and physical
victimization, reflecting negative peer relations were negatively related to both
minor and serious delinquency. Why would these experiences be related to low
delinquency? Youths who are characterized by unreceptive tendencies to the
dominant sub culture tend to emerge as frequent victims of bulling, and as the

result, they could be victimized physically (Schwartz, Farver, Chang & Lee-Shin,

71

2002). In this connection, possibly, one explanation is that there might be some
peer pressure on youths to be delinquent, and when they are not, other youths
would bully them.

Another study finding is that internal pressure was significantly related to
serious delinquency, but the direction of the relationship was opposite to the
hypothesis, as internal pressure was negatively related to the serious
delinquency. The possible reason for the negative relationship may be a
measurement problem. Perhaps this is really a measure of “goals” not of internal
pressure. It may not create much stress or strain for youths. Future research
must focus on the conceptually accurate meanings of Internal pressure.

Finally, this analysis revealed that parent’s socio-economic status was
positively related to delinquency. Hypothesis 6 assumed that parents’ low
socioeconomic status might lead students to experience discrimination in current
Korean school environment by teacher and peers. In this sense, low income
would be a negative strain to students. However, the results suggest that
students who belong to high-income families are more likely to commit
delinquency. Possibly, high-income students might have resources such as
money to get cigarettes and drug, or money to drink alcohol in a bar. This
accessibility to delinquent resources might lead students to commit more
delinquency. Thus the accessibility to delinquent resources may be more

significant than negative strain from parent’s social status.

72

2. Interaction Effects

To investigate interaction effects of moderating variables, dissatisfaction
with family relations and living with both parents were employed. Interestingly,
interaction effects existed only when serious delinquency is considered.
Specifically, the results revealed that when the moderating variable is
dissatisfaction with family, there are significant interactions between 1) grade and
serious delinquency; 2) external pressure and serious delinquency. Similarly,
there Is a significant interaction between grade and serious delinquency when
living with parents is the moderating variable. When students’ grades are high
and the level of dissatisfaction with family relations is high, they are most likely to
commit serious delinquency. This finding suggests that even though students
receive good grades, if they are not emotionally attached to their family, they
might be more likely to commit delinquency. Similarly, students who have high
levels of dissatisfaction with family relations and low external pressure are most
likely to commit serious delinquency. From these findings, a high level of
dissatisfaction with family relations was considered to be the most important
factor that affects serious delinquency. In addition, students who receive high
grades but live with a parent and a stepparent, or single parent were most

involved in serious delinquency.
3. Gender Differences

This study compared boys and girls, and found school dissatisfaction and

negative teacher relations positively related to general, minor and serious

73

 

delinquency In both the boys’ and girls’ group. However, subgroup analysis for
each gender revealed that sexual harassment, Income and dissatisfaction with
family relations were most strongly related to girl’s delinquency. Girls who are not
satisfied with family relations were more likely to commit delinquency than girls
who are satisfied with the relationship.

This finding suggests that emotional attachment is more important to girls
regardless of whether they are living with both parents or not, because
dissatisfaction with family relations was significantly related to delinquency, but
living with both parents was not related to delinquency for the female group. This
finding is consistent with much of the prior delinquency research, which
suggested that positive family relations matter more to girls than boys in
explaining delinquency (e.g., Anderson, Holmes & Ostresh, 1999).

The results show that boys who receive good grade are more likely to be
engaged in minor delinquency than others. Also boys who have many
experiences of physical victimization were less likely to commit minor
delinquency. Interestingly, income is only related to boys’ serious delinquency.
Boys who belong to a high-income family were more likely to commit serious

delinquency.

74

ll. Limitations and Recommendations

There are several limitations of this research that merit mentioning. First,
since the data used in this study has been collected for previous research, this
study cannot represent the overall national - wide population of Korean schools.
The original data were collected among schools In Seoul, which is the capitol city
of South Korea. Therefore, this study focuses only on youths and their delinquent
behaviors in the city area. To compare differences on delinquency between city
and rural area, wider regional sampling Is necessary In future research. Also,
since the present study was based on cross-sectional data, a longitudinal study
might help to explain more variations in negative stimuli and delinquency over
time, and establish time order.

While the primary focus of the present study is on negative stimuli in the
school, other important measures of loss of positively valued stimuli and failure to
achieve positively valued goals were not considered in this study. The available
data to be used for the present research does not include a complete measure of
the loss of positively valued stimuli and the failure to achieve positively valued
goals. The absence of these measures is considered as another limitation of this
study.

Finally, since questions about sexual harassment were only used with girls,
no comparison group for sexual harassment was available. As such, the findings
can only be generalized to a specific population. In future research, if a comparative

sample could have been secured, this explanation might have been clarified.

75

Ill. Conclusion and Policy Implications

Generally, it would be acceptable to argue that no society is delinquency
free. Numerous studies and practical attempts have been made to address this
problem to guard society and to protect the future. Failure to address this existent
problem, however, creates an even greater social problem when these youths
enter society as adults. In one sense, this study aims to provide a more in-depth
understanding of existing delinquency problems In the Korean context. To better
understand the causation of delinquency, empirical analysis was conducted
based on pre—collected data. The multivariate regression analysis results in this
study suggest that negative experiences (negative stimuli) from school
environments including teacher and peer cause delinquency among Korean
youths, and student’s dissatisfaction with school is also significantly associated
with delinquency. Moreover, emotional attachment to parents decreases the
probability of delinquency.

These research results may provide perspectives for effective
interventions into schools and family. One of the major reasons why students
experience school related negative strains is because they are not interested in
school activities. (This low school attachment was described above as school
dissatisfaction). As a result, they do poorly in school, often get Into conflicts with
teachers and others, and develop a general dislike for school. Normally, school In
Korea is focused on higher education, and teacher and parents have excessive
expectations for higher education. The pressure might push certain youths to feel

a great deal of strain, and respond to strain by committing delinquent behavior. In

76

this context, school Itself must change its educational philosophy in ways that not
only focus on academic success but also consider each Individual’s interest and
aptitude. Many of such efforts gradually will reduce school strain. For example,
efforts might be made to Increase the Involvement and experience of success of
students at high risk for delinquency.

Second, the present study suggests that emotional attachment between
youths and parents reduce risk for delinquency. One of the study results, that
youth who belong to high-income families are more likely to engage In
delinquency, suggests that parents’ social support is more Important than
material wealth. Numerous studies also support that well-monitored youths are
less involved in delinquency and other norm-breaking behaviors (Stattin & Kerr,
2000). This finding suggests that parents’ effort to build attachments, monitor
their youths, and spend time with their youths are important because this form of
parenting will keep their youths delinquency free. As mentioned in study limitation,
to try to figure this out, future research needs to have more complete measure on
suchissues.

Overall, General Strain Theory is useful In explaining contemporary
delinquency problems in Korea. However, any singular theory may not explain all
types of behavior. Moreover, any singular effort to keep youth safe may not be
successful. An effective solution for delinquency In current school setting In
Korea requires coordinated efforts among schools, parents, government
authorities and professionals focused on juvenile problems. The schools are a

logical place to begin addressing the problem of juvenile delinquency. However, if

77

 

our efforts go no further, there is little hope for the significant long-term
improvements that are possible through a more comprehensive attack on the

problem.

78

APPENDICES

79

Appendix A. Scatterplots

 

 

 

 

 

 

30
U
20‘
9
CU o D
U 0 CI 1:)
U) o o D o
' 3 D o 3
5‘ 10‘ o o p o a
C] D D D U U
C o n o D n o
g E E E E 3 a 3
L5 OI 0 1:1 :1 g g g 3
U
(D
3
.9
CD
CD '10 _ _ _ - _ _ _
-1 O l 3 4 5 6 7
Negative teacher relation (ntl+nt2) - Scale
14.
o
12'
10‘
2 o
8 8: 0
CD
I u :1
>4 6' D
O
C D D C!
(1:) 41 0 :1 n
U
C D D
E) 21 [J D D
U ///"
U) U U
8 Out a a o a
5
CD -2 ' ' _ _ _ . .

 

 

3.5

Negative Teacher Relationships-Sexual harrassment

8O

 

D D DDnD

D D DDDnD

DD 000090

000000 0000000090

 

3.5

3.0

1.5 2.0 2.5

1.0

 

 

30

O
2

0 0 0
_

9QO 1 3:63:28 macrom

 

Negative peer relationships - Bullying

 

D 0

000000 00

D DODUD

D 000 000030

000000 OODDDDDDn

 

0.0

 

 

30

O
2

0

28m 1 >ocosoc__mu macrom

0

O
_

 

1.5 2.0 2.5 3.0 3.5

1.0

Negative peer relationships-Physical victimization

81

Serious delinquency — Scale

 

 

 

 

14

o
12'
10'

0
8| 0

o 13 o
6' 0

n o
4: D o

o D u
2: 1:1 o D

o a 1:1
OI o o o 1::
-2 . s a r . v .
-.5 0.0 5 1.0 1.5 2.0 2.5 3.0

Negative peer relationships—Sexual harrassment

82

o:_m> u.) 26. ecu 85292 :9: 20 $333 .8285 3_Lmoc___ooz_:E o: 3398 mesons-.3 osmocmflo 3:35:00 6qu

8.1.: 8.1...

 

 

: :2.- ::.- :8. 8. 8. 8. 8.- 8.- 8. 8. 8.- 8.- 2:28 £8 58 9.33.8
-- 8. :2. 8.- .8.- 8. :2. .8. :2. 8. :2. :8. 833228258 8822685 .8
.- :2.- .8.- ..2.- 8.- 8.- .8.- 8. 8. 8. .8. mass 88.823 .2
-- :2. :8. 8.- 8. :2.- 8. 8. 8. 8.- Ecosystems .2
-- :8. 8.- 8. 8.- 8.- :2.- ..2. :8.- osoooa _oEovm .m
- 8.- .8. 8. 8.- 8.- :2.- ..8. ego-88652:. .w
-- :2. :2. :2. :2. :8. :3. Loon 8. 255822 .853 s
- :2. :2. :2. :2. :8. Loos 8 c8888; 62an o
-- :2. 8. ::. :2. 825m 18:29 Loos 9882 .m
-- .8. :2. :8. 890 .v
.- tau. :3. 29.82 3 E2536: _maxow .m
I 3mm. Emom c0339 Emcommu m>zmmm2 .N
I Emom cozomvwzmwwfi _oocow .F
8 .2 2 E m m e m m. e m N e 2:822

 

woBmtm> Eoocoaoos __< 8., 5:22 cozflotoo .m xficoaq<

83

BIBLIOGRAPHY

84

Agnew, R. (1985). A revised strain theory of delinquency. Social Forces, 64(1)-
151 -— 167.
. (1989). A longitudinal test of the revised strain theory. Journal of
Quantitative Criminoloqv- 5(4). 373 — 387.
. (1992). Foundation for a general strain theory of crime and
delinquency. Criminology. 30(1), 47 — 87.
Agnew, R., & White, HR. (1992). An empirical test of general strain theory.
Criminoloqv- 30(4). 475 — 499.
Anderson, B.J., & Holmes, M. D., & Ostresh, E. (1999). Male and female
delinquents’ attachments and effects of attachments on severity of self-
reported delinquency. Criminal Justice and Behavior, 26(4), 435 — 452
Boo, RM, 8. Kim, SR. (2000, April7). 2431188- QEH [School Violence]. D_ogg;
Ah Daily News, pp.A1

Choi.Y.S. & Park, S.J. (1999). Juvenile Delinquency in Korea. Seoul, Korea:
Seoul National University Press.

Cohen, A. (1955). Delinquent _Bovs. New York: Free press

Daly, K and M. Chesney-Lind. (1998). Feminism and Criminology. m
Quarterly, 5. 497-538.

Gottfredson, M. and T. Hirschi. (1990). A General Theorv of Crime. Stanford
CA:Stanford University Press.

Farnworth, M., & Leiber, M.J. (1989). Strain theory revisited: Economic goals,
educational means, and delinquency. American Sociological Review. 54,

263 — 274.

85

Federal Bureau of Investigation. (1992). Uniform Crime Reports for 11th US. 1991
Washington: DC: US. Government Printing Office.
Mazerolle, P., & Maahs, J. (2000). General strain and delinquency: An alternative

examination of conditioning influences. Justice Quarterly. 17(4), 753 —
778.

 

Felsom, R. (1996). Big people hit little people: Sex differences in physical power

and interpersonal violence. Criminology 34(4). 33—52.
Fitzpatrick, K. M. (1999). Violent victimization among America’s school children.
Journal of Interpersonal Violence. 14(10). 1055 — 1070.
Han, K.H. (2000, February 15). 5517—01] 9814-3- 21] ’8 2}. [Bullying in School].
News Maker, 161, p. 46 - 50
Kaufman, P., Chen, X., Choy, S.P., Ruddy, S.A. Miller, A.K., Fleury, J.K., Chandler,

K.A., Rand, M.R., Klaus, P., & Planty, M.G. Indicators of School Crime

 

and Safety. (2000). US. Department of Education and Justice. NCES
2001-017/NCJ-184176. Washington, D.C.:2000.

Kim, J.H.,&Kim, E.K., (1996). itflfic—E 2894 ’éEHSZl EH31 [Violence

around School Environment and Treatment]. Seoul, Korea: KIC

Korean Institute of Criminology. (1999). Delinquency and Prevention, Seoul:

Korea

Lawrence, R. (1991). School performance, peers and delinquency: Implication for

juvenile justice. Juvenile and Family Court Journal. 42(3). 59 - 69.

 

. (1998). School Crime and Juvenile Justice. New York, NY: Oxford
University.

86

Lee. Y.H. (2000). Delinquent Youths. The Journal of Juvenile Delinquency. 12(3),
36-52

Leone, P.E., Mayer, M.J., Malmgren,K., & Meisel, SM, (2000). Focus on
Exceptional Children, 33(1), 1-20.

Mazerolle, P. (1998). Gender, General Strain, and Delinquency: An Empirical
Examination. Justice Quarterly. 15(1). 65-91

Merton, R.K. (1938). Social Structure and anomie. American Sociological Review

32672-682.

Park, Y.H. (2000, February 15). 8'39] 511% 211E. [School system in Korea].

Korea Times. 32, p. 4 - 5.
Piquero, N.L., & Sealok, MD. (2000) Generalizing general strain theory: An
examination of an offending population. Justice Quarterly 17(3) 449-484
Small, M. & Tetrick, K.D. (2001). General strain theory and delinquency Juvenile
Justice and Delinquency Prevention, 8(1). 3-12.
Shwartz, D, Farver, J.M., Chang, L. & Lee-Shin Y. (2002). Victimization in South
Korean children’s peer groups. Journal of Abnormal Child Psychology.
30(2), 113-128.
Stattin, H. & Kerr, M. (2000). Parental monitoring: A reinterpretation. QM
Development, 71(4). 1072-1085
Werner, E., & Stuart, H. (1995). Criminological Theory: An Analysis of Its
_Underlyinflssumptions. Harcourt Brace College Publishers Orlando, FL.
Wright, J.P. & Cullen, ET. (2001) Parental efficacy and delinquent behavior: Do
control and support matter? Criminology, 39(3), 677-705

87